Proefschrift Vermeer

8mm

8mm

Bone-site-specific responses to

bisphosphonates

Long bone and jaw compared

Botspecifieke reacties op

bisfosfonaten

Een vergelijking tussen pijpbeenderen en kaak

The research described in this thesis was carried out at the department of Oral Cell Biology

and Functional Anatomy at the Academic Centre for Dentistry Amsterdam (ACTA), University

of Amsterdam and VU University Amsterdam, MOVE Research Institute Amsterdam,

Amsterdam, the Netherlands.

Printing of this thesis was kindly sponsored by:

ACTA onderzoeksinstituut

Anna Fonds|NOREF te Leiden

BD Biosciences

ChipSoft B.V.

Media Cybernetics

Nederlands Bijwerkingen Fonds

Nederlandse Vereniging voor Calcium- en Botstofwisseling (NVCB)

Vrije Universiteit Amsterdam

Cover design: Ruth Visser, VisserVisuals Amsterdam, the Netherlands

Printed by: Gildeprint, Enschede

ISBN: 9789461087362

© Adriana Francisca Vermeer, Amsterdam 2014. All rights reserved.

No part of this thesis may be reproduced or transmitted in any form or by any means,

electronic or mechanical, including photocopying, recording, or otherwise, without written

permission of the author.

VRIJE UNIVERSITEIT

Bone-site-specific responses to bisphosphonates Long bone and jaw compared

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad Doctor aan

de Vrije Universiteit Amsterdam,

op gezag van de rector magnificus

prof.dr. F.A. van der Duyn Schouten,

in het openbaar te verdedigen

ten overstaan van de promotiecommissie

van de Faculteit der Tandheelkunde

op dinsdag 21 oktober 2014 om 11.45 uur

in het auditorium van de universiteit,

De Boelelaan 1105

door

Adriana Francisca Vermeer

geboren te Heeze

promotor: prof.dr. V. Everts

copromotor: dr.ir. T.J. de Vries

committee members: dr. B.C.J. van der Eerden

prof.dr. S. Gibbs

prof.dr. J. Klein-Nulend

prof.dr. J. de Lange

prof.dr. P.T.A. Lips

prof.dr. C.J.F. van Noorden

paranymphs: Alejandra Maria Ruiz Zapata

Irma Johanna Pauline Vermeer

Contents

Chapter 1. General introduction 9

Chapter 2. Jaw bone marrow-derived osteoclast precursors internalize more bisphosphonate than long-bone marrow precursors 21

Chapter 3. Zoledronic acid differently affects long bone and jaw bone turnover and induces molar root resorption in female mice 47

Chapter 4. The effect of bisphosphonates on human periodontal-ligament-fibroblast-mediated osteoclastogenesis 69

Chapter 5. Osteoclast fusion and fission 83

Chapter 6. Migration, fusion, and CXCL12-CXCR4-mediated chemoattraction of long bone and jaw osteoclast precursors 103

Chapter 7. General discussion 123

General summary 135

Algemene samenvatting 139

Acknowledgments - Dankwoord 143

About the author 149

Chapter 1.

General introduction

Chapter 1

10

Bone cells and remodeling

Bone is important to give the body support, as an attachment site for muscles, and to

maintain mineral homeostasis. Bone also surrounds the bone marrow that contains

hematopoietic progenitors and mesenchymal stem cells. To keep its strength, bone is

continuously renewed; a process called bone remodeling. Bone remodeling starts with bone

resorption, i.e. degradation of bone, which is performed by osteoclasts [1]. Remaining

demineralized bone fragments in the resorption pit are then ingested by bone-lining cells

[2], in what is also referred to as the reversal phase, which couples bone resorption and

bone formation [3]. Finally, osteoblasts deposit new matrix proteins and mineralize the

newly formed osteoid [4]. In response to mechanical loading, bone embedded osteocytes

regulate bone remodeling by influencing osteoclast and osteoblast activities [5,6]. In this

chapter, the role of bone cells in maintaining bone homeostasis is explained. The main focus

will be on osteoclasts and bisphosphonates; drugs that can be used to treat diseases that

are characterized by excessive bone resorption, such as osteoporosis and bone cancer.

Osteoclasts

Osteoclasts are multinucleated cells formed by fusion of mononuclear precursors from the

monocyte/macrophage lineage. These precursors are present in bone marrow and in blood,

and migrate towards bone when resorption is needed. Upon contact with osteoblasts,

osteoclast precursors are activated to fuse and to become osteoclasts [7]. Osteoblasts

express two essential cytokines that stimulate osteoclast formation: macrophage-colony

stimulating factor (M-CSF) [8] and receptor activator of nuclear factor κB ligand (RANKL)

[9,10]. M-CSF binds to its receptor c-Fms, which is present on the osteoclast precursors,

thereby inducing RANK expression and priming the precursors to differentiate into an

osteoclast [11]. RANK activation by RANKL is considered to be the main stimulus for fusion

of osteoclast precursors.

Osteoclast heterogeneity

Osteoclasts from different bone sites have different characteristics, a phenomenon called

osteoclast heterogeneity [12,13]. Osteoclast formation from different precursors was shown

to occur at a different velocity [14,15]. In vitro, osteoclasts from long bone precursors are

formed relatively early in time, whereas jaw osteoclasts are formed later. This can be

explained by the finding that long-bone marrow contains more rapidly-differentiating

myeloid blasts, whereas jaw marrow contains a relatively high number of monocytes that

need more time to differentiate into osteoclasts [14,15]. Therefore, the study by de Souza

Faloni et al. also shows that not only osteoclasts, but also the composition of the bone

marrows are bone-site specific. Furthermore, the osteoclasts that were formed from these


11

bone marrows were shown to have a different morphology [16]. Moreover, the resorption

machineries were shown to be bone-site specific. Osteoclasts derived from long bones, for

instance, use primarily cathepsin K for the digestion of bone matrix, whereas calvarial

osteoclasts use next to cathepsins also matrix metalloproteinases (MMPs) [17-19].

Bone resorption

During bone resorption (reviewed in [20]), the osteoclast first creates a sealing zone. As this

zone contains mainly F-actin, it is also called the ‘actin ring’. This ring also contains αvβ3

integrin, which tightly attaches to the bone. The area encircled by the actin ring is called the

resorption lacuna, where a ruffled border is created, where chloride is secreted by the

chloride channel ClC-7, and where protons are secreted by the proton pump vacuolar-type

H+-ATPase (V-ATPase). Acidification leads to demineralization of the bone. Subsequently,

matrix proteins are degraded by proteases such as cathepsin K. Degraded proteins are

partly ingested by osteoclasts, which secrete them at the functional secretory domain at the

basolateral membrane. However, some proteins remnants are left behind in the resorption

pit and cleared by bone-lining cells [2].

Bone-lining cells and the reversal phase

As their name suggests, bone-lining cells are the cells lining the bone. They are

osteoblast/fibroblast-like cells, and after attraction of osteoclast precursors, the bone-lining

cells migrate away from each other to make space for the osteoclast precursors [21]. This

phenomenon was also seen when periodontal ligament fibroblasts were cultured together

with osteoclast precursors [22,23]. Osteoclasts can also transmigrate through a layer of

osteoblasts, implying that osteoclasts can migrate underneath the bone-lining cells to the

site of bone resorption [24]. After bone resorption, bone-lining cells clear the collagen

fragments that have been left behind in the resorption pit [2]. Without this clearance of the

resorption pit, new bone formation does not occur. Because this is called the reversal phase

of bone remodeling, bone-lining cells are also called reversal cells.

Osteoblasts and bone formation

Osteoblasts are bone-forming cells derived from mesenchymal stem cells in the bone

marrow. Runt-related transcription factor 2 (Runx2) was shown to be an essential

transcription factor during osteoblast differentiation [25]. Mature osteoblasts synthesize new

bone called osteoid by excreting proteins such as collagen type I, osteonectin, bone

sialoprotein, and osteopontin. These and other matrix proteins might play a role in bone

mineralization, which is also carried out by osteoblasts. Alkaline phosphatases play a role in

this process by increasing the local phosphate concentration [26].

Chapter 1

12

By creating a feedback loop towards osteoclasts through M-CSF and RANKL

expression, osteoblasts are also regulating bone homeostasis. Interestingly, they also

express osteoprotegerin (OPG), which inhibits RANK-mediated stimulation of

osteoclastogenesis by competitively binding to its ligand, RANKL [27]. Other cell types such

as periodontal ligament fibroblasts and osteocytes can also stimulate osteoclastogenesis

[28-30].

Osteocytes

Osteocytes belong to the osteoblast lineage and have become deeply embedded in lacunae

in the bone matrix. Thin cytoplasmic extensions protrude from the cell in canaliculi. This

system forms a network which is surrounded by canalicular fluid. Through the canalicular

network, osteocytes are connected to each other by gap junctions [31]. In response to

mechanical loading, they regulate bone homeostasis [6,32]. As a result of loading, they

stimulate osteoblasts to form new bone. On the other hand, when bone is unloaded, the

osteocytes die and stimulate osteoclastogenesis, thereby inducing bone degradation [30].

Recently, it has become clear that osteocytes can, next to osteoblasts, also express RANKL

and stimulate osteoclastogenesis during bone remodeling in vivo [33,34].

Bisphosphonates

In healthy bones, bone resorption and bone formation are in balance. When this balance is

disturbed and directed towards bone resorption, too much bone is degraded, making it

vulnerable to fracture. Such an imbalance occurs for instance in osteoporosis and cancers

that have metastasized to the bone. Those diseases are commonly treated with

bisphosphonates (BPs), which induce apoptosis and inhibit osteoclast activity. Due to their

high affinity for calcium, BPs rapidly bind to bone after administration [35]. During bone

resorption they are released into the resorption lacuna, where they become available for

uptake by the osteoclast. BPs therefore mainly act on osteoclasts in vivo; in vitro, they can

also be toxic to other cells such as osteoblasts [36], and reduce viability of macrophages

[37,38] and periodontal ligament fibroblasts [39,40]. Yet, in vivo BPs were also shown to be

taken up by monocytes [41], and they reduced the number of osteoclast precursors in

human peripheral blood [42,43]. Interestingly, anti-apoptotic effects on osteocytes and

osteoblasts with low concentrations of BPs were also reported [44], and were shown to be

mediated by connexin 43 [45].

Two groups of BPs are known and characterized by the presence or absence of

nitrogen (N), each using its own mechanism to inhibit osteoclast activity. BPs such as

clodronate, which do not contain nitrogen (non-N-BPs), are converted into a non-


13

hydrolyzable form of ATP, leading to apoptosis [46,47]. Nitrogen-containing BPs act through

a more complicated mechanism that is described below.

Nitrogen-containing bisphosphonates

Nitrogen-containing BPs (N-BPs) such as pamidronate, risedronate, and zoledronic acid,

inhibit the enzyme farnesyl pyrophosphate synthase (FPPS), thereby preventing protein

prenylation [48,49]. Prenylation is a post-translational protein modification, during which a

lipid group is added to the protein. Prenylation is important for the homing of small GTPases

to the cell membrane, where they play a role in cytoskeletal organization, migration, cell

adhesion and cell survival [50]. In osteoclasts the formation of the actin ring is disturbed by

N-BPs, leading to diminished osteoclast function.

As well as diminishing protein prenylation, inhibition of FPPS by N-BPs leads to

accumulation of its substrate, isopentenyl diphosphate (IPP). This is converted into

triphosphoric acid 1-adenosin-5-yl ester 3-(3-methylbut-3-enyl) ester (ApppI) leading to

apoptosis [51]. Thus, as well as diminishing osteoclast survival and activity by inhibiting

small GTPase function, N-BPs can also induce apoptosis directly.

Osteonecrosis of the jaw

A rare, but serious side effect in patients receiving a high-dose BP is osteonecrosis of the

jaw (ONJ). This means that necrotic bone has been exposed due to soft tissue damage, not

being the result of radiotherapy (i.e. osteoradionecrosis) [52]. ONJ is most common after

treatment with the N-BPs zoledronate and pamidronate, and can occur both in the mandible

and in the maxilla [53]. Prevalence is highly dependent on the dose and is estimated

between 0.1 and 10% for people that receive high doses for cancer treatment [53,54].

People that are treated for osteoporosis receive a relatively low dose and the incidence of

ONJ is lower. Invasive dental treatment, e.g. a tooth extraction, seems to increase the

chance to develop ONJ [55].

Several hypotheses have been proposed to explain the etiology of BP-related ONJ

(reviewed in [54,56]). Infection, oversuppression of bone turnover, and toxicity of BPs to

cells other than osteoclasts are the most commonly described. How these pathways may

interact in the onset of ONJ is depicted in Figure 1, however, its exact mechanism is

currently unknown. To mimic BP-induced ONJ, animal models were developed in which it

was shown that BP administration resulted in necrotic jaw bone, however, either

immunosuppressive agents were used in conjunction [57,58] or periodontitis was induced

[59,60]. Thus, BP treatment alone was not enough to induce ONJ in these animal models

Chapter 1

14

and its pathogenesis remains unclear. Also, it is not known why especially jaw bones are

affected.

Figure 1. Current hypotheses for BP-related ONJ (squares). ONJ (blue square) consists of 3 events

shown in the squares within the blue box. Green arrows represent positive effects, red blunted arrows

show inhibitions. The blue arrows point out a result of the inhibitions indicated in red. The dashed

arrows with question marks are speculative, since positive effects of BPs on osteoblasts and osteocytes

have also been shown. It is not clear how the inhibition of BPs on osteoclasts on one hand, and a

stimulation of osteoclasts by an infection on the other, can contribute to ONJ.


15

Hypothesis and thesis outline

We hypothesized that ONJ, the bone-site specific, negative effect of BPs, is related to the

phenomenon of osteoclast heterogeneity. Therefore, we investigated whether BPs can have

a different effect on long bone and jaw osteoclasts and their precursors.

In chapter 2, we investigated the internalization of BPs by long bone and jaw

osteoclast precursors and the effect of BPs on long bone and jaw osteoclastogenesis and

apoptosis. These in vitro studies with mouse cells were followed up by an in vivo study that

is described in chapter 3. We subjected female C57BL/6J mice to weekly injections of

zoledronic acid and investigated its effect on osteoclasts and bone formation markers in the

jaw and long bones. This study revealed that BPs were able to induce osteoclast formation

at the molar root. To investigate whether this stimulation may be induced by periodontal

ligament fibroblasts that were treated with BPs we used human cells (chapter 4). In order

to study the effect of BPs on osteoclast formation, time-lapse microscopy was shown to be a

useful tool to study the fusion of long bone osteoclast precursors and multinucleated

osteoclasts (chapter 5). Subsequently, we extended this in chapter 6, where we studied

multiple steps of osteoclastogenesis, i.e. proliferation, migration, and fusion, from long bone

and jaw bone marrow cells. We also analyzed the expression of genes involved in those

processes and investigated the effect of BPs on migration. Finally, we discuss our major

findings in chapter 7 and propose a new hypothesis on how the effect of BPs on different

osteoclasts may explain the pathogenesis of ONJ.

Chapter 1

16

References

1. Boyce BF, Rosenberg E, de Papp AE, Duong le T. The osteoclast, bone remodelling and

treatment of metabolic bone disease. European journal of clinical investigation.

2012;42(12):1332-41.

2. Everts V, Delaisse JM, Korper W, Jansen DC, Tigchelaar-Gutter W, Saftig P, et al. The bone

lining cell: its role in cleaning Howship's lacunae and initiating bone formation. J Bone Miner

Res. 2002;17(1):77-90.

3. Andersen TL, Abdelgawad ME, Kristensen HB, Hauge EM, Rolighed L, Bollerslev J, et al.

Understanding coupling between bone resorption and formation: are reversal cells the

missing link? Am J Pathol. 2013;183(1):235-46.

4. Harada S, Rodan GA. Control of osteoblast function and regulation of bone mass. Nature.

2003;423(6937):349-55.

5. Bonewald LF, Johnson ML. Osteocytes, mechanosensing and Wnt signaling. Bone.

2008;42(4):606-15.

6. Klein-Nulend J, Bakker AD, Bacabac RG, Vatsa A, Weinbaum S. Mechanosensation and

transduction in osteocytes. Bone. 2013;54(2):182-90.

7. Udagawa N, Takahashi N, Akatsu T, Tanaka H, Sasaki T, Nishihara T, et al. Origin of

osteoclasts: mature monocytes and macrophages are capable of differentiating into

osteoclasts under a suitable microenvironment prepared by bone marrow-derived stromal

cells. Proc Natl Acad Sci U S A. 1990;87(18):7260-4.

8. Wiktor-Jedrzejczak W, Bartocci A, Ferrante AW, Jr., Ahmed-Ansari A, Sell KW, Pollard JW, et

al. Total absence of colony-stimulating factor 1 in the macrophage-deficient osteopetrotic

(op/op) mouse. Proc Natl Acad Sci U S A. 1990;87(12):4828-32.

9. Lacey DL, Timms E, Tan HL, Kelley MJ, Dunstan CR, Burgess T, et al. Osteoprotegerin ligand

is a cytokine that regulates osteoclast differentiation and activation. Cell. 1998;93(2):165-76.

10. Yasuda H, Shima N, Nakagawa N, Yamaguchi K, Kinosaki M, Mochizuki S, et al. Osteoclast

differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is

identical to TRANCE/RANKL. Proc Natl Acad Sci U S A. 1998;95(7):3597-602.

11. Arai F, Miyamoto T, Ohneda O, Inada T, Sudo T, Brasel K, et al. Commitment and

differentiation of osteoclast precursor cells by the sequential expression of c-Fms and

receptor activator of nuclear factor kappaB (RANK) receptors. The Journal of experimental

medicine. 1999;190(12):1741-54.

12. Everts V, de Vries TJ, Helfrich MH. Osteoclast heterogeneity: lessons from osteopetrosis and

inflammatory conditions. Biochim Biophys Acta. 2009;1792(8):757-65.

13. Henriksen K, Bollerslev J, Everts V, Karsdal MA. Osteoclast activity and subtypes as a function

of physiology and pathology--implications for future treatments of osteoporosis. Endocr Rev.

2011;32(1):31-63.

14. de Souza Faloni AP, Schoenmaker T, Azari A, Katchburian E, Cerri PS, de Vries TJ, et al. Jaw

and long bone marrows have a different osteoclastogenic potential. Calcif Tissue Int.

2011;88(1):63-74.


17

15. de Vries TJ, Schoenmaker T, Hooibrink B, Leenen PJ, Everts V. Myeloid blasts are the mouse

bone marrow cells prone to differentiate into osteoclasts. J Leukoc Biol. 2009;85(6):919-27.

16. Azari A, Schoenmaker T, de Souza Faloni AP, Everts V, de Vries TJ. Jaw and long bone

marrow derived osteoclasts differ in shape and their response to bone and dentin. Biochem

Biophys Res Commun. 2011;409(2):205-10.

17. Everts V, Korper W, Hoeben KA, Jansen ID, Bromme D, Cleutjens KB, et al. Osteoclastic bone

degradation and the role of different cysteine proteinases and matrix metalloproteinases:

differences between calvaria and long bone. J Bone Miner Res. 2006;21(9):1399-408.

18. Everts V, Korper W, Jansen DC, Steinfort J, Lammerse I, Heera S, et al. Functional

heterogeneity of osteoclasts: matrix metalloproteinases participate in osteoclastic resorption

of calvarial bone but not in resorption of long bone. FASEB J. 1999;13(10):1219-30.

19. Shorey S, Heersche JN, Manolson MF. The relative contribution of cysteine proteinases and

matrix metalloproteinases to the resorption process in osteoclasts derived from long bone and

scapula. Bone. 2004;35(4):909-17.

20. Teitelbaum SL. Bone resorption by osteoclasts. Science. 2000;289(5484):1504-8.

21. Perez-Amodio S, Beertsen W, Everts V. (Pre-)osteoclasts induce retraction of osteoblasts

before their fusion to osteoclasts. J Bone Miner Res. 2004;19(10):1722-31.

22. Bloemen V, Schoenmaker T, de Vries TJ, Everts V. IL-1beta favors osteoclastogenesis via

supporting human periodontal ligament fibroblasts. J Cell Biochem. 2011;112(7):1890-7.

23. de Vries TJ, Schoenmaker T, Wattanaroonwong N, van den Hoonaard M, Nieuwenhuijse A,

Beertsen W, et al. Gingival fibroblasts are better at inhibiting osteoclast formation than

periodontal ligament fibroblasts. J Cell Biochem. 2006;98(2):370-82.

24. Saltel F, Chabadel A, Zhao Y, Lafage-Proust MH, Clezardin P, Jurdic P, et al. Transmigration: a

new property of mature multinucleated osteoclasts. J Bone Miner Res. 2006;21(12):1913-23.

25. Ducy P, Zhang R, Geoffroy V, Ridall AL, Karsenty G. Osf2/Cbfa1: a transcriptional activator of

osteoblast differentiation. Cell. 1997;89(5):747-54.

26. Robey PG, Boskey AL. Extracellular matrix and biomineralization of bone. In: Favus MJ,

editor. Primer on the metabolic bone diseases and disorders of mineral metabolism. j. b. 5th

ed. Washington, DC: ASBMR; 2003. p. 38-46.

27. Simonet WS, Lacey DL, Dunstan CR, Kelley M, Chang MS, Luthy R, et al. Osteoprotegerin: a

novel secreted protein involved in the regulation of bone density. Cell. 1997;89(2):309-19.

28. Kanzaki H, Chiba M, Shimizu Y, Mitani H. Dual regulation of osteoclast differentiation by

periodontal ligament cells through RANKL stimulation and OPG inhibition. Journal of dental

research. 2001;80(3):887-91.

29. Kulkarni RN, Bakker AD, Everts V, Klein-Nulend J. Inhibition of osteoclastogenesis by

mechanically loaded osteocytes: involvement of MEPE. Calcif Tissue Int. 2010;87(5):461-8.

30. Tatsumi S, Ishii K, Amizuka N, Li M, Kobayashi T, Kohno K, et al. Targeted ablation of

osteocytes induces osteoporosis with defective mechanotransduction. Cell metabolism.

2007;5(6):464-75.

31. Bonewald LF. The amazing osteocyte. J Bone Miner Res. 2011;26(2):229-38.

32. Burger EH, Klein-Nulend J, van der Plas A, Nijweide PJ. Function of osteocytes in bone--their

role in mechanotransduction. The Journal of nutrition. 1995;125(7 Suppl):2020S-3S.

Chapter 1

18

33. Nakashima T, Hayashi M, Fukunaga T, Kurata K, Oh-Hora M, Feng JQ, et al. Evidence for

osteocyte regulation of bone homeostasis through RANKL expression. Nature medicine.

2011;17(10):1231-4.

34. Xiong J, Onal M, Jilka RL, Weinstein RS, Manolagas SC, O'Brien CA. Matrix-embedded cells

control osteoclast formation. Nature medicine. 2011;17(10):1235-41.

35. Russell RG, Watts NB, Ebetino FH, Rogers MJ. Mechanisms of action of bisphosphonates:

similarities and differences and their potential influence on clinical efficacy. Osteoporos Int.

2008;19(6):733-59.

36. Idris AI, Rojas J, Greig IR, Van't Hof RJ, Ralston SH. Aminobisphosphonates cause osteoblast

apoptosis and inhibit bone nodule formation in vitro. Calcif Tissue Int. 2008;82(3):191-201.

37. Coxon FP, Thompson K, Roelofs AJ, Ebetino FH, Rogers MJ. Visualizing mineral binding and

uptake of bisphosphonate by osteoclasts and non-resorbing cells. Bone. 2008;42(5):848-60.

38. Moreau MF, Guillet C, Massin P, Chevalier S, Gascan H, Basle MF, et al. Comparative effects

of five bisphosphonates on apoptosis of macrophage cells in vitro. Biochemical pharmacology.

2007;73(5):718-23.

39. Agis H, Blei J, Watzek G, Gruber R. Is zoledronate toxic to human periodontal fibroblasts?

Journal of dental research. 2010;89(1):40-5.

40. Correia Vde F, Caldeira CL, Marques MM. Cytotoxicity evaluation of sodium alendronate on

cultured human periodontal ligament fibroblasts. Dental traumatology : official publication of

International Association for Dental Traumatology. 2006;22(6):312-7.

41. Roelofs AJ, Coxon FP, Ebetino FH, Lundy MW, Henneman ZJ, Nancollas GH, et al. Fluorescent

risedronate analogues reveal bisphosphonate uptake by bone marrow monocytes and

localization around osteocytes in vivo. J Bone Miner Res. 2010;25(3):606-16.

42. D'Amelio P, Grimaldi A, Di Bella S, Tamone C, Brianza SZ, Ravazzoli MG, et al. Risedronate

reduces osteoclast precursors and cytokine production in postmenopausal osteoporotic

women. J Bone Miner Res. 2008;23(3):373-9.

43. D'Amelio P, Grimaldi A, Cristofaro MA, Ravazzoli M, Molinatti PA, Pescarmona GP, et al.

Alendronate reduces osteoclast precursors in osteoporosis. Osteoporos Int.

2010;21(10):1741-50.

44. Plotkin LI, Weinstein RS, Parfitt AM, Roberson PK, Manolagas SC, Bellido T. Prevention of

osteocyte and osteoblast apoptosis by bisphosphonates and calcitonin. The Journal of clinical

investigation. 1999;104(10):1363-74.

45. Plotkin LI, Lezcano V, Thostenson J, Weinstein RS, Manolagas SC, Bellido T. Connexin 43 is

required for the anti-apoptotic effect of bisphosphonates on osteocytes and osteoblasts in

vivo. J Bone Miner Res. 2008;23(11):1712-21.

46. Frith JC, Monkkonen J, Auriola S, Monkkonen H, Rogers MJ. The molecular mechanism of

action of the antiresorptive and antiinflammatory drug clodronate: evidence for the formation

in vivo of a metabolite that inhibits bone resorption and causes osteoclast and macrophage

apoptosis. Arthritis Rheum. 2001;44(9):2201-10.

47. Frith JC, Monkkonen J, Blackburn GM, Russell RG, Rogers MJ. Clodronate and liposome-

encapsulated clodronate are metabolized to a toxic ATP analog, adenosine 5'-(beta, gamma-


19

dichloromethylene) triphosphate, by mammalian cells in vitro. J Bone Miner Res.

1997;12(9):1358-67.

48. Luckman SP, Hughes DE, Coxon FP, Graham R, Russell G, Rogers MJ. Nitrogen-containing

bisphosphonates inhibit the mevalonate pathway and prevent post-translational prenylation of

GTP-binding proteins, including Ras. J Bone Miner Res. 1998;13(4):581-9.

49. van Beek E, Pieterman E, Cohen L, Lowik C, Papapoulos S. Farnesyl pyrophosphate synthase

is the molecular target of nitrogen-containing bisphosphonates. Biochem Biophys Res

Commun. 1999;264(1):108-11.

50. Itzstein C, Coxon FP, Rogers MJ. The regulation of osteoclast function and bone resorption by

small GTPases. Small GTPases. 2011;2(3):117-30.

51. Monkkonen H, Auriola S, Lehenkari P, Kellinsalmi M, Hassinen IE, Vepsalainen J, et al. A new

endogenous ATP analog (ApppI) inhibits the mitochondrial adenine nucleotide translocase

(ANT) and is responsible for the apoptosis induced by nitrogen-containing bisphosphonates.

Br J Pharmacol. 2006;147(4):437-45.

52. Khosla S, Burr D, Cauley J, Dempster DW, Ebeling PR, Felsenberg D, et al. Bisphosphonate-

associated osteonecrosis of the jaw: report of a task force of the American Society for Bone

and Mineral Research. J Bone Miner Res. 2007;22(10):1479-91.

53. Abu-Id MH, Warnke PH, Gottschalk J, Springer I, Wiltfang J, Acil Y, et al. "Bis-phossy jaws" -

high and low risk factors for bisphosphonate-induced osteonecrosis of the jaw. J

Craniomaxillofac Surg. 2008;36(2):95-103.

54. Reid IR. Osteonecrosis of the jaw: who gets it, and why? Bone. 2009;44(1):4-10.

55. Filleul O, Crompot E, Saussez S. Bisphosphonate-induced osteonecrosis of the jaw: a review

of 2,400 patient cases. Journal of cancer research and clinical oncology. 2010;136(8):1117-

24.

56. Otto S, Hafner S, Mast G, Tischer T, Volkmer E, Schieker M, et al. Bisphosphonate-related

osteonecrosis of the jaw: is pH the missing part in the pathogenesis puzzle? J Oral Maxillofac

Surg. 2010;68(5):1158-61.

57. Bi Y, Gao Y, Ehirchiou D, Cao C, Kikuiri T, Le A, et al. Bisphosphonates cause osteonecrosis of

the jaw-like disease in mice. Am J Pathol. 2010;177(1):280-90.

58. Kikuiri T, Kim I, Yamaza T, Akiyama K, Zhang Q, Li Y, et al. Cell-based immunotherapy with

mesenchymal stem cells cures bisphosphonate-related osteonecrosis of the jaw-like disease in

mice. J Bone Miner Res. 2010;25(7):1668-79.

59. Aghaloo TL, Kang B, Sung EC, Shoff M, Ronconi M, Gotcher JE, et al. Periodontal disease and

bisphosphonates induce osteonecrosis of the jaws in the rat. J Bone Miner Res.

2011;26(8):1871-82.

60. Aguirre JI, Akhter MP, Kimmel DB, Pingel JE, Williams A, Jorgensen M, et al. Oncologic doses

of zoledronic acid induce osteonecrosis of the jaw-like lesions in rice rats (Oryzomys palustris)

with periodontitis. J Bone Miner Res. 2012;27(10):2130-43.

Chapter 2.

Jaw bone marrow-derived

osteoclast precursors internalize

more bisphosphonate than long-

bone marrow precursors

Jenny A.F. Vermeer

Ineke D.C. Jansen

Matangi Marthi

Fraser P. Coxon

Charles E. McKenna

Shuting Sun

Teun J. de Vries

Vincent Everts

Published in: Bone. 2013 Nov;57(1):242-51

Chapter 2

22

Abstract

Bisphosphonates (BPs) are widely used in the treatment of several bone diseases, such as

osteoporosis and cancers that have metastasized to bone, by virtue of their ability to inhibit

osteoclastic bone resorption. Previously, it was shown that osteoclasts present at different

bone sites have different characteristics. We hypothesized that BPs could have distinct

effects on different populations of osteoclasts and their precursors, for example as a result

of a different capacity to endocytose the drugs. To investigate this, bone marrow cells were

isolated from jaw and long bone from mice and the cells were primed to differentiate into

osteoclasts with the cytokines M-CSF and RANKL. Before fusion occurred, cells were

incubated with fluorescein-risedronate (FAM-RIS) for 4 or 24 h and uptake was determined

by flow cytometry. We found that cultures obtained from the jaw internalized 1.7 to 2.5

times more FAM-RIS than long bone cultures, both after 4 and 24 h, and accordingly jaw

osteoclasts were more susceptible to inhibition of prenylation of Rap1a after treatment with

BPs for 24 h. Surprisingly, differences in BP uptake did not differentially affect

osteoclastogenesis. This suggests that jaw osteoclast precursors are less sensitive to

bisphosphonates after internalization. This was supported by the finding that gene

expression of the anti-apoptotic genes Bcl-2 and Bcl-xL was higher in jaw cells than long

bone cells, suggesting that the jaw cells might be more resistant to BP-induced apoptosis.

Our findings suggest that bisphosphonates have distinct effects on both populations of

osteoclast precursors and support previous findings that osteoclasts and precursors are

bone-site specific. This study may help to provide more insights into bone-site-specific

responses to bisphosphonates.

The effect of BPs on long bone and jaw osteoclasts and their precursors

23

Introduction

During the process of bone remodeling, bone is resorbed by osteoclasts. These

multinucleated, tartrate-resistant acid phosphatase (TRACP)-expressing cells are formed by

the fusion of mononuclear precursors present in the bone marrow. In diseases associated

with excessive bone resorption, such as osteoporosis and cancers that have metastasized to

the bone, bisphosphonates (BPs) are widely used as a treatment to reduce bone resorption

by osteoclasts, thereby improving bone quality and reducing fracture risk.

Two different classes of BPs can be distinguished by the presence or absence of an

amino group. Each class acts by a distinct mechanism to inhibit osteoclast function. After

uptake, non-N-BPs, such as clodronate are transformed into non-hydrolysable forms of ATP,

resulting in apoptosis [1,2]. Nitrogen-containing BPs (N-BPs) such as zoledronic acid,

pamidronate, and risedronate, inhibit farnesyl pyrophosphate synthase (FPPS), thereby

preventing prenylation of small GTPases [3,4]. Prenylation is a post-translational lipid

modification important for cellular localization and function of small GTPases. Disruption of

this process and inhibition of small GTPase function are therefore thought to be the

mechanism by which N-BPs lead to dysfunctional osteoclasts [5]. Besides their effect on

prenylation, N-BPs can induce apoptosis directly through accumulation of the substrate of

FPPS, IPP, which can be converted to the toxic ATP analogue triphosphoric acid 1-adenosin-

5-yl ester 3-(3-methylbut-3-enyl) ester (ApppI) [6]. In vitro, BPs exert effects on numerous

different cell types [7-9]. In vivo, however, BPs bind rapidly and avidly to bone mineral, and

as a result are thought to selectively affect osteoclasts, which release and take up the BP

during resorption. Nonetheless, uptake of BPs by monocytes in vivo has also been reported

[10]. A side-effect associated with bisphosphonates is osteonecrosis of the jaw (ONJ), which

is defined as persistence of exposed bone in the maxilla or mandible for at least 8 weeks

[11]. The prevalence is very low in patients receiving oral BPs for the treatment of

osteoporosis, however it is estimated to range between 1 and 15% in cancer patients who

receive high intravenous doses of the N-BPs pamidronate or zoledronic acid [12]. Several

hypotheses regarding the etiology of BP-related ONJ have been proposed [12]. First,

suppression of bone turnover and subsequent increase in bone mass could lead to

avascularization followed by necrosis. Since bone turnover in the jaw is taken to be higher

than in other bones [13], this effect might be more pronounced in the jaw. Next, high bone

turnover and thus high local osteoclast activity could also lead to excessive release of BP

from the bone and BP exposure and subsequent uptake by neighboring cells [8]. This might

lead to death of cells other than osteoclasts such as other bone cells, endothelial cells, and

bone marrow cells. Since osteoclasts are more active at a low pH [14], this effect could be

even more pronounced in an acidic environment created by an infection or by the BP itself

[12,15].

Chapter 2

24

A potential explanation for ONJ could lie in the sensitivity of jaw osteoclasts and

their precursors to bisphosphonates. Previously, it was shown that osteoclasts derived from

various bone sites have different characteristics (reviewed in [16,17]). We hypothesize that

precursors and osteoclasts from the jaw are more sensitive to bisphosphonates than those

from long bone, resulting in excessive suppression of bone turnover in the jaw compared to

other skeletal sites.

To investigate the effect of BPs on different osteoclast populations, we isolated jaw

and long-bone marrow from mice and studied the effect of pamidronate on

osteoclastogenesis in vitro. We also quantitatively and qualitatively assessed internalization

of the fluorescently labeled N-BP risedronate (FAM-RIS) by jaw and long-bone marrow cells.

This study will provide more insight into the differences between bone-site-specific

osteoclasts and precursors, and may lead to a better understanding of the association

between certain BPs and osteonecrosis of the jaw.


25

Materials and methods

Synthesis of FAM-RIS

FAM-RIS was synthesized from a commercially available mixture of 5- and 6-

carboxyfluorescein (5(6)-carboxyfluorescein) according to the published procedure [18]. The

sample was purified by TLC on silica gel then by preparative gradient HPLC (RP, A: 10%

MeOH, 0.1 M TEAC, pH 7.0; B: 75% MeOH in 0.1 M TEAC, pH 7.5; A to 40% of B in 12 min

then to 70% B in 100 min). After rotary evaporation to remove the solvent, the residue was

dissolved in water and the concentration determined by UV-VIS (λmax = 492 nm, ε = 72000

M-1cm-1). The sample was then divided into 0.5 mL aliquots (glass vials) which were then

lyophilized and stored at -20 °C. The compound was characterized by 1H and 31P NMR

(single peak); high resolution MS (M+), calcd 715.1089 m/z; found 715.1055; and UV-VIS

and fluorescent emission spectroscopy (λmax, = 518 nm).

Bone marrow isolation

Bone marrow cells from long bones and mandibles of six-week-old male C57BL/6J mice

(Harlan, Horst, The Netherlands) were isolated and used for osteoclastogenesis

experiments. Mice were injected intraperitoneally with a lethal dose of sodium pentobarbital

(0.1 mL Euthestate, Ceva Sante Animale, Naaldwijk, The Netherlands). Tibiae, femurs, and

mandibles were removed and were kept in α-minimal essential medium (α-MEM; GIBCO,

Paisley, UK) supplemented with 5% fetal calf serum (FCS; Hyclone, Logan, UT, USA), 1%

antibiotics (100 U/mL penicillin, 100 µg/mL streptomycin, and 250 ng/mL amphotericin B;

antibiotic antimycotic solution; Sigma, St. Louis, MO, USA), and heparin (170 IE/mL; Leo

Pharmaceutical Products B.V., Weesp, The Netherlands). Soft tissue, condyles and the

incisors were removed from the mandibles, leaving the jaw bone containing the molars and

adjacent bone marrow space [19]. Bones were crushed in mortars and flushed with medium

without heparin through a 21-gauge needle and filtered with a 70-µm pore-size cell strainer

(Falcon/Becton Dickinson, Franklin Lakes, NJ, USA). Animal experiments were approved by

the Animal Welfare Committee of the VU University (Amsterdam, The Netherlands).

Osteoclastogenesis

Freshly isolated bone marrow cells were seeded in 96-well plates (Cellstar, Greiner Bio-One,

Monroe, NC, USA) at a density of 1 x 105 cells per well. Cells were seeded on plastic or on

bovine cortical-long-bone slices. Osteoclastogenesis was induced with 150 µL medium

containing 30 ng/mL recombinant mouse macrophage-colony stimulating factor (rmM-CSF;

R&D Systems, Minneapolis, MN, USA) and 20 ng/mL recombinant mouse receptor activator

of nuclear factor κB ligand (rmRANKL; R&D Systems). Stock solutions of 100 mM

Chapter 2

26

pamidronate (Sigma) in deionized water, 100 mM risedronate (LKT Laboratories, Inc., St.

Paul, MN, USA) in PBS, 10 mM risedronate conjugated with 5(6)-carboxyfluorescein (FAM-

RIS) in PBS (as described in 2.1) [18], and 10 mg/mL FITC-dextran (Mw: 10000; Life

Technologies, Molecular Probes, Carlsbad, CA, USA) in PBS were prepared, filter sterilized

and stored at -20ºC until use. Pamidronate (1-100 µM) was added to the culture medium

during an early (days 0-3) or a later (days 3-6) stage of differentiation. For some

experiments, bone slices were incubated with pamidronate (10-100 µM) for 20 h and PAM

was washed away before seeding of the bone marrow. Culture medium with M-CSF and

RANKL, but without BPs was used as a control. Cultures were maintained in a humidified

atmosphere at 37ºC and 5% CO2. The culture medium was replaced every three days. At

the end of the culture period, cells were rinsed with PBS and fixed in 4% PBS-buffered

formaldehyde (used for TRACP staining and confocal microscopy). Bone slices were stored

in distilled water at 4ºC (used for Resorption Assay).

TRACP staining

Fixed cells on plastic or on bone slices were stained for TRACP using the leukocyte acid

phosphatase kit (Sigma). Nuclei were visualized with 4’6-diamidino-2-phenylindole

dihydrochloride (DAPI). The number of TRACP-positive cells with three or more nuclei was

assessed and expressed per cm2.

RNA isolation and real time quantitative PCR

RNA was isolated from cells cultured on bone slices using the RNeasy Mini Kit (Qiagen,

Hilden, Germany) according to manufacturer’s instructions. cDNA was synthesized using the

MBI Fermentas cDNA synthesis kit (Fermentas, Vilnius, Lithuania). Real time qPCR was

performed on an ABI Prism 7000 using SYBR Green mastermix (Applied Biosystems, Foster

City, CA, USA) and primers are listed in Table 1. Gene expression was divided by expression

of the housekeeping gene porphobilinogen deaminase (PBGD) for normalization.

Caspase 3/7-Glo Assay

Caspase-3/7 activity was measured 5 h after addition of 50 µM PAM at day 3 of culture in

the presence of M-CSF and RANKL. The Caspase 3/7-Glo Assay (Promega, Madison, WI,

USA) was performed according to manufacturer’s instructions and normalized for the

amount of DNA. DNA content was measured using the CyQUANT Cell Proliferation Assay

(Molecular Probes).


27

Table 1. Primer sequences used for real time quantitative PCR.

Gene Primer sequence 5’�3’

PBGD FW: AgTgATgAAAgATgggCAACt

RV: TCTggACCATCTTCTTgCTgA

TRACP FW: gACAAgAggTTCCAggAgACC

RV: gggCTggggAAgTTCCAg

DC-STAMP FW: TgTATCggCTCATCTCCTCCAT

RV: gACTCCTTgggTTCCTTgCTT

Bcl-2 FW: ggCATCTgCACACCTggAT

RV: AgACAgCCAggAgAAATCAAACA

Bcl-xL FW: CggCTgggACACTTTTgTg

RV: CAgTCATgCCCgTCAggAA

Bax FW: TggCAgCTgACATgTTTgCT

RV: TTTAgTgCACAgggCCTTgAg

Atp6v0d2 FW: CATTCCTTggAgCCCCTgAg

RV: TCTCTgTgAAACggCCCAgT

Cathepsin K FW: ACAgCAggATgTgggTgTTCA

RV: gCCgAgAgATTTCATCCACCT

F4/80 FW: gCACCAATgTACCAggCTCCTA

RV: gATCCTTTTgCAgTTgAAgTTTCC

MafB FW: AACgCgTCCAgCAgAAACAT

RV: CTCAggAgAggAggggCTgT

Irf8 FW: TgTCCCCgAggAAgAACAAA

RV: gCCACACTCCATCTCAggAACT

ALP FW: CTCATggAggCCTTTgTCTT

RV: CTCATgATgTCCgTggTCAA

Resorption Assay

To assess osteoclast activity, cells cultured on bone slices were lysed with deionized water

and 10% ammonia. Resorption pits were visualized with Coomassie Brilliant Blue as

described previously [20]. Micrographs were taken of pre-defined areas on each bone slice,

and four fields per bone slice were analyzed. Micrographs covered approximately one fifth of

each bone slice. The resorbed area was measured using Image Pro-Plus Software (Media

Cybernetics, Silver Spring, MD, USA) and expressed as percentage of the total surface area.

Chapter 2

28

Flow cytometry

Uptake of FAM-RIS (10 µM) or FITC-dextran (500 µg/mL) by osteoclast precursors (bone

marrow cells treated with M-CSF and RANKL for 3 days) was studied after 4 and 24 h of

incubation of the cells with the labeled bisphosphonates or after 4 h with dextran.

Incubation was performed at 4ºC (only 4 h) or 37ºC. After the incubation cells were rinsed

with PBS and removed from the surface using 0.05% trypsin-EDTA (Gibco) followed by

scraping the cells off the bottom of the wells. Washing and fixation were performed as

described in paragraph 2.3. Fluorescence was measured using flow cytometry (C6 Flow

Cytometer, Accuri, Cambs, UK). Cells treated with PBS were used to set background

fluorescence.

Immunostaining and confocal microscopy

To visualize BP uptake, bone marrow cells were cultured in the presence of FAM-RIS for 4 h

on glass chamber slides (Lab-Tek; Nunc, Naperville, IL, USA) and fixed with formaldehyde

at day 5 as described above. The plasma membrane of the cells was visualized with anti-

CD44. Cells were kept in the dark during the staining procedure. First, non-specific

background staining was blocked with 20% normal goat serum (Vector Laboratories,

Burlingame, CA, USA) for 30 min. Chamber slides were incubated overnight at 4°C with rat

anti-mouse CD44 antibody IM7.8.1 (2 µg/mL, Cedarlane Laboratories Ltd., Burlington, ON,

Canada) dissolved in PBS/BSA 1% (bovine serum albumin; Sigma). After washing 3 times

with PBS, wells were incubated for 1 h at room temperature with Alexa Fluor 647-labeled

goat anti-rat antibody (1:400, Life Technologies, Molecular Probes). Finally, nuclei were

visualized with propidium iodide (Sigma). Slides were stored at 4ºC in PBS until they were

analyzed by confocal laser scanning microscopy (Leica Microsystems GmbH, Wetzlar,

Germany).

To determine the activity per osteoclast, mouse bone marrow cells were cultured

on bovine cortical bone slices and fixed with formaldehyde after eight days as described

above. Resorption pits were visualized by staining collagen type I and the plasma

membranes of the cells were visualized with anti-CD44. First, non-specific background

staining was blocked with 20% normal goat serum (Vector Laboratories, Burlingame, CA,

USA) for 30 min. Bone slices were washed three times with PBS, followed by a one hour

incubation with rat anti-mouse CD44 antibody IM7.8.1 (see above) and rabbit anti-human

collagen I antibody (2 µg/ml, Abcam, Cambridge, UK). After another washing step, bone

slices were incubated for 1 h in the dark with Alexa Fluor 647-labelled goat anti-rat antibody

(1:400) and Alexa Fluor 488-labelled goat anti-rabbit antibody (1:400, both Invitrogen,

Molecular Probes, Carlsbad, CA, USA). Finally, nuclei were visualized with propidium iodide

(Sigma). All incubations were performed at an ambient temperature. Bone slices were


29

stored at 4ºC in PBS until they were analyzed by confocal laser scanning microscopy (Leica

Microsystems GmbH, Wetzlar, Germany). A representative area per bone slice was analyzed

for resorption and divided by the number of multinucleated cells as a measure for activity

per osteoclast.

Western blotting

Cells were lysed using radioimmunoassay precipitation buffer (RIPA; Enzo Life Sciences,

Stressgen, Ann Arbor, MI, USA) containing protease inhibitors (Roche Diagnostics,

Mannheim, Germany). Protein content was measured with the BCA Protein Assay (Thermo

Scientific, Pierce, Rockford, IL, USA). Equal amounts of protein were denatured and

separated on a NuPAGE 10% Bis-Tris gel (Life Technologies, Novex) under reducing

conditions. Proteins were transferred to a nitrocellulose membrane using the iBlot system

(Life Technologies, Novex). Background was blocked overnight at 4°C with PBS/BSA 1%

containing 0.1% Tween 20 (blocking buffer). Incubation with a polyclonal antibody

specifically binding to the unprenylated form of Rap1a (sc-1482, 1:200; Santa Cruz

Biotechnology, Santa Cruz, CA, USA) was done at ambient temperature for 3 h.

Subsequently, the blots were rinsed 3 times with blocking buffer followed by a 1 h

incubation step with HRP-labeled rabbit anti-goat antibody (Pierce). Chemiluminescent

signal was detected using SuperSignal West Pico Chemiluminescent Substrate (Thermo

Scientific) and the Biospectrum AC Imaging System and VisionWorksLS software (UVP,

Cambridge, UK). Gamma settings were optimized for clarity of the bands. After

chemiluminescent detection, blots were washed with blocking buffer and incubated for 1 h

at room temperature with mouse anti-β-actin antibody (1:5000, Sigma). Antibodies were

washed away by rinsing 3 times with blocking buffer. An alkaline phosphatase labeled

secondary antibody (Pierce) was incubated for 1 h and stained with NBT/BCIP according to

manufacturer’s instructions (Roche Diagnostics). Density of the bands was measured using

ImageJ software (National Institutes of Health, Bethesda, MD, USA)

Statistical analyses

The effect of BP treatment on long bone and jaw osteoclastogenesis was analyzed by

separate Friedman tests for long bone and jaw, followed by Dunn’s Multiple comparison

tests to compare the medians of BP treatment to controls. Gene expression after BP

treatment was analyzed with a Repeated Measures ANOVA test and post-hoc analyses were

done with Dunnett’s Multiple Comparison Tests. A one-sample t-test was used to analyze

caspase 3/7 activity. Differences between long-bone marrow cells and jaw cells regarding

caspase 3/7 activity and FAM-RIS and dextran uptake were analyzed with a paired t-test

(two-tailed). A paired t-test was also used to compare gene expression of Bcl-XL, Bcl-2, and

Chapter 2

30

Bax. Data are expressed as means of at least three measurements and SD. Differences were

considered significant when p<0.05.


31

Results

Jaw osteoclast precursors internalize more FAM-RIS than long bone osteoclast precursors

To gain insights into the effect of bisphosphonates on different subpopulations of osteoclast

precursors, we first assessed uptake of fluorescently labeled risedronate (FAM-RIS) by jaw

and long-bone marrow cultures by flow cytometry. FAM-RIS was added after 3 days of

culture in the presence of M-CSF and RANKL. At this time point the cultures were primed to

become osteoclasts but they contained mainly mononuclear osteoclast precursors and

expressed at least 40 times more TRACP mRNA than at day 1 of culture. Two cell

populations were identified based on their size (Figure 1A). The median fluorescence

intensity (MFI) of the small population was lower than the MFI of large cells. Jaw cultures

contained significantly more small cells than long bone cultures, both in the absence and in

the presence of FAM-RIS (Figure 1B). Internalization of bisphosphonates was seen in the

entire population in both small and large long bone and jaw cultures (not shown). Uptake

per cell, 4 and 24 h after addition of FAM-RIS, was assessed and expressed as a ratio of

fluorescence of FAM-RIS treated cells divided by the MFI of untreated controls. Examples of

the histograms of large long bone and jaw cells and how the ratios were calculated are

shown (Figure 1C). Jaw cells internalized significantly more FAM-RIS than long bone cells

(Figure 1D). This was observed for both small and large cells. This difference persisted over

a longer period; similar results were found after 24 h (Figure 1D, right panel).

FAM-RIS uptake at 4°C was investigated to assess passive passage through and

adsorption to the plasma membrane. Uptake at 4°C ranged from 3 to 16% of uptake at

37°C (not shown) and was similar for jaw cells and long bone cells. This suggests that the

higher BP uptake by jaw cells than by long bone cells was the result of mainly active

endocytosis. Uptake was further confirmed with confocal microscopy, where FAM-RIS was

found mainly intracellularly (Figure 1E).

Jaw and long bone osteoclast precursors internalize similar levels of dextran

Previously, Thompson et al. have shown in J774 cells that the most likely mechanism of BP

uptake is fluid-phase endocytosis [21]. To get more insight into the mechanism of uptake in

primary long bone and jaw bone marrow cells, internalization of the fluid-phase endocytosis

marker dextran was assessed. In contrast to uptake of BPs, jaw cells did not internalize

more dextran than long bone cells (Figure 2). This implies that the rate of fluid-phase

endocytosis is similar in those two types of osteoclast precursors and that increased

accumulation of BPs in jaw cells (Figure 1D) is therefore not the result of differences in

fluid-phase endocytosis.

Chapter 2

32

Figure 1. BP uptake by marrow cells of long bone and jaw cultured for 3 days in the presence of M-

CSF and RANKL. (A) Gates set on living cells to distinguish between small (P2) and big (P4) cells. (B)

The percentage of small cells is higher in jaw bone marrow cell cultures than in long-bone marrow cell

cultures. (C) Examples of histograms of large long bone (left) and jaw (right) cells. (D) Uptake was

higher by jaw cells than by long bone cells, both 4 (left) and 24 (right) h after addition at day 3 of

culture. (E) Micrograph of FAM-RIS (green, asterisks) uptake by long-bone marrow cells, stained for the

membrane marker CD44 (red, arrows), demonstrating mainly intracellular BP. For clarity, the nuclear

staining was not shown. Results (D) are expressed as median fluorescence intensity of the treatment

divided by untreated controls. Means of at least six marrow isolates from mice and standard deviations

are shown. *p<0.05, **p<0.01, ***p<0.001


33

Figure 2. FITC-dextran uptake by bone marrow cells (day 3 of culture with M-CSF and RANKL). Four h

after addition of 0.5 mg/mL of dextran at day 3, uptake by small and large long bone and jaw

osteoclast precursors is not significantly different. Means and standard deviations of six measurements

performed in two independent experiments are shown.

RIS more strongly inhibits Rap1a prenylation in jaw cells than in long bone cells

Using western blotting, we investigated whether the higher RIS uptake by jaw osteoclast

precursors resulted in greater inhibition of protein prenylation. Risedronate (RIS) was added

at day 3 of culture and total cell lysates were collected 24 h later. Unprenylated Rap1a

(uRap1a) was detected in both long bone and jaw cell lysates treated with 10 µM RIS, but

with a lower concentration of RIS, 1 µM, uRap1a accumulation was two-fold higher in jaw

cells (Figure 3). Our results show that the differences in uptake of FAM-RIS are reflected by

differences in the accumulation of uRap1a in jaw and long bone osteoclast precursors.

Figure 3. Western blot analysis of unprenylated Rap1a (uRap1a) in bone marrow cells (day 3 of

culture) after 24 h of RIS treatment. uRap1a accumulates in long bone cells and in jaw cells when they

are treated with 10 µM RIS. With 1 µM RIS, uRap1a is only seen in extracts of jaw cells and hardly in

those of long bone cells. Values represent fold induction of the uRap1a band intensity of long bone cells

treated with 1 µM RIS. Gamma settings were adjusted for clarity of the bands. One representative blot

is shown as an example of 5 independent experiments. Quantification was possible for two independent

experiments.

Chapter 2

34

Osteoclastogenesis of long bone and jaw precursors on plastic is similarly affected by BPs

We investigated whether differences in internalization and inhibition of Rap1a prenylation

resulted in different effects on osteoclastogenesis of long bone and jaw bone marrow

precursors. Various concentrations of pamidronate (PAM, an N-BP) were added after 3 days

of culture. After another 3 days of culture in the presence of PAM, the number of TRACP-

positive, multinucleated cells (>2 nuclei, Figure 4A) was assessed. The formation of

osteoclasts from both populations of precursors was inhibited by the bisphosphonate. This

inhibitory effect was only seen in the presence of 100 µM PAM (Figure 4B). Lower

concentrations (1 and 10 µM) did not affect the number of osteoclasts formed from both

bone marrow populations. PAM similarly affected long bone and jaw osteoclastogenesis.

Similar results were obtained when PAM was added during early osteoclastogenesis (t=0-3),

followed by 3 days without BP (Figure 4B, right panel). Again, the number of long bone and

jaw osteoclasts was inhibited with 100 µM PAM, whereas lower concentrations had no

effect. These results show that PAM did not differentially affect osteoclastogenesis of long-

bone marrow cells versus jaw bone marrow cells on plastic, and that the effect of PAM is

independent of the differentiation stage at which it is added. To mimic in vivo conditions

more closely, osteoclastogenesis was induced on bovine cortical-bone slices. Comparable to

the effect on plastic, 100 µM PAM added at day 3 significantly inhibited the number of both

long bone and jaw osteoclasts to a similar extent (Figure 4C), whereas 1 and 10 µM PAM did

not affect osteoclastogenesis. Similar results were obtained when the BP was added during

early stages of osteoclastogenesis, from the start of culture until day 3 (Figure 4C, right

panel).

Consistent with the decreased osteoclastogenesis on bone, gene expression of

both TRACP and dendritic cell-specific transmembrane protein (DC-STAMP; Figure 4D) was

significantly lower with 100 µM PAM in both long bone and jaw cultures. Lower

concentrations of PAM did not affect the expression of TRACP and DC-STAMP by long bone

or jaw osteoclast cultures. These data show that PAM similarly affects the gene expression

of the analyzed osteoclast marker genes by long bone and jaw osteoclast cultures.

BPs induce caspase 3/7 activity in long bone cells

We investigated whether BPs induce apoptosis in bone marrow cells cultured on plastic.

PAM was added after 3 days of culture in the presence of M-CSF and RANKL. Five hours

after addition, PAM (50 µM) significantly induced caspase 3/7 activity in long bone cultures

(Figure 5A). Jaw osteoclast cultures treated with PAM did not show higher caspase 3/7

activity than controls. Caspase 3/7 activity was significantly higher in long bone cells than in

jaw cells (Figure 5A).


35

Figure 4. Osteoclastogenesis of long bone and jaw precursors incubated for 6 days with M-CSF and

RANKL on plastic (B) and bone (A,C,D). (A) Examples of TRACP-positive, multinucleated cells (arrows)

on cortical bone slices. PAM was added from day 0-3 (B,C, right panels) or day 3-6 (B,C, left panels).

(B,C) No significant differences were observed between the effect of PAM on long bone and jaw

osteoclastogenesis on plastic (B) or bone (C). (D) Gene expression of TRACP (left panel) and DC-STAMP

(right panel) in long bone and jaw osteoclasts was similarly affected. Data represent means and SD

(n>4). *p<0.05, **p<0.01, ***p<0.001 represent significant differences from controls. There were no

significant differences between the number of long bone and jaw osteoclasts.

Chapter 2

36

Hence, despite more BP uptake and concomitant accumulation of uRap1a in the

jaw cells (Figure 1D, Figure 3), BPs do not seem to induce more caspase 3/7 activity in

those cells. Perhaps, this was the result of differential expression of anti-apoptotic

mechanisms by both cell types. Therefore, we investigated whether jaw osteoclast

precursors more potently expressed the anti-apoptotic genes Bcl-xL and Bcl-2. After 3 days

of culture with M-CSF and RANKL, both Bcl-xL (Figure 5B) and Bcl-2 (Figure 5C) gene

expression were significantly higher in jaw cells than in long bone cells. The expression of

the pro-apoptotic gene Bax was similarly expressed by both cell types (Figure 5D). These

results indicate that indeed, jaw osteoclast precursors intrinsically express more of the

measured anti-apoptotic genes than long bone osteoclast precursors.

Figure 5. Caspase 3/7 activity and apoptotic gene expression by long bone and jaw osteoclast

precursors. (A) Caspase 3/7 activity was induced by PAM in long bone cells (**p<0.01 compared to the

control set at100%) and was significantly higher in long bone cells than in jaw cells (n=6, # p<0.05).

(B) Gene expression of Bcl-xL and (C) Bcl-2 at day 3 of culture in the presence of M-CSF and RANKL is

higher in jaw cells than in long bone cells. (D) Bax is similarly expressed in long bone and jaw cells.

Gene expression data represent means and SD (n=5) **p<0.01, ***p<0.001, n.s. not significant.


37

Resorption by long bone and jaw osteoclasts is similarly affected by BPs

We investigated the effect of bisphosphonates on osteoclast activity by assessing the

resorption pit area (Figure 6A) after 10 days of culture. When added from day 3, 100 µM of

PAM significantly inhibited osteoclast activity of long bone and jaw osteoclasts (Figure 6B).

Long bone osteoclast activity was also inhibited by 10 µM PAM. Since 10 µM PAM did not

affect the number of osteoclasts on bone slices (see Figure 4C), the inhibitory effect of BPs

on osteoclast activity was more pronounced than its effect on the generation of these cells

and this effect was stronger in long bone cells than in jaw cells. Jaw and long bone

osteoclast activity was not affected by 1 µM PAM.

We also studied the mRNA expression of Atp6v0d2, encoding a subunit of the

vacuolar H+-ATPase, and cathepsin K, that are both essential for osteoclastic bone

resorption. Expression of Atp6v0d2 and cathepsin K in long bone osteoclasts was

significantly inhibited by 10 and 100 µM PAM (Figure 6C). Jaw osteoclasts, however,

expressed significantly less Atp6v0d2 and cathepsin K only when 100 µm PAM was added

(Figure 6C). Therefore, the effect of PAM treatment on the expression of genes needed for

bone resorption was slightly stronger in long bone osteoclasts than in jaw osteoclasts.

Figure 6. Resorptive activity of osteoclasts after 10 days of culture in the presence of M-CSF and

RANKL. (A) Examples of resorption pits (arrows), stained with Coomassie Brilliant Blue. (B) 10 and 100

µM of PAM inhibited bone resorption by osteoclasts. (C) 100 µM PAM significantly inhibited Atp6v0d2

(left) and cathepsin K (right) gene expression. Results represent means (n=4-6) and SD. *p<0.05,

**p<0.01, ***p<0.001 represent significant differences from controls.

Chapter 2

38

BPs similarly inhibit mature long bone and jaw osteoclasts

To investigate their effect on more mature osteoclasts BPs were added after 5 days of

culture. Cells and resorption pits on the same bone slice were stained and analyzed by

confocal microscopy (Figure 7A). One µM PAM did not affect mature osteoclasts, however

after treatment with 10 µM PAM, the number of both long bone and jaw osteoclasts was 3

times lower than in the controls (Figure 7B). This difference was significant for long bone

cells. Resorption by those long bone cells was also significantly inhibited (12-fold, Figure

7C). Bone resorption by jaw osteoclasts after 10 µM PAM was 5 times lower than in the

control, however this was not significant. These results confirm that BPs have a stronger

inhibiting effect on osteoclast activity than on osteoclast numbers.

The resorbed area per field was divided by the number of multinucleated cells as a

measure for activity per osteoclast. Long bone osteoclast activity was inhibited by 10 µM

pamidronate, however this was not significant (Figure 7D). This implies that BPs have a

stronger effect on mature long bone osteoclast activity than on mature jaw osteoclast

activity. This effect was also seen when BPs were added during an earlier stage of

differentiation and confirms the findings shown in Figure 6 and described above.

To represent the in vivo situation more closely, bone slices were pre-incubated with

BPs (10-100 µM) for 20 h, BPs were washed away and bone marrow cells were seeded on

top of the pre-incubated bone slices. Therefore, BPs can only be taken up during bone

resorption and affect mainly mature osteoclasts. Mature osteoclast death was induced by 50

and 100 µM PAM; however, this was only significant for jaw osteoclasts (Figure 7E). A lower

concentration, 10 µM, did not affect the number of osteoclasts or the resorption by those

cells (Figure 7E,F). Resorption by both types of osteoclasts was significantly inhibited by 50

and 100 µM PAM (Figure 7F), indicating that BPs similarly inhibit mature long bone and jaw

osteoclasts. Taken together, our results indicate that BPs similarly affect mature long bone

and jaw osteoclasts and that the inhibiting effect of BPs is stronger on their activity than on

the osteoclast numbers.

Gene expression of non-osteoclastic genes

Gene expression of all tested osteoclast markers was significantly inhibited after treatment

with 100 µM PAM (Figure 4D, 6C). To investigate whether this inhibition was also seen for

non-osteoclast markers, we analyzed the expression of the macrophage marker F4/80 and

the osteoblast marker tissue non-specific alkaline phosphatase (ALP). Also, we investigated

the expression of the transcription factors V-maf musculoaponeurotic fibrosarcoma

oncogene homolog B (MafB) and interferon regulatory factor 8 (Irf8), which are highly

expressed in macrophages and inhibit osteoclastogenesis. Interestingly, expression of ALP,

F4/80, and MafB was higher after treatment with 100 µM PAM from day 3-8 (Figure 8A,B).


39

This effect was not significant for MafB expression by jaw cells though. These results

indicate that PAM specifically inhibits the analyzed osteoclast marker gene expression and

not ALP, F4/80 and MafB gene expression. This could also imply that the potent inhibition of

osteoclastogenesis by PAM alters the cellular balance in the cultures from predominantly the

osteoclast lineage towards macrophages and osteoblasts.

Figure 7. The effect of BPs on mature osteoclasts on bone. (A) confocal microscopic image of

multinucleated cells (arrow heads) and resorption pits (arrows). 10 µM PAM slightly decreased the

number of multinucleated cells (B) and inhibited bone resorption (C) when added from day 5 until the

end (day 8) of culture. (D) Resorption per long bone osteoclast was slightly inhibited by 10 µM PAM. (E)

PAM-incubated bone slices (50 and 100 µM) decreased the number of osteoclasts, however this was

only significant for jaw osteoclasts. (F) Long bone and jaw osteoclast resorption was significantly

inhibited by 50 and 100 µM PAM. Results represent means (n=3-4 (B-D) or n=6 (E-F)) and SD.

*p<0.05, **p<0.01, ***p<0.001 represent significant differences from controls.

Chapter 2

40

Surprisingly, gene expression of Irf8 was also inhibited by 100 µM PAM in both jaw

and long bone cells. Possibly the macrophages that were present at the measured time

point, i.e. after 8 days of culture in the presence of M-CSF and RANKL, were mature. Since

Irf8 induces early macrophage differentiation [22], expression at day 8 may already be

decreased and similar to osteoclastic Irf8 expression. Therefore, a changed balance towards

a more macrophage-like phenotype would not lead to higher Irf8 expression.

Figure 8. Expression of non-osteoclastic genes in bone marrow cells cultured with M-CSF and RANKL.

(A) ALP, (B) F4/80, and (C) MafB gene expression at day 8 in bone marrow cultures were increased

with 100 µM PAM. (D) Irf8 expression was inhibited by 100 µM PAM. Results represent means (n=3-6)

and SD. *p<0.05, ***p<0.001 represent differences from controls.


41

Discussion

We hypothesized that long bone and jaw bone marrow-derived osteoclasts and/or their

precursors are differentially affected by bisphosphonates. By directly comparing the effect of

nitrogen-containing bisphosphonates on bone marrow cells isolated from these sources, we

showed for the first time that indeed, N-BP treatment affected long bone and jaw osteoclast

precursors in different ways.

Jaw osteoclast precursors internalized more fluorescently labeled N-BP than

precursors from long-bone marrow. This was mainly due to active uptake, since little BP

passively passed the membrane when cells were cultured at 4°C. This suggests that the

higher BP uptake by jaw cells was primarily the result of active endocytosis by these cells.

Yet, we observed more BP uptake at 4°C than previously shown by Thompson et al. [21].

This could be attributed to the totally different cell types used, since Thompson et al. used

the mouse macrophage cell line J774 rather than primary bone marrow cultures.

To investigate whether jaw osteoclast precursors have a higher endocytic capacity,

we analyzed uptake of dextran, a marker for fluid-phase endocytosis [21]. Since jaw

osteoclast precursors did not internalize more dextran, our results suggest that jaw bone

marrow cells do not have higher fluid-phase endocytic activity than long-bone marrow cells,

but accumulate more BPs intracellularly. Therefore, jaw osteoclast precursors might use

another mechanism of endocytosis to internalize BP. Another explanation could be that BP is

actively removed from the cells, and that differences in this process account for the higher

intracellular BP concentration in osteoclast precursors from jaw than in those from long

bone. Taken together, these findings support other studies that have shown differences

between osteoclasts at different bone sites [23-25] and the bone marrow cells derived from

those sites [19,26].

It is not known whether jaw bone marrow cells can also take up more BP in vivo. It

is accepted that in vivo, BPs are taken up from the bone in significant amounts only during

bone resorption, explaining why they primarily affect osteoclasts. However, cells other than

osteoclasts in the bone microenvironment, can internalize BPs that are released from the

bone, in particular macrophages, as a result of their high endocytic activity [8,10]. Since

bone turnover is high in the jaw, one might expect a considerable amount of BP to be

released from this bone site [27]. Also, the distribution of a fluorescently labeled

zoledronate (fluorescein-zoledronate, FAM-ZOL) to the alveolar bone was recently

documented in a rat model [28]. Together with our finding that jaw osteoclast precursors

actively internalized more BP than long bone osteoclast precursors in vitro, we expect that

cellular uptake of BPs by bone marrow cells in vivo is also higher in the jaw than in long

bones. This implies that cells other than bone cells, e.g. macrophages, could also play a role

in the pathogenesis of ONJ and supports a previously proposed hypothesis [29].

Chapter 2

42

Surprisingly, despite the higher degree of N-BP internalization in osteoclast

precursors from the jaw, and accumulation of uRap1a with a lower BP concentration in jaw

cells, the subsequent formation of osteoclasts derived from long-bone and jaw marrow

precursors was similarly affected. This finding strongly suggests that jaw osteoclast

precursors are less sensitive to bisphosphonates after internalization. Support for this

assumption is provided by the differential expression of the osteoclast activity genes,

Atp6v0d2 and cathepsin K. In jaw osteoclasts, their expression was less affected by BP

treatment than their expression in long bone osteoclasts. Furthermore, we showed that jaw

cells express more of the anti-apoptotic genes Bcl-xL and Bcl-2 than long bone cells and that

BPs induced higher caspase 3/7 activity in long bone cells than in jaw cells. These results

support our hypothesis and show that indeed long bone and jaw osteoclasts or precursors

respond differently to bisphosphonates.

Despite the accumulation of unprenylated Rap1a with 1 and 10 µM RIS,

osteoclastogenesis was not affected by those concentrations and we saw an effect on

osteoclast activity only with 10 µM PAM. This discrepancy might be caused by a different

response of osteoclasts to the different bisphosphonates. Alternatively, it is possible that a

partial inhibition of protein prenylation (as with 1 µM) is insufficient to compromise the

function of the small GTPases necessary for osteoclast function. The fact that inhibition is

stronger on activity than on the number of osteoclasts is not surprising, since in addition to

their effect on apoptosis, N-BPs also affect osteoclast activity by disturbing actin ring and

ruffled border formation, which are highly dependent on prenylated small GTPases.

We showed that per long bone osteoclast the activity was inhibited. This was

supported by our findings that the effect of PAM treatment on the expression of genes

needed for bone resorption was stronger in long bone osteoclasts than in jaw osteoclasts.

These data also indicate that, contrary to our initial hypothesis, jaw osteoclasts may be less

sensitive to BPs than long bone osteoclasts and suggest that it is worthy to further

investigate the effect of BPs on bone cells and bone marrow cells in the jaw and in long

bones. This may help to gain more insight into the etiology of BP-related ONJ.

In our cultures, gene expression of the macrophage markers F4/80 and MafB and

the osteoblast marker ALP was higher after BP treatment. Although we cultured the bone

marrow cells with an excess of M-CSF and RANKL and therefore strongly stimulated the cells

towards the osteoclast lineage, stromal cells are still present in the culture. Increased ALP

expression after BP treatment may result from the impaired osteoclastogenesis, changing

the balance in these cultures towards the osteoblastic phenotype. Inhibition of

osteoclastogenesis also may have resulted in the maintenance of the macrophage

phenotype, resulting in higher F4/80 and MafB expression in cultures treated with a high

dose of BPs. Surprisingly, gene expression of Irf8 was also inhibited by 100 µM PAM in both


43

jaw and long bone cells. This might be related to the differentiation stage of the

macrophages and further studies are needed to explore these findings in more detail.

In conclusion, our findings are the first to demonstrate different responses of bone-

site-specific osteoclast precursors to bisphosphonates. More FAM-RIS accumulated in

osteoclast precursors from jaw than from long bones, indicating that there may be

differences in the ability of the cells to endocytose BPs, or to expel them following uptake.

This strongly supports previous findings that osteoclasts and precursors are bone-site

specific.

Chapter 2

44

Acknowledgments

The authors would like to thank Manon Ketting, Emine Özokcu, Sophie Kroon, Lieneke

Bakker, Mei-Ling Tsui, and Jolanda Hogervorst for technical assistance. For support and

advice regarding flow cytometry we thank Kamran Nazmi and Dr. Anke Roelofs, and for

advice regarding Western blotting we thank Dr. Behrouz Zandieh Doulabi.


45

References

1. Frith JC, Monkkonen J, Auriola S, Monkkonen H, Rogers MJ. The molecular mechanism of

action of the antiresorptive and antiinflammatory drug clodronate: evidence for the formation

in vivo of a metabolite that inhibits bone resorption and causes osteoclast and macrophage

apoptosis. Arthritis Rheum. 2001;44(9):2201-10.

2. Frith JC, Monkkonen J, Blackburn GM, Russell RG, Rogers MJ. Clodronate and liposome-

encapsulated clodronate are metabolized to a toxic ATP analog, adenosine 5'-(beta, gamma-

dichloromethylene) triphosphate, by mammalian cells in vitro. J Bone Miner Res.

1997;12(9):1358-67.






Commun. 1999;264(1):108-11.

5. Russell RG, Watts NB, Ebetino FH, Rogers MJ. Mechanisms of action of bisphosphonates:

similarities and differences and their potential influence on clinical efficacy. Osteoporos Int.

2008;19(6):733-59.

6. Monkkonen H, Auriola S, Lehenkari P, Kellinsalmi M, Hassinen IE, Vepsalainen J, et al. A new

endogenous ATP analog (ApppI) inhibits the mitochondrial adenine nucleotide translocase

(ANT) and is responsible for the apoptosis induced by nitrogen-containing bisphosphonates.

Br J Pharmacol. 2006;147(4):437-45.

7. Cornish J, Bava U, Callon KE, Bai J, Naot D, Reid IR. Bone-bound bisphosphonate inhibits

growth of adjacent non-bone cells. Bone. 2011;49(4):710-6.











12. Reid IR, Cornish J. Epidemiology and pathogenesis of osteonecrosis of the jaw. Nat Rev

Rheumatol. 2012;8(2):90-6.

13. Allen MR, Burr DB. Mandible matrix necrosis in beagle dogs after 3 years of daily oral

bisphosphonate treatment. J Oral Maxillofac Surg. 2008;66(5):987-94.

14. Arnett T. Regulation of bone cell function by acid-base balance. Proc Nutr Soc.

2003;62(2):511-20.

Chapter 2

46

15. Helsloot RSJ, van den Berg T, Frank MH, Everts V. Bisphosphonate-related osteonecrosis of

the jaw; a literature review and a new hypothesis. International Journal of Oral Research.

2011;2.





2011;32(1):31-63.

18. Kashemirov BA, Bala JL, Chen X, Ebetino FH, Xia Z, Russell RG, et al. Fluorescently labeled

risedronate and related analogues: "magic linker" synthesis. Bioconjug Chem.

2008;19(12):2308-10.



2011;88(1):63-74.

20. de Vries TJ, Schoenmaker T, Beertsen W, van der Neut R, Everts V. Effect of CD44 deficiency

on in vitro and in vivo osteoclast formation. J Cell Biochem. 2005;94(5):954-66.

21. Thompson K, Rogers MJ, Coxon FP, Crockett JC. Cytosolic entry of bisphosphonate drugs

requires acidification of vesicles after fluid-phase endocytosis. Mol Pharmacol.

2006;69(5):1624-32.

22. Yang J, Hu X, Zimmerman M, Torres CM, Yang D, Smith SB, et al. Cutting edge: IRF8

regulates Bax transcription in vivo in primary myeloid cells. J Immunol. 2011;187(9):4426-30.

23. Everts V, Korper W, Hoeben KA, Jansen ID, Bromme D, Cleutjens KB, et al. Osteoclastic bone

degradation and the role of different cysteine proteinases and matrix metalloproteinases:

differences between calvaria and long bone. J Bone Miner Res. 2006;21(9):1399-408.

24. Jansen ID, Mardones P, Lecanda F, de Vries TJ, Recalde S, Hoeben KA, et al. Ae2(a,b)-

deficient mice exhibit osteopetrosis of long bones but not of calvaria. FASEB J.

2009;23(10):3470-81.

25. Zenger S, Ek-Rylander B, Andersson G. Long bone osteoclasts display an augmented

osteoclast phenotype compared to calvarial osteoclasts. Biochem Biophys Res Commun.

2010;394(3):743-9.




27. Kozloff KM, Volakis LI, Marini JC, Caird MS. Near-infrared fluorescent probe traces

bisphosphonate delivery and retention in vivo. J Bone Miner Res. 2010;25(8):1748-58.

28. Hokugo A, Sun S, Park S, McKenna CE, Nishimura I. Equilibrium-dependent bisphosphonate

interaction with crystalline bone mineral explains anti-resorptive pharmacokinetics and

prevalence of osteonecrosis of the jaw in rats. Bone. 2013;53(1):59-68.

29. Pazianas M. Osteonecrosis of the jaw and the role of macrophages. J Natl Cancer Inst.

2011;103(3):232-40.

Chapter 3.

Zoledronic acid differently affects

long bone and jaw bone turnover

and induces molar root resorption in

female mice

Jenny A.F. Vermeer

Greetje A.P. Renders

Marion A. van Duin

Ineke D.C. Jansen

Lieneke F. Bakker

Sophie A. Kroon

Teun J. de Vries

Vincent Everts

Submitted for publication

Chapter 3

48

Abstract

Bisphosphonates (BPs) are widely used to treat bone diseases such as osteoporosis.

However, BPs can negatively affect the jaw bone by causing osteonecrosis of the jaw.

Previously, we showed that BPs differently affected long bone and jaw osteoclast

precursors. Here, we investigated whether in vivo exposure to BPs has a different effect on

long bone and jaw osteoclasts and the turnover of these two types of bone. Zoledronic acid

(ZA, 0.5 mg/kg weekly) was administered i.p. to 3-month-old female C57BL/6J mice for up

to 6 months and its effect on osteoclasts and bone formation was studied. Long-term

treatment with ZA reduced the number of jaw bone marrow cells, without affecting the

number of long-bone marrow cells. ZA treatment did not affect the osteoclastogenic

potential of long bone and jaw bone marrow cells in vitro nor the number of osteoclasts in

vivo. Yet, ZA treatment increased bone volume and mineral density of both long bone and

jaw. Interestingly, 6 months of treatment inhibited bone formation in the long bones

without affecting the jaw. Finally, we showed that BPs can cause molar root resorption. Our

results show that BPs differently affect long bone and jaw bone marrow cells and bone

turnover in those bones. These findings provide more insight into bone-site-specific effects

of bisphosphonates. Also, we showed that BPs can stimulate osteoclasts and their activity at

the molar roots.

The effect of BPs on long bone and jaw bone remodeling

49

Introduction

Bisphosphonates (BPs) such as zoledronic acid (ZA) are commonly used to treat diseases

that are characterized by excessive bone resorption. They inhibit osteoclastic bone

resorption, thereby improving bone quality and reducing fracture risk. As BPs have a strong

affinity for calcium, they rapidly bind to bone mineral after administration. During bone

resorption they are released from the bone and taken up by osteoclasts. However, BPs have

also been shown to be internalized by surrounding cells such as monocytes and

macrophages [1-3]. Long-term BP treatment also reduced the number of osteoclast

precursors in human peripheral blood [4,5]. These data show that next to osteoclasts, other

cell types can also be affected by BPs, indicating that BPs may affect bone turnover in

several ways.

A side effect of high-dose BP use in humans is osteonecrosis of the jaw (ONJ),

which is defined as an exposure of bone that lasts for at least 8 weeks [6]. Risk factors for

this condition are dental extractions and other oral trauma [7], however it is not clear why

specifically the jaw is negatively affected by BPs. With the increasing life expectancy, it is

estimated that these drugs will be subscribed more often, and the side effects will become

an emerging problem. Although several in vivo models have shown that BP treatment

induces necrotic jaw bone, these signs of ONJ were not solely the result of BP treatment.

After tooth extraction, BPs delayed alveolar bone healing, especially in combination with an

immunosuppressive agent [8,9]. Also, BP treatment induced signs of ONJ in rat models for

periodontal disease [10,11]. Despite the development of these models, it remains unclear

how and why specifically the jaw is affected by BP treatment. To explain bone-site-specific

effects of BPs, it is essential to gain more insight into the bone remodeling activity of the

different bones.

Previously, we hypothesized that different effects of BPs on osteoclasts in the jaw

and in long bone may explain the etiology of BP-related ONJ [12]. We showed that BPs did

not differently affect long bone and jaw osteoclastogenesis in vitro. However, osteoclast

precursors from the jaw internalized more BPs than long-bone marrow precursors. It

remains unclear whether long bone and jaw bone marrow cells are differently affected by

BPs in vivo.

Here, we studied the effect of BP treatment on the jaw and on long bones of

female mice in vivo. We also investigated the osteoclastogenic potential of bone marrow

cells isolated from BP-treated mice. Finally, we studied the effect of BP treatment on

markers of bone turnover in the different bones. This study will give us more insight into

bone-site-specific effects of bisphosphonates and may provide a better understanding of the

pathophysiology of osteonecrosis of the jaw.

Chapter 3

50


Mice and bone marrow isolation

Animal experiments were approved by the Animal Welfare Committee of the VU University

(Amsterdam, The Netherlands). Female, 3-month-old C57BL/6J mice were divided into 4

groups. Two groups were ovariectomized and two groups were sham operated. The

ovariectomized mice were used for other experiments, not mentioned in this study. Sham

operated mice were injected intraperitoneally with a high dose (0.5 mg/kg) zoledronic acid

(ZA; Novartis, Basel, Switzerland) or saline once a week. Mice (6 per group) were sacrificed

at baseline and 1, 3, and 6 months after the start of the experiment. Eight days before

sacrifice, mice were injected intraperitoneally with fluorescently labeled calcein (10 mg/kg;

Sigma, St. Louis, MO, USA), and 2 days before sacrifice with alizarin complexone (20 mg/kg;

Sigma). Bone marrow cells were isolated from the right mandibles and long bones (tibia and

femur) as described before [12]. Left mandibles, maxillas, and left humeri were fixed with

4% phosphate-buffered formaldehyde and stored at 4°C until microCT analysis and

processing for histology.

Osteoclastogenesis and TRACP staining

Cells were stained with Türk’s solution (Merck, Darmstadt, Germany) and counted with a

hemocytometer. Freshly isolated long bone and jaw bone marrow cells were seeded in 96-

well plates (105 cells per well) and osteoclastogenesis was induced with 30 ng/mL

recombinant mouse macrophage-colony stimulating factor (rmM-CSF; R&D Systems,

Minneapolis, MN, USA) and 20 ng/mL recombinant mouse receptor activator of nuclear

factor κB ligand (rmRANKL; R&D Systems). After 6 days, cells were fixed and a TRACP

staining was performed with the leukocyte acid phosphatase kit (Sigma). Nuclei were

visualized with 4’6-diamidino-2-phenylindole dihydrochloride (DAPI). The number of

tartrate-resistant acid phosphatase (TRACP)-positive, multinucleated (≥3 nuclei) cells was

assessed using an inverted microscope (Leica Microsystems GmbH, Wetzlar, Germany).

Micro-computed tomography

Micro-computed tomography (microCT; Scanco Medical AG, Brüttisellen, Switzerland) was

used to assess bone volume and the tissue mineral density (TMD) of humeri and hemi

mandibles. Scanning was performed in fixation fluid with an 8 µm voxel size, a peak voltage

of 55 kV, and an integration time of 250 ms. Imaging processing included Gaussian filtering

and segmentation: sigma = 0.8, support = 1, threshold unit 170 (= 618 mg HA cm-3) and

200 (= 727 mg HA cm-3) for humeri and mandibles, respectively, for all analyzed time points

and groups. Total humeri were analyzed by exactly outlining the bones, i.e. without


51

including empty volume. The volume of interest for the jaw is indicated in Figure 2A. For

both long bone and jaw, 20-30 slices per sample were analyzed.

Histology and histomorphometry

Following fixation, tissue samples were dehydrated with ethanol and subsequently

embedded in methyl methacrylate (MMA). Transverse sections of the left maxillary bone and

longitudinal sections of the proximal humeri were made using a Jung K microtome (Leica).

Five µm sections were attached to gelatin-coated microscope slides and dried for at least 3

days at 37°C. Sections were mounted in depex and stored at room temperature in the dark

for the analysis of dynamic bone formation in the diaphysis of the proximal humeri and the

alveolar bone surrounding two or three roots of the second molar. Analyses were performed

using Leica QWin (Leica Microsystems Image Solutions, Rijswijk, The Netherlands) and

Image Pro-Plus Software (Media Cybernetics, Silver Spring, MD, USA).

Right maxillae and distal humeri were decalcified in 4.2% EDTA containing 0.8%

formaldehyde for 5 weeks at room temperature, dehydrated, and embedded in paraffin.

Transverse sections of the maxilla and longitudinal sections of the distal humeri were cut (7

µm thickness) and attached to silane coated microscope slides. Paraffin was removed with

xylene substitute, sections were rehydrated and a TRACP staining was performed as

described previously with minor changes [13]. A sodium-potassium tartrate concentration of

2 mM was used and sections were incubated for 3 h at 37°C. Counterstaining was

performed with Mayer’s hematoxylin. TRACP-positive cells covering the bone in the

endocortical, distal diaphysis and in the alveolar bone surrounding the roots of the second

molar were analyzed and this was expressed as the percentage of bone covered by

osteoclasts. Also, the osteoclasts attached to the cementum of the roots of the second

molar were determined. Here, a distinction was made between osteoclasts that were

attached to root-resorbed areas and those that attached to cementum that showed no signs

of resorption. The periodontal ligament (PDL) area was also determined.


The Kruskal-Wallis test, followed by Dunn’s multiple comparison test were performed to test

the effect of age on the number of bone marrow cells, osteoclast formation in vitro and in

vivo, osteoclast size, and bone formation. To test the effect of age on microCT parameters

we used a One-way ANOVA followed by Tukey’s multiple comparison tests. Unpaired, two-

tailed t-tests were used to test the effect of ZA treatment on TMD and bone volume

fraction. To test the effect of ZA treatment on the number of bone marrow cells, on the

osteoclastogenic potential of bone marrow cells, bone formation parameters, and on the

number of osteoclasts in vivo we used the Mann-Whitney U test. Graphs represent means

Chapter 3

52

and standard deviations of six measurements, unless indicated otherwise. Differences were

considered significant when p<0.05.


53

Results

Bisphosphonates affect jaw bone marrow cells without affecting long-bone marrow cells

Bone marrow cells were isolated from mice treated with ZA to assess the osteoclastogenic

potential of those cells. With age, fewer bone marrow cells were isolated from the long bone

(Figure 1A). There was no effect of ZA treatment on the number of bone marrow cells

isolated from long bones (Figure 1A). Interestingly, ZA gradually reduced the number of jaw

bone marrow cells and after 6 months treatment, ZA significantly reduced the number of

bone marrow cells that was isolated from the jaw (Figure 1B). These results indicate that

long-term ZA treatment reduces the number of bone marrow cells present in the jaw. These

results also show that BPs differently affect bone marrow cells in the jaw and in the long

bones.

Bisphosphonates do not affect the osteoclastogenic potential of bone marrow cells

We investigated the osteoclastogenic potential of bone marrow cells from the long bones

and jaws of mice treated with ZA and their controls at several time points. After 6 months,

at the age of 9 months, significantly fewer long bone osteoclasts were formed from control

bone marrow than after 1 and 3 months (Figure 1C). The age of the mice did not affect the

number of osteoclasts that were formed from control bone marrow from the jaw (Figure

1D). Since 6 months of ZA treatment reduced the number of jaw bone marrow cells (see

above), we were not able to isolate enough cells from the jaw at that time point to assess

their osteoclastogenic potential. Therefore, the effect of ZA treatment on jaw

osteoclastogenesis was only studied after 1 and 3 months of treatment (Figure 1D). There

was no significant effect of ZA treatment on the number of osteoclasts that was formed

from both long bone and jaw bone marrow precursors at any time point (Figure 1C,D). This

indicates that BP treatment in vivo does not affect the in vitro osteoclastogenic potential of

bone marrow cells at the included time points.

Chapter 3

54

Figure 1. Bone marrow cell isolation and osteoclastogenic potential of bone marrow cells from long

bones and jaws from mice treated with ZA. (A) The number of long-bone marrow cells from tibia and

femur was not affected by ZA treatment. (B) Six months of ZA treatment inhibited the number of jaw

bone marrow cells in the hemi-mandible. (C,D) The osteoclastogenic potential of (C) long bone and (D)

jaw bone marrow was not affected by ZA treatment. The x-axes show the time after the start of

treatment at the age of 3 months. Means of 6 (A,B) or 4-6 (C,D) measurements and standard

deviations are shown. **p=0.01 (Mann-Whitney U test) represents a significant difference between ZA

treatment and control, #p<0.05, ##p<0.01 represent a significant difference of the controls at that

time point versus baseline (0), $p<0.05 represents a significant difference of the control group at 6

months versus 3 months.

Bisphosphonates increase bone volume and bone mineral density

ZA may reduce the number of bone marrow cells either directly by inducing bone marrow

cell apoptosis, or indirectly by decreasing bone marrow space due to a higher bone volume

or by a combination of both. To investigate whether a decrease in the number of jaw bone

marrow cells (Figure 1B) was the result of such an indirect effect, we assessed bone volume

by microCT (Figure 2A). With age, bone volume increased up to 6 months in the long bones

and up to 9 months in the jaw (Figure 2B). ZA increased bone volume fraction (BV/TV) of


55

the long bones and of the jaw after 1, 3, and 6 months of treatment (Figure 2B). This effect

was similar in long bones (Figure 2B, left panel) and in the jaw (Figure 2B, right panel).

However, the jaw bone volume fraction was higher than long bone BV/TV and increased to

above 90% after ZA treatment. This might explain the decreased number of jaw bone

marrow cells as indicated in Figure 1B. An increased bone volume by ZA decreased the

periodontal ligament space of the lingual root after 6 months of treatment, whereas the PDL

space surrounding the other roots were not affected (Supplementary Figure 1).

The TMD of both types of bone increased with the age of the mice and was higher

in the jaw than in long bones (Figure 2C). After 3 and 6 months of ZA treatment, both long

bone (Figure 2C, left panels) and jaw (right panels) TMD were significantly higher than the

TMD of the control animals. This indicates that BPs similarly increase the mineral content in

long bones and jaws.

Figure 2. MicroCT analysis of long bones and jaws from mice treated with BPs. (A) 3D reconstructions

and the volume of interest (VOI) of the humeri (top) and the 1st and 2nd molar region of the mandible

(bottom, arrows) after 6 months of treatment. M1=first molar, bar=1 mm. (B) Bone volume (% of total

volume) of long bones (left) and jaws (right) was higher after 1,3, and 6 months of ZA treatment. (C)

Tissue mineral density (TMD) was increased by 3 and 6 months of ZA treatment. The x-axes show the

time after the start of treatment at the age of 3 months. Graphs show the means and standard

deviations of 5-6 measurements. *p<0.05, **p<0.01, ***p<0.001 represent significant differences

between ZA treatment and control, determined by an unpaired, two-tailed t-test.

Chapter 3

56

Osteoclast numbers in the long bones increase with age

The bone surface that was covered by osteoclasts and the number of osteoclasts in vivo

was assessed (see Supplementary Figure 2 for an example). In the long bones the

osteoclast surface and osteoclast numbers increased with the age of mice (#p<0.05, 6

months versus 0, and 6 months versus 1 month). In the jaw, the osteoclast numbers did not

change with age. Interestingly, after 3 and 6 months of study, when the mice were 6 and 9

months old, respectively, osteoclasts covered a 2 to 3.6 times larger long bone surface than

jaw bone surface (p<0.05, Figure 3A versus B). ZA treatment did not significantly affect the

number of osteoclasts in the long bones (Figure 3A), nor in the jaw (Figure 3B). The effect

on the number of osteoclasts per bone perimeter was similar (Figure 3C,D). Yet, the long

bone and jaw osteoclasts were slightly larger after BP treatment (Figure 3E,F). This was

only significant in the jaw after 3 months of treatment. Since the bone volume fraction and

TMD were higher after ZA treatment, these data may indicate that the osteoclasts are less

active after treatment.

ZA differently affects mineralization in long bone and jaw

The higher BV/TV and TMD after ZA treatment could not be attributed to a lower number of

osteoclasts. Therefore, we investigated the effect on the bone formation markers

mineralizing surface and mineral apposition rate. The mineralizing surface in the control

mice decreased with age, only in long bones (Figure 4A). Interestingly, the long bone

mineralizing surface was 1.4 to 2.4 times higher than the mineralizing surface in the jaw

(Figure 4A versus 4B). The mineral apposition rate was even 2.2 to 3 times higher in the

long bones than in the jaw (Figure 4C versus 4D).

ZA treatment significantly reduced the bone formation parameters in the long

bones after 3 and 6 months of treatment, whereas the jaw was only affected after 3 months

(Figure 4). This might indicate that 6 months of ZA treatment reduces bone turnover in the

long bones, whereas it is unaffected in the jaw. These results also indicate that the higher

bone volume fraction and TMD were probably induced at an earlier stage of treatment, or

that they were the result of reduced osteoclast activity.


57

Figure 3. Osteoclast counts on decalcified sections. (A) Long bone and (B) jaw osteoclasts were not

significantly affected by ZA treatment. (C-D) The number of osteoclasts per bone perimeter in (C) long

bone and (D) jaw, showing similar results as the osteoclast surface covering the bone. (E-F) Osteoclasts

are generally larger after ZA treatment, however this was only significant in the jaw (F) after 3 months

of treatment. The x-axes show the time after the start of treatment at the age of 3 months. Graphs

show the means and standard deviations of 4-6 measurements or 3 measurements for the jaw at

baseline. # indicates p<0.05, between 6 months and baseline, and between 6 months and 1 month, as

determined by The Kruskal-Wallis test, followed by Dunn’s multiple comparison. **p<0.01 using the

Mann-Whitney U test comparing ZA and control at that time point.

Chapter 3

58

Figure 4. Mineralization parameters. (A) Long bone mineralizing surface and (C) mineral apposition

rate were inhibited by 3 and 6 months of ZA. (B) Jaw mineralizing surface and (D) mineral apposition

rate were only affected after 3 months of ZA. Graphs show the means and standard deviations of n=4-6

for the long bone and n=3-6 for the jaw. The x-axes show the time after the start of treatment at the

age of 3 months. *p<0.05, **p<0.01 using the Mann-Whitney U test between ZA and control at that

time point. #p<0.05 as determined by The Kruskal-Wallis test, followed by Dunn’s multiple comparison.

ZA induces molar root resorption

Analysis of the root surfaces of the molars revealed the presence of osteoclasts attached to

the cementum layer (Figure 5A). Osteoclasts were commonly found in the animals treated

with ZA and very rare in the controls (Figure 5C). These molar-root-associated osteoclasts

proved to be associated with resorption pits (Figure 5B). After 3 months, 9.1% of the

osteoclasts were active, and after 6 months 15.4% of the osteoclasts was associated with a

resorption pit (Figure 5B,D), thus indicating resorptive activity despite the presence of the

bisphosphonate. These results show, for the first time to our knowledge, that BPs can

induce osteoclast activity at a certain location.


59

Figure 5. Osteoclast attachment to root cementum. (A) Active (arrow) and non-active (dashed arrow)

osteoclast attached to the root cementum. d: dentin, ab: alveolar bone, pdl: periodontal ligament, *

cementum. (B) Higher magnification of the resorbing osteoclast, clearly showing disrupted cementum.

Bars: 30 µm. (C) Resorption is exclusive to ZA treatment; in controls, cementum associated osteoclasts

are very rare. (D) The percentage of osteoclasts after 3 and 6 months that was associated with a

resorption pit. The x-axes show the time after the start of treatment at the age of 3 months. The

graphs show the means and standard deviations of 5-6 measurements or 3 measurements for the jaw

at baseline. ##p<0.01 represents a significant differences between ZA treatment and control using the

Mann-Whitney U test.

Chapter 3

60

Discussion

In this study we showed that osteoclast numbers and bone turnover rate are bone-site

specific. We showed that the osteoclast numbers increased with age in the long bones,

whereas their number remained the same in the jaw (Figure 3). Also, the diaphyses of the

long bones contained more osteoclasts per bone perimeter than the alveolar bone after 3

and 6 months of study, when the mice were 6 and 9 months old, respectively. This was

accompanied by higher mineral apposition rates in the long bones than in the alveolar bone,

as measured by dynamic bone parameters (Figure 4). These results strongly suggest that

the bone turnover rate in old mice is higher in the long bones than in the jaw. This is in

contrast with the paradigm, stating that bone turnover in the jaw is higher than in long

bones [14]. However, only few studies directly compared turnover at both sites. Also,

variation in turnover rate exists between locations within both bone sites [15]. Huja and co-

workers showed that the bone formation rate in young, growing dogs is higher in the femur

than in the jaw [16], whereas the opposite was found in mature dogs [17]. In line with our

studies, Kubek et al. found similar mineral apposition rates in 6-months-old C3H female

mice, which were also higher in the femur than at the periodontal ligament surface [15]. It

has to be noted that they studied intracortical turnover in the femur, whereas we

investigated endocortical turnover in the humerus. Together these results show that bone

turnover is not necessarily higher in the jaw than in the long bones; on the contrary, in mice

it seems to be higher in the long bones.

Besides the difference in long bone and jaw osteoclast numbers found in the

current study, we previously showed that osteoclast precursors from these sites can be

differently affected by bisphosphonates. Jaw osteoclast precursors internalized more BPs

and accumulated more unprenylated Rap1a than long bone precursors, however, this was

not accompanied by differences in osteoclastogenesis [12]. In the present study, ZA did not

significantly affect the number of osteoclasts in the long bones and in the jaw, thereby

confirming that ZA has a similar effect on long bone and jaw osteoclasts. Yet, more and

larger osteoclasts seemed present in the jaw after 6 months of treatment, whereas the

number was inhibited in the long bones (Figure 3). This is in line with a previous study by

Kuroshima et al., who showed fewer long bone osteoclasts and slightly more jaw osteoclasts

after subcutaneous injections of 0.1 mg/kg ZA twice a week for 13 months [18]. Therefore,

it requires further investigation whether jaw osteoclasts are less susceptible to BPs than

long bone osteoclasts after a longer treatment.

Although we previously showed that jaw bone marrow cells might be less sensitive

to BPs than long-bone marrow cells in vitro [12], the current study showed that the bone

marrow cells in the jaw were more susceptible to BP treatment than those in the long

bones. Six months of ZA treatment reduced the number of bone marrow cells in the jaw and


61

not in long bones (Figure 1B). This might be attributed to the very limited bone marrow

cavity in the jaw (Figure 2) or by higher BP uptake by the bone marrow cells at that site

[12]. Perhaps, jaw bone marrow cells are exposed to higher BP concentrations in vivo. Jaw

and long bones from rats were shown to contain a similar amount of BPs per unit of dry

weight or calcium [19,20]. Since the jaw contains more mineral per bone volume [20]

(Figure 2), it likely adsorbed more BPs than long bones. Together with the finding that jaw

osteoclasts resorb at least as much bone or dentin as long bone osteoclasts in vitro [12,21],

more BP is likely released from the jaw than from long bone, making it available for the

surrounding cells in the bone marrow [1,2]. The resulting high local BP concentration might

cause apoptosis of bone marrow cells, leading to a reduced number of jaw bone marrow

cells. Also, Hokugo et al. recently showed that newly administered BPs can replace

previously adsorbed BP [22]. Since we applied ZA weekly, it is likely that the BP replaces the

previously administered ZA, releasing it into the bone microenvironment and making it

available for internalization by other cell types than osteoclasts.

The reduced number of jaw bone marrow cells after 6 months of ZA treatment did

not result in fewer osteoclasts at the alveolar bone (Figure 1,3). This may indicate that the

osteoclasts are not formed from precursors in the bone marrow or that the osteoclast

precursors still present in the bone marrow were not affected by BPs. This was supported by

the finding that ZA treatment did not affect the osteoclastogenic potential of bone marrow

cells, although it should be noted that the osteoclastogenic potential after 6 months could

not be measured (Figure 1D). Yet, since ZA reduced the number of jaw bone marrow cells,

our findings seem to provide support for the hypothesis that BPs may affect other cell types

in the bone marrow, such as macrophages [23]. Also, the jaw is more exposed to microbes

and pathogens and a negative effect of BPs on macrophages and the inflammatory response

may explain why specifically the jaw is vulnerable for osteonecrosis. Inhibition of

inflammation was also shown to be involved in the induction of ONJ [8,9]. Therefore, it

would be of interest to study the BP release from the bone and uptake by bone marrow cells

in the jaw in a mouse model of inflammation and bisphosphonates, such as the previously

designed models for ONJ [10,11]. This could also provide more insight into whether the

different response of long bone and jaw osteoclasts to BPs we showed here are related to

the development of ONJ. This requires further investigation, since with our study design, the

mice did not develop any signs of osteonecrosis of the jaw.

Another hypothesis regarding the etiology of ONJ is the oversuppression of bone

turnover. A reduced bone remodeling capacity and reduced ability to repair bone damage

can lead to the accumulation of microdamage [24,25]. Our results however, do not support

a reducing effect on bone turnover markers in the jaw, whereas bone turnover was inhibited

by ZA in the long bones on the long term (Figure 4). Also, the bone surface covered with

Chapter 3

62

osteoclasts and osteoclast numbers increased with age in mice only in long bone, to levels

that were 2 to almost 4 times higher in the long bones than in the jaw (Figure 3A-D).

Therefore, our results suggest that oversuppression of bone turnover and the alleged

microdamage would more likely cause side effects in the long bones than in the jaw.

Indeed, others have shown that BPs were associated with atypical femur fractures, which

are probably caused by the inability to heal microcracks [26].

Interestingly, under the influence of BPs, active osteoclasts accumulated at the

molar root cementum (Figure 5). Previous studies on the effect of bisphosphonates during

orthodontic treatment reported an inhibiting effect of BPs on root resorption during

orthodontic tooth movement [27,28]. However, under our regime with ZA, but without an

external force, approximately 10-15% of the root osteoclasts proved to be active. Therefore,

we demonstrate that under the influence of BPs, more osteoclasts are associated with tooth

cementum and moreover, that these cells actively resorb this mineralized tissue. A

stimulating effect on osteoclast activity has, as far as we know, never been reported. This

was unexpected and surprising, since BPs are used as a treatment to inhibit osteoclast

activity. Possibly, those molar-root-associated osteoclasts were resistant to BPs as little BP

accumulates at the root cementum [29].

Another explanation for the induction of osteoclast activity may be similar to the

mechanism of osteoclast activity that is induced during orthodontic tooth movement.

Mechanical loading induces bone resorption in the direction of the load. We hypothesize that

through a similar mechanism, osteoclasts accumulate at the root after BP treatment. The

increased bone volume after BP treatment (Figure 2) may induce mechanical loading that is

transmitted through the periodontal ligament, and induces root resorption in order to make

space for the newly formed bone. The periodontal ligament fibroblasts themselves could

play a critical role in this process as they can stimulate osteoclast formation [30,31].

However, further studies are required to investigate the role of the periodontal ligament

fibroblasts in osteoclast formation after BP treatment. Also, it would be highly interesting to

investigate whether root resorption and a possible loss of teeth could be related to the

positive correlation between a tooth extraction and the occurrence of BP-related ONJ.

In conclusion, our results support previous findings that osteoclasts are bone-site

specific. Also, we show that BPs can differently affect long bone and jaw osteoclasts and

bone turnover in vivo. Furthermore, we found resorption of tooth root cementum, in spite of

the presence of the bisphosphonate ZA, indicating that the drug is able to stimulate

osteoclast-mediated resorption. This stimulated resorption was site-specific, again indicating

the diverse effects the drug has on osteoclasts at different bone sites. Finally, the evidence

that BPs affect specifically jaw bone marrow cells may improve our understanding of the

pathogenesis for BP-related ONJ.


63

Acknowledgments

The authors would like to thank Dr. Geerling E.J. Langenbach for his input in the design of

the study and Dr. Clara M. Korstjens for the development of analysis protocols for the

measurement of dynamic bone parameters. We are grateful to Carla Prins and Rika van der

Laan for conducting the animal experiments. We thank Dr. Behrouz Zandieh Doulabi for

development of the TRACP staining protocol on decalcified sections. Finally, we thank The

Dutch Organization for Scientific Research for funding (NWO grant number: 021.001.050 to

GAPR).

Chapter 3

64

Supplementary data

Supplementary Figure 1. Periodontal ligament area. ZA reduced the periodontal ligament area

surrounding the root on the lingual side, only after 6 months of treatment as tested with a Mann-

Whitney U test (**p<0.01). The x-axes indicate the locations of the molar roots.

Supplementary Figure 2. TRACP staining on decalcified sections. (A) transverse section of the

maxilla, showing roots of the second molar. ab: alveolar bone, pdl: periodontal ligament, d: dentin. (B)

Longitudinal section of the distal humerus. Squares (A-B) indicate the area that is magnified in C

(maxilla) and D (humerus). Arrows indicate osteoclasts. Bars: 100 µm (A-B), 30 µm (C-D)


65

References








4. D'Amelio P, Grimaldi A, Cristofaro MA, Ravazzoli M, Molinatti PA, Pescarmona GP, et al.

Alendronate reduces osteoclast precursors in osteoporosis. Osteoporos Int.

2010;21(10):1741-50.

5. D'Amelio P, Grimaldi A, Di Bella S, Tamone C, Brianza SZ, Ravazzoli MG, et al. Risedronate

reduces osteoclast precursors and cytokine production in postmenopausal osteoporotic

women. J Bone Miner Res. 2008;23(3):373-9.






24.

8. Bi Y, Gao Y, Ehirchiou D, Cao C, Kikuiri T, Le A, et al. Bisphosphonates cause osteonecrosis of

the jaw-like disease in mice. Am J Pathol. 2010;177(1):280-90.

9. Kikuiri T, Kim I, Yamaza T, Akiyama K, Zhang Q, Li Y, et al. Cell-based immunotherapy with

mesenchymal stem cells cures bisphosphonate-related osteonecrosis of the jaw-like disease in

mice. J Bone Miner Res. 2010;25(7):1668-79.

10. Aghaloo TL, Kang B, Sung EC, Shoff M, Ronconi M, Gotcher JE, et al. Periodontal disease and

bisphosphonates induce osteonecrosis of the jaws in the rat. J Bone Miner Res.

2011;26(8):1871-82.

11. Aguirre JI, Akhter MP, Kimmel DB, Pingel JE, Williams A, Jorgensen M, et al. Oncologic doses

of zoledronic acid induce osteonecrosis of the jaw-like lesions in rice rats (Oryzomys palustris)

with periodontitis. J Bone Miner Res. 2012;27(10):2130-43.

12. Vermeer JA, Jansen ID, Marthi M, Coxon FP, McKenna CE, Sun S, et al. Jaw bone marrow-

derived osteoclast precursors internalize more bisphosphonate than long-bone marrow

precursors. Bone. 2013;57(1):242-51.

13. Ersan N, van Ruijven LJ, Bronckers AL, Olgac V, Ilguy D, Everts V. Teriparatide and the

treatment of bisphosphonate-related osteonecrosis of the jaw: a rat model. Dento maxillo

facial radiology. 2013;43(1):20130144.

14. Allen MR, Burr DB. The pathogenesis of bisphosphonate-related osteonecrosis of the jaw: so

many hypotheses, so few data. J Oral Maxillofac Surg. 2009;67(5 Suppl):61-70.

Chapter 3

66

15. Kubek DJ, Burr DB, Allen MR. Ovariectomy stimulates and bisphosphonates inhibit

intracortical remodeling in the mouse mandible. Orthodontics & craniofacial research.

2010;13(4):214-22.

16. Huja SS, Beck FM. Bone remodeling in maxilla, mandible, and femur of young dogs.

Anatomical record (Hoboken, NJ : 2007). 2008;291(1):1-5.

17. Huja SS, Fernandez SA, Hill KJ, Li Y. Remodeling dynamics in the alveolar process in skeletally

mature dogs. The anatomical record Part A, Discoveries in molecular, cellular, and

evolutionary biology. 2006;288(12):1243-9.

18. Kuroshima S, Go VA, Yamashita J. Increased numbers of nonattached osteoclasts after long-

term zoledronic acid therapy in mice. Endocrinology. 2012;153(1):17-28.

19. Bauss F, Pfister T, Papapoulos S. Ibandronate uptake in the jaw is similar to long bones and

vertebrae in the rat. Journal of bone and mineral metabolism. 2008;26(4):406-8.

20. Wen D, Qing L, Harrison G, Golub E, Akintoye SO. Anatomic site variability in rat skeletal

uptake and desorption of fluorescently labeled bisphosphonate. Oral diseases.

2011;17(4):427-32.



2011;88(1):63-74.

22. Hokugo A, Sun S, Park S, McKenna CE, Nishimura I. Equilibrium-dependent bisphosphonate

interaction with crystalline bone mineral explains anti-resorptive pharmacokinetics and

prevalence of osteonecrosis of the jaw in rats. Bone. 2013;53(1):59-68.


2011;103(3):232-40.

24. Hoefert S, Schmitz I, Tannapfel A, Eufinger H. Importance of microcracks in etiology of

bisphosphonate-related osteonecrosis of the jaw: a possible pathogenetic model of

symptomatic and non-symptomatic osteonecrosis of the jaw based on scanning electron

microscopy findings. Clinical oral investigations. 2010;14(3):271-84.

25. Mashiba T, Hirano T, Turner CH, Forwood MR, Johnston CC, Burr DB. Suppressed bone

turnover by bisphosphonates increases microdamage accumulation and reduces some

biomechanical properties in dog rib. J Bone Miner Res. 2000;15(4):613-20.

26. Shane E, Burr D, Abrahamsen B, Adler RA, Brown TD, Cheung AM, et al. Atypical

subtrochanteric and diaphyseal femoral fractures: second report of a task force of the

american society for bone and mineral research. J Bone Miner Res. 2014;29(1):1-23.

27. Fujimura Y, Kitaura H, Yoshimatsu M, Eguchi T, Kohara H, Morita Y, et al. Influence of

bisphosphonates on orthodontic tooth movement in mice. European journal of orthodontics.

2009;31(6):572-7.

28. Sirisoontorn I, Hotokezaka H, Hashimoto M, Gonzales C, Luppanapornlarp S, Darendeliler MA,

et al. Orthodontic tooth movement and root resorption in ovariectomized rats treated by

systemic administration of zoledronic acid. American journal of orthodontics and dentofacial

orthopedics : official publication of the American Association of Orthodontists, its constituent

societies, and the American Board of Orthodontics. 2012;141(5):563-73.


67








research. 2001;80(3):887-91.

Chapter 4.

The effect of bisphosphonates on

human periodontal-ligament-

fibroblast-mediated

osteoclastogenesis

Jenny A.F. Vermeer

Jeroen Langeveld

Ton Schoenmaker

Tymour Forouzanfar

Vincent Everts

Teun J. de Vries

Manuscript in preparation

Chapter 4

70

Abstract

Bisphosphonates (BPs) such as pamidronate are used to treat bone destructive diseases and

act by inhibiting bone resorption by osteoclasts, thereby improving bone quality. In contrast,

BPs were shown to stimulate osteoclast formation and resorption at the mouse molar root.

Periodontal ligament (PDL) fibroblasts could be seen as the mediators of osteoclast

formation at this location. Here, we investigated whether BP exposure of PDL fibroblasts

isolated from molar roots alters osteoclast formation. Human PDL fibroblasts were isolated

from healthy donors, and subjected to pamidronate (PAM; 1-100 µM) in vitro for 24 hours,

after which cell viability was measured. PAM was removed, and freshly isolated peripheral

blood mononuclear cells (PBMCs) were added to the PAM-treated or vehicle-treated

fibroblasts. Osteoclasts were counted after 21 days of co-culture, and osteoclast-related

gene expression was measured. Cell viability was not affected after 24 hours of treatment

with any concentration of PAM. However, in the long term, transient exposure to 100 µM

PAM was toxic to PDL fibroblasts. Osteoclast formation was induced by the vehicle-treated

PDL fibroblasts, and this was not affected by pre-treatment with 1 or 10 µM PAM. Also, gene

expression of the osteoclast marker tartrate-resistant acid phosphatase was unaffected by

pre-treatment with PAM. Yet, 100 µM completely blocked osteoclast formation, probably due

to the absence of PDL fibroblasts. Moreover, fewer PBMCs survived in this condition than in

a mono-culture of PBMCs. In conclusion, BPs inhibited periodontal-ligament-fibroblast-

mediated osteoclast formation, probably due to a toxic effect on PDL fibroblasts, however,

there was no effect on osteoclastogenesis with a non-toxic BP concentration.

The effect of BPs on PDL-fibroblast-mediated osteoclastogenesis

71

Introduction

Bone remodeling, a life-long process of bone renewal, is performed by bone-resorbing

osteoclasts that decalcify and degrade bone matrix, and bone forming osteoblasts. The

coupling between bone resorption and bone formation was shown to play an important role

in maintaining bone homeostasis (reviewed in Sims and Martin 2014 [1]). Osteoblasts can

signal and stimulate osteoclastogenesis by expressing M-CSF and RANKL [2,3]. Periodontal

ligament fibroblasts were also shown to stimulate osteoclastogenesis in vitro [4,5].

When the balance between bone resorption and bone formation is disturbed and

directed more towards bone resorption, such as in osteoporosis and bone metastasis,

bisphosphonates (BPs) can be used as a treatment. BPs induce osteoclast apoptosis and

inhibit bone resorption. The signaling towards osteoblasts however, is likely lost due to BP

activity, and therefore long-term BP treatment results in reduced bone formation along with

inhibition of resorption [6]. BPs were also shown to have a direct inhibiting effect on

osteoblast viability and differentiation [7,8] and the viability of periodontal ligament cells

[9,10]. Anti-apoptotic effects on osteocytes and osteoblast with low concentrations of BPs

were also reported [11,12].

Recently, we have shown that BP administration can cause osteoclast formation at

the molar root in mice (manuscript submitted). Surprisingly, and in contrast to the generally

accepted mode of action of bisphosphonates, actively resorbing osteoclasts were observed

at the molar roots, exclusively in the bisphosphonate-treated animals. We hypothesize that

the periodontal ligament (PDL) may play a role in stimulating osteoclast formation at this

site. The major cell type in the periodontal ligament, the PDL fibroblasts, were shown to

stimulate osteoclastogenesis in vitro [4,5]. In direct contact with osteoclast precursors, PDL

fibroblasts produced osteoclast stimulating factors such as RANKL and M-CSF [13].

In the current study, we investigated whether BP exposure to human PDL

fibroblasts affects PDL-fibroblast-mediated osteoclastogenesis in vitro. PDL fibroblasts were

pre-treated with pamidronate for 24 hours, rinsed with PBS, and human peripheral blood

mononuclear cells (PBMCs) were seeded on top of the fibroblasts. The effect of BP

treatment on markers of osteoclastogenesis and on osteoclast formation was studied. With

this study, we aim to get more insight into the effect of bisphosphonates on PDL fibroblasts

and their ability to induce osteoclastogenesis.

Chapter 4

72


Isolation of PDL fibroblasts and cell culture

Experiments with human PDL fibroblasts were approved by the Medical Ethical Committee of

the VU University and informed consents from healthy donors undergoing third molar

extractions were obtained. Cells were isolated and cultured as described previously with

minor modifications in the culture conditions [4,13]. Briefly, fourth to sixth passage PDL

fibroblasts (1.5x104 viable cells) were seeded in 48-well plates and allowed to attach for 24

h. Medium was replaced and pamidronate (PAM; Sigma-Aldrich, St. Louis, MO, USA) or

vehicle was added in several concentrations. Another 24 h later, PAM was removed, cells

were rinsed with PBS and freshly isolated PBMCs (5x105) were added on top. For mono-

cultures without PBMCs, cells were also washed with PBS and complete medium (DMEM;

GIBCO, Paisley, UK, containing 10% FCS; Hyclone, Logan, UT, USA, and 1% antibiotic

antimycotic solution; Sigma, St. Louis, MO, USA) was added. Medium was replaced twice a

week and micrographs were taken weekly. At 21 days, cells were fixed with 4% phosphate-

buffered formaldehyde and stored at 4°C for TRACP staining, or lysed with RLT lysis buffer

(Qiagen, Hilden, Germany) containing β-mercaptoethanol and stored at -80°C until RNA

isolation.

Cell viability and CyQuant Assay

Directly after PAM treatment, PDL fibroblasts were detached from the plastic by a mild

trypsin treatment, were counted and viability was determined using a Count and Viability

Assay which is based on membrane permeability, and a Muse Cell Analyzer (Merck Millipore,

Darmstadt, Germany). Cells were also lysed and DNA content was measured using the

CyQUANT Cell Proliferation Assay according to manufacturer’s instructions (Invitrogen,

Molecular Probes Carlsbad, CA, USA).


RNA from co-cultures and PDL fibroblast mono-cultures was isolated with the RNeasy Mini

Kit (Qiagen, Hilden, Germany) according to manufacturer’s instructions. Subsequently, cDNA

was synthesized using the MBI Fermentas cDNA synthesis kit (Fermentas, Vilnius,

Lithuania). Real time qPCR was performed on an ABI Prism 7000 using SYBR Green

mastermix (Applied Biosystems, Foster City, CA, USA) as described previously [14]. Gene

expression was normalized for the housekeeping gene β2-microglobulin and primers had the

following sequences (5’�3’): β2-microglobulin forward (FW): AAgATTCAggTTTACTCACgTC,

β2-microglobulin reverse (RV): TgATgCTgCTTACATgTCTCg; Ki67 FW:


73

CgAgACgCCTggTTACTATCAA, Ki67 RV: ggATACggATgTCACATTCAATACC; TRACP FW:

CACAATCTgCAgTACCTgCAAgAT, TRACP RV: CCCATAgTggAAgCgCAgATA

TRACP staining

Tartrate-resistant acid phosphate (TRACP) staining was performed with the leukocyte acid

phosphatase kit (Sigma-Aldrich) and nuclei were visualized with 4’6-diamidino-2-

phenylindole dihydrochloride (DAPI). Five micrographs per well were taken in standardized

regions, using a 10x magnification with an inverted microscope (Leica Microsystems GmbH,

Wetzlar, Germany). The number of TRACP-positive, multinucleated (≥3 nuclei) cells was

assessed in those 5 fields.


Gene expression data were analyzed using a Repeated Measures ANOVA, followed by

Dunnett’s Multiple Comparison tests. All other data were analyzed with a Friedman test,

followed by Dunn’s Multiple Comparison to compare BP-treated samples with the controls.

Graphpad Prism 5 Software was used (GraphPad Software, Inc., La Jolla, CA, USA) and

graphs show the mean and standard deviations of 3-4 healthy donors.

Chapter 4

74

Results

PAM does not affect PDL fibroblast viability in a short term

To investigate the effect of BP treatment on PDL fibroblast-mediated osteoclast formation,

we pre-treated PDL fibroblasts for 24 hours with the bisphosphonate PAM. After 24 hours of

pre-treatment, the viability of those cells was not affected by PAM (Figure 1A), neither were

the DNA (Figure 1B) and RNA concentration (Figure 1C). Also, gene expression of the

proliferation marker Ki67 was not affected by 24 hours of PAM treatment (Figure 1D). At

this time point, PAM was washed away with PBS after which PBMCs were added. These data

indicate that at the time PBMCs were added, the viability of PDL fibroblasts is not affected

by the PAM treatment.

Figure 1. PDL fibroblast viability was unaffected after 24 h of PAM. (A) The viability of PDL fibroblast

was unaffected by 24 h of PAM. (B) The DNA and (C) RNA content were also unaffected after 24 h. (D)

Gene expression of the proliferation marker Ki67 was not significantly affected by 100 µM PAM. One

representative experiment of 2 is shown (A-C), the mean and standard deviations of n=4 donors are

shown (A-D).


75

PAM reduces the number of PDL fibroblasts in the long term

After 24 hours of treatment with 100 µM PAM, no striking differences were observed

compared to control cultures (Figure 2A,B). After 7 days of co-culture with PBMCs, no PDL

fibroblasts could be identified anymore after the pre-treatment with 100 µM PAM (Figure

2D). After 21 days of culture, all the PDL fibroblasts, as well as the PBMCs had died (Figure

2E,F). This indicates that a short-term treatment with PAM reduces the number of PDL

fibroblasts in the long term. With lower concentrations of PAM, no effect on the number

fibroblasts was noted (not shown).

Figure 2. Micrographs of PDL fibroblasts (t0) and co-cultures of fibroblasts with PBMCs (t7,t21). (A-B)

Directly after PAM treatment, PDL fibroblasts were not strikingly different from the controls. (C-D) After

7 days of culture in the absence of PAM, pre-treated PDL fibroblasts were almost absent from cultures.

PBMCs were clearly visible at this time point. (E,F) PBMCs were also inhibited after a longer culture

period. Bars: 100 µm.

Chapter 4

76

Pre-treatment of PDL fibroblasts with 100 µM PAM inhibits osteoclast formation in co-culture

PDL fibroblasts were treated for 24 hours with PAM, rinsed with PBS, and PBMCs were

added. After 21 days, the number of osteoclasts (TRACP-positive cells with 3 or more nuclei,

Figure 3A) was counted. Low concentrations of 1 and 10 µM PAM did not affect PDL

fibroblast-mediated osteoclastogenesis (Figure 3B). Yet, a pre-treatment of the fibroblasts

with 100 µM PAM completely blocked osteoclastogenesis (Figure 3B). Likely, this was related

to the PDL fibroblasts disappearing under the influence of PAM (Figure 2). The 10-fold lower

amount of RNA (p<0.01) isolated from these cell is further indicative that most cells had

disappeared after treatment with 100 µM PAM. The cells that were still present in the

culture, did express a similar level of TRACP mRNA as controls, indicating that the few

remaining cells were most likely PBMCs directed towards the osteoclast lineage.

Figure 3. Osteoclast formation was completely abolished in co-culture with 100 µM PAM pre-treated

PDL fibroblasts. (A) Osteoclasts (arrows) formed from PBMCs in co-culture with PDL-fibroblasts, bar: 50

µm. (B) Pre-treatment of PDL fibroblasts with 1 and 10 µM PAM did not affect their osteoclast-inducing

potential, however, 100 µM completely abolished osteoclast formation (n.d.: not detectable). (C) TRACP

expression by mono-cultures and co-cultures was not affected by PAM pre-treatment of PDL fibroblasts.

The mean and standard deviations of n=4 donors are shown.


77

Apart from the absence of multinucleated osteoclasts (Figure 3A), PDL fibroblasts

caused a dramatic decrease in the number of TRACP-positive mononuclear cells when 100

µM PAM was added in pre-culture (Figure 4A). In this condition, fewer cells were present

after 21 days of culture than in mono-cultures of PBMCs (Figure 4B). These data suggest

that PDL fibroblasts that are pre-treated with BPs can be lethal to PBMCs.

Figure 4. PAM pre-treatment of PDL fibroblasts is toxic to PBMCs. (A) Co-culture of PBMCs with pre-

treated PDL fibroblasts, where PBMCs have almost completely died. (B) Mono-culture of PBMCs where

more cells have survived. Bars: 100 µm

Chapter 4

78

Discussion

Human periodontal ligament fibroblasts are capable of inducing osteoclastogenesis from

human PBMCs in vitro [4,5,13]. We investigated whether this process was affected by

bisphosphonates. Treatment with PAM for 24 hours did not affect the viability of PDL

fibroblasts. Similar results were found in a previous study, where 30 and 100 µM zoledronic

acid were not toxic to PDL fibroblast after 24 hours, whereas they were toxic after 48 hours

[9]. Agis and coworkers showed that transient exposure to ZOL for only 1 hour resulted in

toxicity after 48 hours. This is in line with our findings, where PAM was not toxic after 24

hours, whereas at later time points and after removal of PAM, the majority of the PDL

fibroblasts had died. These data indicate that although PAM did not affect cell viability after

24 hours of treatment, the PDL fibroblasts are affected later on. Possibly, we were not able

to detect decreased cell viability at 24 hours yet, since the viability test we used is based on

membrane permeability, which is a late marker of cell death [15]. Also, we found that

proliferation as measured by Ki67 gene expression was slightly reduced by 24 hours of

treatment with 100 µM PAM.

So far, no reports have described the effect of BPs on PDL-fibroblast-mediated

osteoclastogenesis. Osteoblasts from human bone biopsies were shown to secrete more

OPG [16,17], and express less RANKL after BP treatment [16]. Yet, it is not known whether

these changes affect osteoclastogenesis. Here, we showed that PDL fibroblast death after 1

day of 100 µM PAM resulted in a completely abolished osteoclastogenesis, whereas lower

concentrations did not affect PDL fibroblasts or their effect on osteoclast formation. These

data may indicate that after treatment of PDL fibroblasts with 100 µM PAM,

osteoclastogenesis did not occur due to the absence of PDL fibroblasts. There was no

additional effect of BPs on osteoclastogenesis through PDL fibroblasts.

Based on our previous study we questioned whether osteoclast formation at the

murine molar root induced by BPs is regulated by the adjacent periodontal ligament. The

data provided in the current study may indicate that BPs do not affect PDL-fibroblast-

induced osteoclast formation in vitro. BP treatment, however, increases bone volume in

vivo, thereby likely increasing the compressive loading on the PDL. Therefore it would be of

interest to study the effect of BPs on the osteoclast-stimulating potential of PDL fibroblasts

under compressive loading [18].

Interestingly, after BPs were washed away, the fibroblasts treated with 100 µM

PAM were toxic to PBMCs. Perhaps, this was the result of BP that was released from dying

fibroblasts, making the BP available for uptake by PBMCs. Another explanation could be that

BP-treated and dying PDL fibroblasts express or release factors that are toxic to PBMCs. The


79

mechanism behind this, and the possible effect of lower BP concentrations on PDL-mediated

toxicity to PBMCs requires further investigation.

Taken together, we demonstrated that a high concentration of BPs was toxic to

PDL fibroblasts in a long term, leading to completely abolished osteoclast formation and

toxicity to precursors. Non-toxic concentrations of BPs to PDL fibroblasts did not affect

osteoclast formation induced by those cells.

Chapter 4

80

Acknowledgments

The authors would like to thank Jolanda Hogervorst for excellent technical assistance.


81

References

1. Sims NA, Martin TJ. Coupling the activities of bone formation and resorption: a multitude of

signals within the basic multicellular unit. BoneKEy reports. 2014;3:481.

2. Lacey DL, Timms E, Tan HL, Kelley MJ, Dunstan CR, Burgess T, et al. Osteoprotegerin ligand

is a cytokine that regulates osteoclast differentiation and activation. Cell. 1998;93(2):165-76.

3. Wiktor-Jedrzejczak W, Bartocci A, Ferrante AW, Jr., Ahmed-Ansari A, Sell KW, Pollard JW, et

al. Total absence of colony-stimulating factor 1 in the macrophage-deficient osteopetrotic

(op/op) mouse. Proc Natl Acad Sci U S A. 1990;87(12):4828-32.






research. 2001;80(3):887-91.

6. Allen MR, Follet H, Khurana M, Sato M, Burr DB. Antiremodeling agents influence osteoblast

activity differently in modeling and remodeling sites of canine rib. Calcif Tissue Int.

2006;79(4):255-61.



8. Pozzi S, Vallet S, Mukherjee S, Cirstea D, Vaghela N, Santo L, et al. High-dose zoledronic acid

impacts bone remodeling with effects on osteoblastic lineage and bone mechanical

properties. Clinical cancer research : an official journal of the American Association for Cancer

Research. 2009;15(18):5829-39.






11. Bellido T, Plotkin LI. Novel actions of bisphosphonates in bone: preservation of osteoblast and

osteocyte viability. Bone. 2011;49(1):50-5.

12. Plotkin LI, Weinstein RS, Parfitt AM, Roberson PK, Manolagas SC, Bellido T. Prevention of

osteocyte and osteoblast apoptosis by bisphosphonates and calcitonin. The Journal of clinical

investigation. 1999;104(10):1363-74.

13. Bloemen V, Schoenmaker T, de Vries TJ, Everts V. Direct cell-cell contact between periodontal

ligament fibroblasts and osteoclast precursors synergistically increases the expression of

genes related to osteoclastogenesis. Journal of cellular physiology. 2010;222(3):565-73.



2011;88(1):63-74.

Chapter 4

82

15. Darzynkiewicz Z, Li X, Gong J. Assays of cell viability: discrimination of cells dying by

apoptosis. In: Darzynkiewicz Z, Robinson JP, Crissman HA, editors. Methods in Cell Biology.

41: Academic Press, Inc.; 1994. p. 15-38.

16. Pan B, Farrugia AN, To LB, Findlay DM, Green J, Lynch K, et al. The nitrogen-containing

bisphosphonate, zoledronic acid, influences RANKL expression in human osteoblast-like cells

by activating TNF-alpha converting enzyme (TACE). J Bone Miner Res. 2004;19(1):147-54.

17. Viereck V, Emons G, Lauck V, Frosch KH, Blaschke S, Grundker C, et al. Bisphosphonates

pamidronate and zoledronic acid stimulate osteoprotegerin production by primary human

osteoblasts. Biochem Biophys Res Commun. 2002;291(3):680-6.

18. Kanzaki H, Chiba M, Shimizu Y, Mitani H. Periodontal ligament cells under mechanical stress

induce osteoclastogenesis by receptor activator of nuclear factor kappaB ligand up-regulation

via prostaglandin E2 synthesis. J Bone Miner Res. 2002;17(2):210-20.

Chapter 5.

Osteoclast fusion and fission

Ineke D.C. Jansen

Jenny A.F. Vermeer

Veerle Bloemen

Jan Stap

Vincent Everts

Published in: Calcif Tissue Int. 2012 Jun;90(6):515-22

Chapter 5

84

Abstract

Osteoclasts are specialized multinucleated cells with the unique capacity to resorb bone.

Despite insight in the various steps of the interaction of osteoclast precursors leading to

osteoclast formation, surprisingly little is known about what happens with the

multinucleated cell itself after it has been formed. Is fusion limited to the short period of its

formation or have osteoclasts the capacity to change their size and number of nuclei at a

later stage? To visualize these processes we analyzed osteoclasts generated in vitro with M-

CSF and RANKL from mouse bone marrow and native osteoclasts isolated from rabbit bones

by live cell microscopy. We showed that osteoclasts fuse not only with mononuclear cells

but also with other multinucleated cells. The most intriguing finding, was fission of the

osteoclasts. Osteoclasts were shown to have the capacity to generate functional

multinucleated compartments as well as compartments that contained apoptotic nuclei.

These compartments were separated from each other; each giving rise to a different

functional osteoclast or to a compartment that contained apoptotic nuclei. Our findings

suggest that osteoclasts have the capacity to regulate their own population in number and

function, probably to adapt quickly to changing situations.


85

Introduction

Osteoclasts are multinucleated, polarized cells, with a unique function: resorption of

mineralized substrates such as bone, dentin and mineralized cartilage. They originate from

mononuclear hematopoietic cells of the monocyte lineage. Supported by osteoblasts and

bone-lining cells these mononuclear cells fuse and form multinucleated tartrate resistant

acid phosphatase (TRACP) positive polarized cells [1]. The process of differentiation and

fusion is modulated by the cytokines M-CSF and RANKL which are expressed in vivo by

osteoblast-like cells.

The formation of multinucleated bone resorbing osteoclasts is a multistep process

comprising (i) recruitment of mononuclear precursors from the bone marrow or peripheral

blood, (ii) attraction of these cells by bone-lining cells to the bone site where resorption is

needed, (iii) attachment of the precursors to the bone-lining cells [2], (iv) a subsequent

differentiation of the attached precursors into mononuclear TRACP-positive cells, (v)

migration of these osteoclast precursors to the mineralized surface, and finally, (vi) fusion

and the formation of multinucleated osteoclasts.

Cell-cell interaction between the osteoblast-like bone-lining cells and osteoclast

precursors is crucial in these processes and it has been shown that this interaction

significantly alters gene expression and highly promotes the formation of osteoclasts [2,3].

Zambonin et al. showed already in 1984 with live cell imaging that monocytes fuse with

osteoclasts and that these cells actively migrated to and from each other prior to the actual

fusion, in this way allowing contact by continuous formation and retraction of lamellipodia

and filopodia [4]. Despite insight in the various steps of osteoclast precursor and osteoclast

interaction, surprisingly little is known about what happens with the multinucleated cell itself

after it has been formed. Is fusion limited to the short period of its formation or have

osteoclasts the capacity to change their size and number of nuclei at a later stage, thus

responding to new situations in bone degradation during their life span? Is it possible that

next to fusion of mononuclear cells with multinucleated ones, also multinucleated cells fuse

with each other? Is even the alternative possible that multinucleated osteoclasts split up in

more than one different multinucleated cells? To gain insight in these different possibilities

we made use of a live cell imaging approach and visualized the interaction of osteoclast

precursors and mature osteoclasts during a period of several days.

Chapter 5

86


Mouse bone marrow cell culture with RANKL and M-CSF for the generation of osteoclasts

Osteoclasts were generated as described earlier [5]. Briefly, 6 week old C57BL/6J mice were

sacrificed following a lethal peritoneal injection of sodium pentobarbital. Tibiae were

dissected, cleaned of soft tissue, and ground in a mortar with alpha Minimal Essential

Medium (α-MEM Invitrogen, Paisley, UK) supplemented with 5% fetal calf serum (FCS,

HyClone, Logan, UT), 100 U/ml penicillin, 100 µg/ml streptomycin and 250 ng/ml

amphotericin B (Antibiotic Antimycotic solution, Sigma, St. Louis, MO), and heparin (170

IE/ml; Leo Pharmaceutical Products B.V., Weesp, The Netherlands). The cell suspension was

aspirated through a 21-gauge needle and filtered over a 70 µm pore-size Cell Strainer filter

(Falcon/Becton Dickinson, Franklin Lakes, NJ, USA). Cells were washed in culture medium,

centrifuged (5 min, 200 g), and plated (1.6 x 106 cells/ml) in 2-well glass-bottom chamber

slides (Lab-Tek II, Nunc, Roskilde, Denmark) with 1 ml culture medium containing 30 ng/ml

recombinant murine M-CSF (R&D systems, Minneapolis, MI) and 20 ng/ml recombinant

murine RANKL (R&D systems), 5% fetal calf serum and antibiotics. The chamber slides were

coated with carbon to promote cell attachment and spreading [6]. Culture media were

refreshed on the third day and the cells were cultured for another 68 h while they were

simultaneously followed by live cell imaging. In addition, cells (1.3 x 105 cells/ml) were

seeded on bovine cortical bone slices with a thickness of 650 µm.

Native osteoclasts

Native osteoclasts were isolated from 5-day-old New Zealand White rabbits. Calvariae and

long bones (tibiae) were dissected and collected in 10 ml α-MEM, with 1% antibiotics but

without FCS. The bones were cut into very small fragments and this homogenate was

transferred to a 50 ml tube in 35 ml α-MEM without FCS and with 1% antibiotic antimycotic

solution. The fragments were gently shaken for 30 seconds to release the osteoclasts from

the bone. After 90 seconds of sedimentation the supernatant was collected. The last part of

the procedure was repeated once more with 25 ml of α-MEM. Supernatants were collected

and centrifuged for 2 min at ambient temperature at 700 rpm. The pellet, containing the

osteoclasts, was washed once with 50 ml α-MEM containing 5% FCS, subsequently

centrifuged and collected in 10 ml of α-MEM containing 5% FCS and 1% antibiotics and

finally seeded in 25 cm2 costar (Corning Inc., Corning, NY, USA) culture flasks. After 48 h at

37°C in an atmosphere containing 5% CO2 the osteoclasts were monitored for 80 h by time-

lapse microscopy as described below.


87

Time-lapse microscopy and image processing

Cells were imaged using a Leica IR-BE (Leica Microsystems GmbH, Germany) inverted wide

field microscope at 37°C in an atmosphere containing 5% CO2 [7]. Phase contrast images

were acquired at 5 or 10 min time intervals using a x40 objective. Multi-field imaging

allowed simultaneous monitoring of different sites in one flask or well. Images were

processed and analyzed using custom-made software and Image Pro Plus

(Mediacybernetics, Carlsbad, CA, USA). The movies described in this chapter can be found

on line through the following link: http://link.springer.com/article/10.1007%2Fs00223-012-

9600-y.

Immunolocalisation of CD31, Ly-6C, F4/80, Moma2, ICAM1, and MMP9 in osteoclastogenesis cultures

Osteoclastogenesis cultures were performed as mentioned above and they were fixed after

3 and 4 days of culture with 4% PBS buffered formaldehyde and subsequently washed with

PBS. Before incubation with the primary antibodies, aspecific binding was blocked with

“image it Fx signal enhancer” (Invitrogen/Molecular Probes, Carlsbad, CA, USA) for 30 min

at ambient temperature. Primary antibodies were: anti-MMP9 (goat anti-mouse MMP9, R &

D, used in a 1:100 dilution in PBS), anti-ICAM1 (rat anti-mouse ICAM1, R&D, 1:100 diluted

in PBS), anti-CD31 (ER-MP12), anti-Ly-6C (ER-MP20), anti-Moma2, and anti-F4/80 (the last

four antibodies were all rat anti-mouse and a gift of P.Leenen, Erasmus University,

Rotterdam, The Netherlands; these antibodies were used in a 1:20 dilution in PBS).

Incubation was at 4°C overnight and subsequently by 1 h at ambient temperature, then

washed 2 times with PBS and subsequently incubated for 2 h with a goat anti-rat Alexa 488

(Invitrogen; for MMP9) or goat anti-mouse Alexa-488 (Invitrogen; for CD31, Ly-6C, Moma2,

F4/80, ICAM1). Nuclei were visualized with a DAPI staining (1.5 µg/ml DAPI for 10 min).

After intensive washing, the procedure was finished by adding a drop of vectashield to

enhance the fluorescence. Staining was visualized by a Leica IMDR converted fluorescence

microscope equipped with a digital camera (Leica DFC 320).

Actin and CD44 staining of osteoclasts generated from mouse bone marrow

Mouse bone marrow cells were seeded on cortical bone slices and osteoclastogenesis was

induced during a culture period of 8 days in the presence of M-CSF and RANKL, as described

above. Osteoclast plasma membranes were visualized by staining these with anti-CD44 as

described previously [5]. In short, bone slices were washed in PBS, fixed in 4% PBS

buffered formaldehyde for 5 min, and subsequently washed in PBS. Non-specific binding to

cells was blocked for 30 min with 10% normal goat serum (Vector Laboratories, Burlingame,

CA) followed by an overnight incubation at 4°C with rat anti-mouse CD44 antibody (IM7.8.1)

Chapter 5

88

1:200 in PBS/1%BSA (Cedarlane Laboratories Ltd., Burlington, Canada). Subsequently,

slices were washed three times with PBS and incubated for 60 min with the secondary goat

anti-rat Alexa 647-conjugated antibody (Invitrogen). Following three PBS washes, F-actin

was stained as described previously [8] using Alexa 488-phalloidin (Invitrogen). Finally,

nuclei were stained with propidium iodide (Sigma). Image stacks were generated with a

confocal laser scanning microscope (Leica) using an argon laser (for Alexa 488 and

propidium iodide) and a helium laser (for Alexa 647).


89

Results

Formation of multinucleated cells by fusion

Bone marrow cells isolated from mouse tibiae were cultured on plastic in the presence of M-

CSF and RANKL and monitored after 3 days of culture by live cell imaging for a subsequent

68 hours. Frequently, fusion was noted between mononuclear cells, but also between two

multinucleated cells and between a mononuclear cell and a multinucleated cell. Prior to

fusion the cells migrated towards each other and subsequently made contact as if to find an

appropriate site for fusion. They interacted with each other by membrane extensions. These

interactions were characterized by a relatively short moment of contact with the plasma

membrane of the neighboring cell (Figures 1, 2; Movie A:

http://link.springer.com/article/10.1007%2Fs00223-012-9600-y, Supplementary Material).

During most fusions, next to the fusing cells a round mononuclear cell was seen in the direct

vicinity (Figure 3).

Figure 1. Mouse bone marrow cells were pre-cultured for 3 days in the presence of M-CSF and RANKL.

Culture media were refreshed on day three and the cells were cultured for another 68 h and

simultaneously followed by live cell imaging. Fusion is seen of a multinucleated with another

multinucleated osteoclast (OC). (A-B) Before fusion, the cells make contact with each other (arrow in A

and B) as if to find the appropriate site to fuse. (B) Cells are in close contact with each other. In C

fusion has occurred.

Chapter 5

90

Figure 2. Mouse bone marrow cells (cultured in α-MEM with M-CSF and RANKL) were followed by live

cell imaging for 68 h after a pre-culture period for 3 days. In the micrograph, fusion (arrow) is shown of

a mononuclear cell (mnc) with a multinucleated osteoclast (OC).

Native and in vitro generated osteoclasts can undergo fission

Next to in vitro generated osteoclasts, we used native osteoclasts isolated from the rabbit.

We choose the rabbit for this purpose since rabbit osteoclasts are much easier to isolate

than native osteoclasts from mice.

Isolated native rabbit osteoclasts together with co-isolated osteoblast-like cells

were cultured (ex vivo) and monitored for 4 days. Initially, osteoblast-like cells encircled the

osteoclast, leaving a relatively small cell-free space between them and the osteoclast. The

osteoclast appeared to make contact with those encircling osteoblasts by cellular extensions

that touched upon the surrounding cells (Movie B). During the culture period, the density of

osteoblast-like cells increased due to their proliferation and the cell-free area eventually

became occupied by these cells.

The osteoclast moved quite extensively and during this movement the osteoblast-

like cells made space for the osteoclast. During these activities the osteoclast formed

different compartments that were connected to each other with thin tubular cytoplasmic

bridge-like structures. Each compartment thus formed contained a number of nuclei. The

thin tubular cytoplasmic structures bridged relatively long distances; distances up to 150 µm

were seen to span between the different parts of the osteoclast. These tubular structures

were not firmly attached to the bottom because osteoblasts were able to move underneath

them (Figure 4C; Movie B). The different compartments were highly motile and migrated

away from each other, hereby elongating the tubular connection (Figure 4B,D).


91

Alternatively, the compartments moved towards each other again, in the meantime

shortening the tubular connecting structures. The moment the connections became very

thin and long they often broke, resulting in the generation of two separate multinucleated

osteoclasts (Figure 4E).

This process of fission resulted in the generation of two or more osteoclasts, each

containing a number of nuclei. The separation of the “new” cells could be either

simultaneously or sequentially, thus multinucleated osteoclasts could split directly in three

cells or first in two followed by another round of fission. Strikingly, we observed that the just

separated cell bodies could return to each other and then fuse again.

The phenomenon of fission was also seen with mouse osteoclasts that were

generated in vitro on plastic or on cortical bone slices. The osteoclasts generated on plastic

were followed for 68 hours by live cell imaging (Figure 5; Movie C). Also here tubular

cytoplasmic structures were formed between multinucleated compartments, which was

followed by fission.

Figure 3. Mouse bone marrow cells cultured for 6 days with M-CSF and RANKL. After refreshment of

the media at day 3 the cells were followed by time-lapse imaging. Fusion is shown of a large osteoclast

with a smaller one. Note the two small mononuclear cells (smc) that are present in the direct vicinity of

the site where fusion occurs.

Chapter 5

92

Figure 4. Co-culture of mature rabbit osteoclasts and osteoblast-like cells. (A) Osteoclast (OC) forms

different compartments (C1, C2, C3), shown in (B-E) that are connected to each other by thin tubular

structures (closed arrow in B, D and E). Each compartment contains a number of nuclei. These tubular

structures were not firmly attached to the bottom of the culture well because osteoblasts were able to

move underneath (asterisks in C and D). Following elongation, the connections became very thin and

they often broke, resulting in the generation of two separate multinucleated osteoclasts (E; OC1, OC2).

Time scale of the micrographs: Micrograph A is made after 13 h of culturing, 11 h later micrograph B

was taken, and C, D and E were taken thereafter every 3 h.

Figure 5. Mouse bone marrow cells pre-cultured for 3 days in the presence of M-CSF and RANKL.

Culture media were refreshed on day three and the cells were cultured for another 68 h and

simultaneously followed by live cell imaging. Tubular cytoplasmic structures (arrow) were formed

between multinucleated compartments (C1,C2). Just prior to breaking up of the connection between

the compartments small mononuclear cells (smc) moved across the bridging extensions and at the site

where these cells made contact the extension was broken. Two osteoclasts (B; OC1,OC2) were formed.

phenomenon. Small, very motile mononuclear cells moved across the bridging extensi

the site where contact between the mononuclear cell and the cytoplasmic bridge occurred

the extension was broken. This observation strongly suggests that separation of the

connection was mediated by this small mononuclear cell. Such cell

the connecting tubular structures occurred very frequently; it was found in 98% of the

separation events (Figure 5; Movie C). To investigate the nature of this mononuclear cell we

used a series of antibodies directed against certain subsets of

an anti

Ly

the myeloid lineage and was differentiated into

No positive labeling for this small mononuclear cell was found for CD31, Moma2, and F4/80

(not shown).

Figure 6.

in t

anti

labeled mononuclear cell. The asterisk marks the site where

cytoplasmic extension that connects different osteoclast parts. OC: osteoclast.








an anti

Ly-6C, ICAM1, and MMP9. The positive labeling of Ly



(not shown).

Figure 6.

in the separation of the osteoclast compartments.

anti-MMP9










an anti-ICAM1 antibody and one against MMP9. The small cells were positively labeled for

6C, ICAM1, and MMP9. The positive labeling of Ly



(not shown).

Figure 6.

he separation of the osteoclast compartments.

MMP9



During the process of the breaking up of the connection, we noted an intriguing








ICAM1 antibody and one against MMP9. The small cells were positively labeled for




(not shown).

Figure 6. Green fluorescent staining (Alexa


MMP9, or















(not shown).

Green fluorescent staining (Alexa


or (C)

















(C) anti

















nti-ICAM1

















ICAM1

















ICAM1. Nuclei stained with DAPI show up in blue. The arrow indicates the

















. Nuclei stained with DAPI show up in blue. The arrow indicates the

















































Green fluorescent staining (Alexa-488) of the small mononuclear cell that could be involved

















488) of the small mononuclear cell that could be involved

he separation of the osteoclast compartments. (A)














the myeloid lineage and was differentiated into a myeloid blast or monocyte



(A) Cells were labe














a myeloid blast or monocyte



Cells were labe













6C, ICAM1, and MMP9. The positive labeling of Ly-6C showed that this cell belonged to




Cells were labe


labeled mononuclear cell. The asterisk marks the site where the labeled cell is in close contact with the











6C showed that this cell belonged to




Cells were labe


the labeled cell is in close contact with the















Cells were labeled with anti












used a series of antibodies directed against certain subsets of mononuclear cells as well as






led with anti









connection was mediated by this small mononuclear cell. Such cell-mediated se



mononuclear cells as well as






led with anti









mediated se









led with anti-








mediated se









-Ly-6C (ER








mediated se






a myeloid blast or monocyte [9]



6C (ER








mediated separations of






[9] (Figure 6).



6C (ER-








parations of






(Figure 6).



-MP20)





phenomenon. Small, very motile mononuclear cells moved across the bridging extension. At



parations of






(Figure 6).



MP20),




93


on. At



parations of






(Figure 6).



, (B)




on. At



parations of






(Figure 6).



(B)



Chapter 5

94

Some of the newly formed osteoclasts had the appearance of an apoptotic cell;

their shape became more round and they partially detached from the surface, but after a

while they attached again and fused with other mononuclear or multinucleated cells (Figure

7).

The formation of compartments connected by thin extensions was also noted in

cultures of osteoclasts seeded on cortical-bone slices. We were not able to monitor this with

live cell imaging, but frequently osteoclasts were observed consisting of different nuclei-

containing compartments connected with each other by thin cytoplasmic extensions.

To analyze whether the cells were involved in bone resorption, we visualized

filamentous actin with phalloidin 488. We found the presence of actin rings in these different

osteoclast compartments (Figure 8). Next to this, in some of the osteoclast compartments

we observed nuclei which were reduced in size and had an apoptotic appearance (Figure 8).

Figure 7. Fission of an osteoclast following the formation of two compartments (C1 and C2) results in

the formation of two “new” osteoclasts (OC1 and OC2; shown in A and B). Subsequently OC2 fuses with

another multinucleated cell (OC3). Time span between micrographs A and C is 3 h. The separation of

the osteoclast starts 20 hours after the start of the visualization. Note the small mononuclear cells

(smc) close at the thin tubular structure in micrograph A.

Figure 8.

arrows) was also noted with osteoclasts seeded on cortical bone slices. Actin rings (green, thin arrows)

were present in these different osteoclast compartments indicat

membrane is stained for CD44 (blue). Nuclei are red. In osteoclast compartment 2 (C2), the nuclei are

reduced in size and appear apoptotic (arrowheads).

Figure 8.





Figure 8.





Figure 8. The formation of compartments (C1, C2, C3) connected by thin tubular structures (thick





The formation of compartments (C1, C2, C3) connected by thin tubular structures (thick

































































were present in these different osteoclast compartments indicating bone resorption activity. Osteoclast




ing bone resorption activity. Osteoclast











































95









Chapter 5

96

Discussion

We visualized native mature osteoclasts and in vitro generated osteoclasts by live cell

imaging and observed fusion of all possible combinations: mononuclear with mononuclear,

mononuclear with multinucleated, and multinucleated with multinucleated cells. Yet, the

most exciting novel series of observations was the fission of osteoclasts. Multinucleated

osteoclasts proved to have the capacity to split up in different compartments, each part

containing a number of nuclei. Sometimes the nuclei of one of the newly formed parts

seemed to be apoptotic, thus suggesting the ability of the cell to get rid of non-functional

parts of the polykarion. However, the most frequent finding was that the newly formed

osteoclasts appeared to be functional given the clear presence of actin rings and their active

movement.

Osteoclast fusion and fission is probably beneficial for the cell and its functional

properties. The process of fusion and fission is also a common phenomenon occurring in

mitochondria. In these organelles fission and fusion was thought to play a role in apoptosis

and elimination of damaged fragments, but recently it was considered more likely that

fusion and fission act in mitochondrial quality control to form healthy and functional

organelles [10]. In these organelles fusion serves to mix and unify the mitochondrial

compartment, whereas fission generates new mitochondria. Fusion and fission in osteoclasts

can occur for comparable reasons: to form osteoclasts with different subsets of nuclei and

therefore with a different functionality. Recently, Youn et al. [11] reported that only a

limited number of nuclei of a multinucleated osteoclast are transcriptionally active.

Separation of nuclei with different expression patterns can be useful to generate osteoclasts

with somewhat different functions, such as osteoclasts involved in resorption of trabecular

bone and those resorbing cortical bone. In this respect it is of interest to note that Zenger

and Andersson described differences among osteoclasts associated with these different

bone sites [12-14]; reviewed in [15]. But also other functional properties of osteoclasts,

such as their participation in the immune response [16,17] secretion of cytokines,

interaction with osteoblasts and recruitment of mononuclear cells from the bone marrow

[18,19], may lead to the presence of osteoclasts that differ in their nuclear composition.

Fusion and fission of osteoclasts resemble phenomena occurring with the

syncytiotrophoblasts in the placenta. The syncytium is a single multinucleated cell layer that

covers the placenta and is in direct contact with maternal blood [20]. The syncytium

regulates the exchange of nutrients and other compounds between mother and fetus.

Syncytiotrophoblast cells are formed by fusion of cytotrophoblast cells. During this process

the protein syncytin plays an important role [21]. It is of considerable interest to note that

recently syncytin was shown to be expressed also by osteoclasts [22], thus suggesting a


97

similarity between the fusion process of these different cell types. During pregnancy parts of

the syncytiotrophoblast are shed into the maternal blood system. These shed parts contain

not only cytoplasm but also nuclei, a process comparable to the osteoclast fission noted in

the present study.

Prior to fission, tubular cytoplasmic structures bridge the different compartments.

The occurrence of such bridging structures was noted previously by Vesely and coworkers

[23], and Abe et al. [24]. Yet, that these structures may form part of a rather unique

property of osteoclasts, the fission of these cells, has not been described before. Zambonin

and Teti [25] described the presence of cytoplasmic extensions between osteoclast parts

present in medullary hen bones during hypocalcemia and suggested that osteoclasts

probably shed their apoptotic nuclei. They also mentioned the presence of a mononuclear

cell in close connection to the bridging extension. They suggested that this mononuclear cell

could either become part of the osteoclast or was just detached from the osteoclast [25].

We visualized similar mononuclear cells in close relationship to the cellular

extensions between osteoclast parts. This small mononuclear cell was found migrating over

the extension just shortly before the extension broke. Given the observation that the cellular

extension breaks at the site where this mononuclear cell crosses it, we propose an active

participation of these cells in the process of fission/separation. Positive labeling for Ly-6C

showed that this cell, comparable to osteoclast precursors, originates from the monocyte

lineage. The expression of MMP9 suggests that this proteolytic enzyme plays a role in

breaking of the cytoplasmic extension. The high expression level of ICAM1 could imply that

this molecule is involved in the attraction and/or binding of this cell to the cytoplasmic

connection. How these cells perform such a task is unknown and needs further

investigation.

The reason why osteoclasts show fission is not clear yet, but in line with

mitochondria and syncytiotrophoblasts it is reasonable to assume that the osteoclast can

regulate its own activity in this way more efficiently.

Collectively, the data presented in this study provide new insights in the dynamics

of cell-cell interactions during osteoclast formation and show for the first time that mature

osteoclasts can undergo fission and separate themselves into functional smaller, yet still

multinucleated cells.

Fusion and fission of osteoclasts show that osteoclasts are very flexible cells, which

have the capacity to regulate their own population in number and function, probably to

adapt quickly to changing situations.

Chapter 5

98

Acknowledgments

We would like to thank Dr. Teun J. de Vries and Ton Schoenmaker for providing the

confocal image shown in Figure 8.


99

Supplementary data

The movies described here can be found on line on this webpage:

http://link.springer.com/article/10.1007%2Fs00223-012-9600-y.

Movie 1. Mouse bone marrow cells pre-cultured for 3 days in the presence of M-CSF and

RANKL. Culture media were refreshed on day three and the cells were cultured for another

68 hours and simultaneously followed by live cell imaging. Fusion of a multinucleated

osteoclast with other multinucleated osteoclasts can be seen.

Movie 2. Isolated native rabbit osteoclasts together with co-isolated osteoblast-like cells

were cultured (ex vivo) and monitored for 4 days. Initially, osteoblast-like cells encircle the

osteoclast, leaving a relative small cell-free space between them and the osteoclast. The

osteoclast appears to make contact with those encircling osteoblasts by cellular extensions

that touch upon the surrounding cells. During these activities the osteoclast forms

compartments that are connected to each other with thin tubular cytoplasmic, bridge-like

structures. Eventually the connection becomes very thin and breaks, thus generating two

new osteoclasts.

Movie 3. Mouse bone marrow cells pre-cultured for 3 days in the presence of M-CSF and

RANKL. At day 3 medium was refreshed and from then, cells were followed by live imaging

for 68 hours. Tubular cytoplasmic structures are formed between multinucleated

compartments and these connections break. Subsequently, after breaking of the connection

one part of the osteoclast fuses with another osteoclast.

Chapter 5

100

References

1. Boyle WJ, Simonet WS, Lacey DL. Osteoclast differentiation and activation. Nature.

2003;423(6937):337-42.

2. Perez-Amodio S, Beertsen W, Everts V. (Pre-)osteoclasts induce retraction of osteoblasts

before their fusion to osteoclasts. J Bone Miner Res. 2004;19(10):1722-31.

3. Bloemen V, de Vries TJ, Schoenmaker T, Everts V. Intercellular adhesion molecule-1 clusters

during osteoclastogenesis. Biochem Biophys Res Commun. 2009;385(4):640-5.

4. Zambonin Zallone A, Teti A, Primavera MV. Monocytes from circulating blood fuse in vitro

with purified osteoclasts in primary culture. Journal of cell science. 1984;66:335-42.

5. de Vries TJ, Schoenmaker T, Beertsen W, van der Neut R, Everts V. Effect of CD44 deficiency

on in vitro and in vivo osteoclast formation. J Cell Biochem. 2005;94(5):954-66.

6. Stap J, Van Marle J, Van Veen HA, Aten JA. Coating of coverslips with glow-discharged carbon

promotes cell attachment and spreading probably due to carboxylic groups. Cytometry.

2000;39(4):295-9.

7. Krawczyk PM, Stap J, Hoebe RA, van Oven CH, Kanaar R, Aten JA. Analysis of the mobility of

DNA double-strand break-containing chromosome domains in living mammalian cells.

Methods in molecular biology (Clifton, NJ). 2008;463:309-20.

8. van Beek EM, de Vries TJ, Mulder L, Schoenmaker T, Hoeben KA, Matozaki T, et al. Inhibitory

regulation of osteoclast bone resorption by signal regulatory protein alpha. FASEB J.

2009;23(12):4081-90.



10. Otera H, Mihara K. Molecular mechanisms and physiologic functions of mitochondrial

dynamics. Journal of biochemistry. 2011;149(3):241-51.

11. Youn MY, Takada I, Imai Y, Yasuda H, Kato S. Transcriptionally active nuclei are selective in

mature multinucleated osteoclasts. Genes to cells : devoted to molecular & cellular

mechanisms. 2010;15(10):1025-35.

12. Hu Y, Ek-Rylander B, Karlstrom E, Wendel M, Andersson G. Osteoclast size heterogeneity in

rat long bones is associated with differences in adhesive ligand specificity. Experimental cell

research. 2008;314(3):638-50.

13. Zenger S, Hollberg K, Ljusberg J, Norgard M, Ek-Rylander B, Kiviranta R, et al. Proteolytic

processing and polarized secretion of tartrate-resistant acid phosphatase is altered in a

subpopulation of metaphyseal osteoclasts in cathepsin K-deficient mice. Bone.

2007;41(5):820-32.

14. Zenger S, Ek-Rylander B, Andersson G. Long bone osteoclasts display an augmented

osteoclast phenotype compared to calvarial osteoclasts. Biochem Biophys Res Commun.

2010;394(3):743-9.



16. Boyce BF, Yao Z, Zhang Q, Guo R, Lu Y, Schwarz EM, et al. New roles for osteoclasts in bone.

Annals of the New York Academy of Sciences. 2007;1116:245-54.


101

17. Boyce BF, Yao Z, Xing L. Osteoclasts have multiple roles in bone in addition to bone

resorption. Critical reviews in eukaryotic gene expression. 2009;19(3):171-80.

18. Boyce BF, Xing L. Osteoclasts, no longer osteoblast slaves. Nature medicine.

2006;12(12):1356-8.

19. Ishii M, Egen JG, Klauschen F, Meier-Schellersheim M, Saeki Y, Vacher J, et al. Sphingosine-1-

phosphate mobilizes osteoclast precursors and regulates bone homeostasis. Nature.

2009;458(7237):524-8.

20. Guller S. Role of the syncytium in placenta-mediated complications of preeclampsia.

Thrombosis research. 2009;124(4):389-92.

21. Handwerger S. New insights into the regulation of human cytotrophoblast cell differentiation.

Molecular and cellular endocrinology. 2010;323(1):94-104.

22. Soe K, Andersen TL, Hobolt-Pedersen AS, Bjerregaard B, Larsson LI, Delaisse JM.

Involvement of human endogenous retroviral syncytin-1 in human osteoclast fusion. Bone.

2011;48(4):837-46.

23. Vesely P, Boyde A, Jones SJ. Behaviour of osteoclasts in vitro: contact behaviour of

osteoclasts with osteoblast-like cells and networking of osteoclasts for 3D orientation. Journal

of anatomy. 1992;181 ( Pt 2):277-91.

24. Abe K, Ohno K, Hasegawa K. Morphological relationships between osteoclasts and bone

resorption surfaces on mouse parietal bones. Archives of histology and cytology.

1990;53(1):95-102.

25. Zambonin Zallone A, Teti A. The osteoclasts of hen medullary bone under hypocalcaemic

conditions. Anatomy and embryology. 1981;162(4):379-92.

Chapter 6.

Migration, fusion, and CXCL12-

CXCR4-mediated chemoattraction of

long bone and jaw osteoclast

precursors

Jenny A.F. Vermeer

Ton Schoenmaker

Mei-Ling Tsui

Jan Stap

Teun J. de Vries

Vincent Everts

Submitted for publication

Chapter 6

104

Abstract

Bone is important to give the body support, but it also contains marrow that harbors

mesenchymal stem cells and hematopoietic progenitors. An important chemoattractant for

hematopoietic cell homing towards the bone marrow, CXCL12, was shown to inhibit

expression of apoptosis related genes. Previously, it was shown that jaw osteoclast

precursors express more anti-apoptotic genes than long bone osteoclast precursors. Also,

jaw and long-bone marrow cells have a different osteoclastogenic potential. Here, we used

time-lapse microscopy to follow long bone and jaw osteoclast precursors during three

important steps of osteoclastogenesis, i.e. proliferation, migration, and fusion. Gene

expression analyses were performed on markers of the different steps of

osteoclastogenesis, and the directional migration of jaw and long-bone osteoclast precursors

towards CXCL12 was studied. Long bone and jaw osteoclast precursors had similar rates of

proliferation, random and directional migration, and fusion. Jaw osteoclast precursors

expressed more CXCL12 and its receptors CXCR4 and CXCR7 than long bone osteoclast

precursors. Therefore, we provide more evidence that osteoclast precursors are bone-site

specific. Higher CXCL12 expression may explain the higher anti-apoptotic gene expression

that was previously found in jaw cells than in long bone cells.

Migration of long bone and jaw osteoclast precursors

105

Introduction

Bone is continuously remodeled throughout life to keep it as strong as possible. Bone

formation is carried out by osteoblasts derived from mesenchymal cells, whereas osteoclasts

from hematopoietic origin are the bone-resorbing cells. When bone resorption is needed,

osteoclast precursors are attracted from the blood or from the bone marrow and they

migrate towards the sites where resorption has to take place. Upon contact with osteoblasts

they fuse to become multinucleated osteoclasts due to stimulation by the cytokines

macrophage-colony stimulating factor (M-CSF) and receptor activator of nuclear factor κB

ligand (RANKL) [1].

Several specialized cytokines called chemoattractants have been shown to be

involved in the attraction of osteoclast precursors [2,3]. Also, the chemokine stromal cell-

derived factor-1α (SDF-1α or CXCL12) and its receptor CXCR4 were shown to be involved in

osteoclast precursor recruitment [4,5]. The expression of CXCL12 by stromal cells and the

subsequent attraction of hematopoietic cells towards the bone marrow [6] can be

counteracted by sphingosine-1-phosphate (S1P), a lipid that attracts osteoclast precursors,

found at high concentrations in peripheral blood [7,8]. Moreover, CXCL12 has been

associated with cell proliferation, cell survival, and tumor metastasis [4,9,10].

CXCL12 may play a prominent role in diseases such as bone cancer and

osteoporosis, where excess of osteoclast activity is observed, leading to loss of bone mass

and fragile bones. Bisphosphonates (BPs) such as pamidronate can be used to inhibit

excessive bone resorption by causing osteoclast apoptosis and inhibiting their resorbing

activity [11]. A specific type of BPs containing nitrogen (N-BPs) also inhibits the prenylation

of small GTPases [12,13]. As these proteins are important for the cytoskeletal

rearrangements necessary during adhesion and migration, it is likely that osteoclast

precursor migration is affected by N-BPs [14,15].

Previously, we have shown that osteoclasts and their bone marrow precursors are

different in long bones and jaws [16,17]. Long bone and jaw osteoclast morphology were

shown to differ in vitro [18] and long bone osteoclasts differentiated faster than jaw

osteoclasts, which was likely related to the cellular composition of the bone marrows

[17,19]. Also, bone marrow cells from the jaw take up more BPs than those from long bone.

Interestingly, this did not differently affect the number of osteoclasts formed from both cell

types, indicating that the jaw osteoclasts or precursors may be less sensitive to BPs [20].

Here, we investigated whether osteoclast precursors from the jaw can compensate for

cytotoxicity by forming more osteoclasts. Therefore, we studied the proliferation, migration

and fusion of osteoclast precursors from jaw and long bones by time-lapse microscopy. Also,

we compared the expression of CXCR4, CXCR7, and CXCL12, and studied the directional

migration of long bone and jaw bone marrow cells towards CXCL12. With this study, we aim

Chapter 6

106

to get more insight into the differences between long bone and jaw osteoclasts and their

precursors regarding (i) proliferation, (ii) migration, and (iii) fusion, and the genes

regulating these processes.


107


Bone marrow cell isolation and cell culture

Animal experiments were approved by the Animal Welfare Committee of the VU University

(Amsterdam, The Netherlands). Bone marrow cells were isolated from the mandibles and

long bones (tibia and femur) as described previously [20]. Freshly isolated bone marrow

cells were seeded and incubated with 150 µl α-MEM (GIBCO, Paisley, UK) complete culture

medium with 5% fetal calf serum (FCS; Hyclone, Logan, UT, USA) and 1% antibiotic

antimycotic solution containing 30 ng/mL recombinant mouse macrophage-colony

stimulating factor (rmM-CSF; R&D Systems, Minneapolis, MN, USA) and 20 ng/mL

recombinant mouse receptor activator of nuclear factor κB ligand (rmRANKL; R&D Systems)

to stimulate osteoclastogenesis. For gene expression analyses, long bone and jaw bone

marrow cells were seeded in 96 well plates (105 cells per well) and lysed after 1 or 3 days of

culture. Freshly isolated bone marrow cells were also lysed for gene expression analyses

and cell lysates were stored at -80°C until RNA isolation.

Time-lapse microscopy and image analyses

For time-lapse microscopy, cells (8x105 per well) were seeded in 2-well glass chamber slides

(Nunc, Roskilde, Denmark) that were coated with carbon as described previously [21]. Cells

were cultured for 3 days in the presence of 1 mL complete culture medium. Before live

imaging started, cells were rinsed twice with PBS, the medium was replaced with fresh

medium containing 10 µM pamidronate (PAM; Sigma-Aldrich, St. Louis, MO, USA) or control

medium with vehicle, and this was covered with mineral oil (Sigma-Aldrich).

Cells were imaged using a Leica IR-BE (Leica Microsystems GmbH, Germany)

inverted wide field microscope at 37°C in an atmosphere containing 5% CO2. Phase contrast

images, (3-5 locations per chamber) were acquired at 5, 6, or 10 minute time intervals

using a x40 or x10 objective. Images were processed and analyzed using custom-made

software and Image Pro Plus (Media Cybernetics, Carlsbad, CA, USA). For migration data,

cell tracks were followed for 18 h after the start of imaging. The velocity was calculated by

the total distance traveled divided by the time imaged, with a maximum of 18 h. For the

analyses of cell division and cell fusion, cells were followed for 80 h in 3 independent

experiments. Dividing and fusing cells were expressed as percentage of the total number of

cells analyzed (4-23 per movie).


RNA was isolated with the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to

manufacturer’s instructions and cDNA was synthesized using the MBI Fermentas cDNA

Chapter 6

108

synthesis kit (Fermentas, Vilnius, Lithuania). Real time qPCR was performed on an ABI

Prism 7000 using SYBR Green mastermix (Applied Biosystems, Foster City, CA, USA) as

described previously [17], and primers are mentioned in Table 1. Gene expression was

normalized for the housekeeping gene porphobilinogen deaminase (PBGD).

Table 1. Primer sequences used for real time quantitative PCR.

Gene Primer sequence 5’�3’

PBGD FW: AgTgATgAAAgATgggCAACt

RV: TCTggACCATCTTCTTgCTgA

Ki67 FW: CAAAAGGCGAAgTggAgCTT

RV: TgTTTCgCAACTTTCgTTTgTg

DC-STAMP FW: TgTATCggCTCATCTCCTCCAT

RV: gACTCCTTgggTTCCTTgCTT

TRACP FW: gACAAgAggTTCCAggAgACC

RV: gggCTggggAAgTTCCAg

NFATc1 FW: CATgCgAgCCATCATCgA

RV: TgggATgTgAACTCggAAgAC

c-Fos FW: TCACCCTgCCCCTTCTCA

RV: CTgATgCTCTTgACTggCTCC

MafB FW: AACgCgTCCAgCAgAAACAT

RV: CTCAggAgAggAggggCTgT

F4/80 FW: gCACCAATgTACCAggCTCCTA

RV: gATCCTTTTgCAgTTgAAgTTTCC

Runx2 FW: TgCCCAggCgTATTTCAg

RV: TgCCTggCTCTTCTTACTgAg

CXCR4 FW: TggAACCgATCAgTgTgAgT

RV: gggCAggAAgATCCTATTgA

CXCL12 FW: TgTgCATTgACCCgAAATTA

RV: TCTCACATCTTgAgCCTCTTgT

CXCR7 FW: CTCACCgTCAggAAggCAAA

RV: AggCTCTgCATAgTCAAACAAgTg


109

Transwell Migration Assay

The bottom chambers (24-well plate) were filled with 600 µL complete culture medium.

Freshly isolated bone marrow cells were seeded into the upper chamber of cell culture

inserts with 3 µm pore size (ThinCert; Greiner Bio-One, Monroe, NC, USA). Cells were

incubated overnight to allow attachment to the ThinCerts. CXCL12 (R&D systems) was

dissolved in 0.1% bovine serum albumin (BSA; Sigma) in PBS (0.1% PBS/BSA), filter

sterilized, aliquoted, and stored at -20°C. Fresh medium was added in the bottom wells of

new 24-well plates and contained the indicated concentration of CXCL12 or vehicle

(PBS/BSA) as a control. ThinCerts were transferred to the wells containing fresh medium

and incubated for 4 h at 37°C. The cells remaining in the upper chamber were swopped off

the filter, were washed with PBS, and the bottom was fixed with 4% phosphate-buffered

formaldehyde. Filters were stained with DAPI, cut from the ThinCerts, and covered with

Vectashield on a microscope slide. Micrographs of 5 standardized areas were taken and

analyzed using Image Pro-Plus Software (Media Cybernetics, Silver Spring, MD, USA).


Migration was analyzed using a Kruskal-Wallis test followed by Dunn’s Multiple Comparison

tests. A Spearman correlation was used for the correlation between the number of cells and

the velocity. Differences in cell division, fusion between long bone and jaw cells were

analyzed with a Mann-Whitney U test. Gene expression was analyzed with a paired t-test.

All data are expressed as means and standard deviations and differences were considered

significant when p<0.05.

Chapter 6

110

Results

Cell proliferation

An explanation why jaw osteoclasts or precursors may be less sensitive to BPs than those

from long bone [20], may be that jaw osteoclast precursors exhibit an increased or

prolonged precursor proliferation. We compared proliferation of long bone and jaw bone

marrow cells using time-lapse microscopy and expressed the number of cells going through

mitosis as a percentage of the total number of cells. Figure 1 shows the cumulative

percentages at each time point, starting from day 3 of culture with M-CSF and RANKL. The

majority of the dividing cells, divided within the first 18 hours of the experiment (Figure 1A).

There was no difference between cell division in long bone and jaw bone marrow cells

(Figure 1A). Gene expression of the proliferation marker Ki67 was also similar in long bone

and jaw cells (Figure 1B). These data indicate that long bone and jaw osteoclast cultures

have similar proliferation rates.

Figure 1. Long bone and jaw cell proliferation rates were similar. (A) The cumulative number of cells

that had fused at the indicated time point after the start of imaging at day 3 of culture with M-CSF and

RANKL, expressed as a percentage of the total number of cells in the field. A similar number of cell

divisions was seen in long bone and jaw osteoclast precursors. Three to five movies (4-23 cells per

movie) were analyzed in n=3 independent experiments. (B) mRNA expression of the proliferation

marker Ki67 was similar in long bone and jaw bone marrow directly after isolation and after 3 days of

culture with M-CSF and RANKL (n=3 for t0, n=8 for t1 and t3).


111

Long bone and jaw osteoclast precursors migrate at the same velocity

Using time-lapse microscopy, we investigated the random migration of long bone and jaw

bone marrow cells that had been primed towards the osteoclast lineage with M-CSF and

RANKL for 3 days. Cells were followed for up to 18 hours. There was no significant

difference between the migration velocity of long bone and jaw osteoclast precursors

(Figure 2A). The bisphosphonate PAM (10 µM) did not affect the migration velocity of both

cell types (Figure 2A).

Dynamics of fusing cells

In Figure 2A, the effect of PAM on the migration velocity of all visible cells is depicted. When

we distinguished between mononuclear and multinucleated cells in the control condition,

multinucleated cells moved slightly slower than mononuclear cells, both in long bone and in

jaw cultures (not significant, Figure 2B). Also, fusing cells, i.e. the mononuclear and

multinucleated cells that fused within the time span of the experiment, moved slightly

slower than the cells that did not fuse, both in long bone and in jaw cultures (Figure 2C).

Furthermore, there was a negative correlation (Spearman’s rs= -0.47 and p=0.05 for long

bone, Spearman’s rs= -0.45 and p=0.09 for jaw) between the number of cells per field and

the velocity (Figure 2D), suggesting that in the presence of more potential fusion partners,

cells may move slower. Altogether, our results suggest that cells migrate faster when they

do not fuse or when fewer cells are in their vicinity. We also observed that not all cells

adjacent to each other fused, whereas these cells were able to fuse with other cells, further

away (Figure 3A-C), indicating that they did not reach a fusion competent status yet or that

they were not fusion compatible. However, after fusion with other cells, they were in close

contact with each other without fusing again (Figure 3D). These data indicate that fusion is

a selective process and that active migration is necessary to come into contact with fusion

competent and compatible cells.

Cell fusion

The fusion capacity of long bone and jaw osteoclast precursors was assessed in time-lapse

movies from day 3 of culture in the presence of M-CSF and RANKL, until 80 hours after the

start of imaging. The cumulative percentage of cells that had fused at that time point is

depicted in Figure 4a. There was no difference between the fusion rate in long bone and

jaw bone marrow cells (Figure 4A). The number of multinucleated cells present at the start

of imaging was similar in long bone and jaw cultures (Figure 4B).

Chapter 6

112

Figure 2. Long bone and jaw osteoclasts precursors migrate at a similar velocity. Migration velocity

was expressed as µm/h and measured for a maximum of 18 h after the start of imaging at day 3 of

culture. (A) The velocity of long bone and jaw osteoclasts and precursors was similar and unaffected by

10 µM pamidronate (PAM), as measured in 3 or 5 movies (4-23 cells per movie), performed in n=4

(control) or n=3 (PAM) independent experiments. (B) Multinucleated cells moved 1.3 times (long bone)

and 1.8 times (jaw) slower than mononuclear cells (not significant, n=2-4). (C) Fusing cells (n=1 for

jaw) moved 1.4 times (long bone) and 1.2 times (jaw) slower that non-fusing cells (n=3-4). (D)

Negative correlation (Spearman) between the velocity and the number of cells in each field, rs=-0.47

and p=0.05 for long bone, and rs=-0.45 and p=0.09 for jaw.

We also measured the gene expression levels of the osteoclast marker tartrate-

resistant acid phosphatase (TRACP), and dendritic cell-specific transmembrane protein (DC-

STAMP), an essential fusion protein [22]. At day 3, when fusion measurements were

started, long bone cells expressed significantly more TRACP and DC-STAMP than jaw cells

(Figure 5A,B). This may indicate that they were further in differentiation and had a higher

fusion capacity. Also, expression of the transcription factor NFATc1, was higher in the long

bone cells at this time point (Figure 5C). c-Fos expression was similar in the long bone and

in the jaw cells at day 3, whereas it was higher in the jaw cells at day 1 of culture (Figure

5D). Altogether, in line with previously reported data [17], these data indicate that


113

differentiation of long bone osteoclasts happens faster than jaw osteoclast differentiation.

Yet, this was not represented by a difference in fusion rate between long bone and jaw

osteoclast precursors (Figure 4A).

Figure 3. Selective fusion of long bone osteoclast precursors. Cells 1a and 2a are in contact (A), but do

not fuse. Yet, they fuse at a later time-point where they fuse with cells 1b (B) and 2b (C), respectively.

Later on, the fused cells 1ab and 2ab are in contact again without fusing (D). Bar: 20 µm.

Chapter 6

114

Figure 4. Long bone and jaw cell fusion rates were similar. (A) The number of cells that fused as a

percentage of the total number of cells in the field at the indicated time points. (B) The percentage of

multinucleated cells at the start of imaging was similar for long bone and jaw cells. 3 or 5 movies were

analyzed in 3 independent experiments.

Figure 5. Gene expression of osteoclast markers. (A) DC-STAMP was absent directly after isolation and

higher in long bone osteoclast precursors than in jaw osteoclast precursors after 3 days of culture with

M-CSF and RANKL. At this time point, (B) TRACP and (C) NFATc1 expression were also higher in the

long bones. (D) c-Fos expression was higher in the jaw cells directly after isolation, however this

difference disappeared after culture with M-CSF and RANKL. *p<0.05, **p<0.01, using a paired t-test,

n=3 for t0 and n=8 for t1 and t3.


115

Non-osteoclast differentiation

To further investigate why fusion rates in long bone and jaw cultures did not correlate with

markers for osteoclastogenesis, we measured gene expression of differentiation markers of

other cell types present in the bone marrow, i.e. macrophages and osteoblast precursors.

Expression of the macrophage markers MafB and F4/80 was significantly higher in jaw than

in long-bone marrow cells cultured with M-CSF and RANKL (Figure 6A,B). Expression of the

early osteoblast marker Runx2, was slightly higher in jaw cultures than in long bone

cultures, however, this was only significant at the start of culture, i.e. before M-CSF and

RANKL were added (Figure 6C). These results indicate that the presence of other cell types

such as macrophages, may, at least in part, explain the relatively low expression of

osteoclastogenesis markers in jaw bone marrow cultures compared to long bone cultures.

Figure 6. Gene expression of macrophage and osteoblast markers. (A) MafB and (B) F4/80 expression

were higher in jaw than in long bone osteoclast cultures with M-CSF and RANKL. (C) Jaw bone marrow

cells expressed more Runx2 than long-bone marrow cells directly after isolation. *p<0.05, **p<0.01

using a paired t-test, n=3 for t0 and n=8 for t1 and t3.

Chapter 6

116

Directional migration

There was no difference between the migration velocity of long bone and jaw osteoclast

precursors (Figure 2A). However, attraction of osteoclast precursors towards the resorption

site is also an important step during osteoclastogenesis. Therefore, we analyzed gene

expression of CXCL12 and CXCR4, an important chemoattractant-receptor pair that may be

involved in osteoclasts precursor attraction [4,5]. Jaw cells expressed more CXCR4 (Figure

7A) and CXCL12 (Figure 7B) than long bone cells. This was significant for CXCL12 after 1

day of culture and for both genes after 3 days of culture with M-CSF and RANKL.

To assess whether higher CXCR4 expression resulted in an increased attraction of

jaw cells towards CXCL12 we performed a Transwell Migration Assay. Both long bone and

jaw bone marrow cells that were pre-cultured for 1 day with M-CSF and RANKL migrated

towards CXCL12 in a dose dependent manner (Figure 7C). However, migration towards

CXCL12 did not differ between long bone and jaw bone marrow cells, indicating that those

cells similarly responded to CXCL12. Possibly, this could be attributed to higher expression

of the decoy receptor for CXCL12, CXCR7, that accompanied higher CXCR4 expression in

jaw cells (Figure 7D).

Figure 7. Jaw osteoclast precursors express more CXCR4, CXCR7, and CXCL12 than those from long

bone. (A) Gene expression of the chemoattractant-receptor pair CXCR4 and (B) CXCL12 was higher in

the jaw than in long-bone marrow cells cultured with M-CSF and RANKL. (C) Long bone and jaw bone

marrow cells migrated towards CXCL12 in a similar, dose dependent manner. (D) Jaw bone marrow

cells expressed more CXCR7 than long-bone marrow cells after culture with M-CSF and RANKL.

*p<0.05, **p<0.01 using a paired t-test, n=5-7 for t0 and n=8 for t1 and t3.


117

Discussion

Previously, we have shown that despite an enhanced BP uptake in jaw osteoclast

precursors, similar numbers of osteoclasts differentiate from jaw and long bone osteoclast

precursors after exposure to BPs [20]. This could be attributed to a higher capacity of jaw

osteoclast precursors to prevent apoptosis. Another explanation could be that jaw

osteoclasts compensate for cell death by increasing formation in vitro by an altered (i)

proliferation, (ii) migration, and/or (iii) fusion. Here, we showed that jaw and long bone cells

did not have a different capacity in either of those parameters. Yet, we found that long bone

and jaw bone marrow cells express different levels of CXCL12, a regulator of osteoclast

precursor attraction and an inhibitor of pro-apoptotic protein expression [4].

By undergoing fission, we have already shown that osteoclast formation is a very

flexible process [23]. We showed here that osteoclast fusion is also a selective process, and

that osteoclast precursor migration is an important step during osteoclastogenesis in vitro.

Both long bone and jaw cells were flexible, and able to adapt their velocity to the situation,

i.e. they migrated faster when they were mononuclear, when they did not fuse, and when

fewer potential fusing partners were present. Together with previous studies which showed

that human osteoclast precursors migrate further in an early stage of differentiation than at

a later stage [24], our data indicate that both long bone and jaw osteoclast precursors are

actively migrating, probably in search for appropriate fusion partners that are fusion

competent and compatible.

As the small GTPases that are affected by N-BPs are important for the cytoskeletal

rearrangements necessary during adhesion and migration, it is likely that osteoclast

precursor migration is affected by N-BPs [14,15]. Surprisingly, pamidronate (10 µM) did not

affect the migration velocity of the cells. Perhaps, the concentration used in these

experiments was too low to cause an effect on small GTPase function. Yet, a higher

concentration of 50 µM caused cell death within 6 hours after addition (not shown). These

data may indicate that PAM does not cause an effect on migration of these osteoclast

precursors without being toxic to those cells.

Previous data have shown that long bone osteoclasts differentiate faster than jaw

osteoclasts in vitro, possibly due to higher numbers of committed osteoclast precursors, i.e.

myeloid blasts, in long bone marrow [17]. In the present study, we confirmed that

osteoclast marker expression was higher in the long bones than in the jaw during early

differentiation. Surprisingly, this was not represented by a higher fusion rate in the long

bone marrow cells. Perhaps, lower osteoclast marker expression was the result of the

abundance of macrophages in jaw cultures, as shown by high expression of the macrophage

markers MafB and F4/80. This would lead to a relatively lower expression of osteoclast

markers in the jaw cultures, not due to the presence of fewer osteoclasts, but due to the

Chapter 6

118

presence of other cell types in the culture. Another explanation for the discrepancy between

osteoclast marker expression and fusion rates could be attributed to the different culture

substrates. Gene expression data are based on cells growing on plastic whereas fusion was

studied on carbon-coated glass chamber slides.

Interestingly, jaw bone marrow cells expressed more CXCR4 than long bone cells

after 3 days of culturing with M-CSF and RANKL. Since both long bone and jaw bone

marrow cells migrated towards CXCL12, CXCR4 was likely present on those cells as well.

Yet, there was no difference in the migration towards CXCL12 between long bone and jaw

bone marrow cells after 1 day of culture with M-CSF and RANKL. Likely, this was the result

of the higher expression of the decoy receptor for CXCL12, CXCR7, leading to an opposite,

chemorepulsive effect [25,26].

Next to CXCR4, jaw bone marrow cells also expressed more CXCL12 than long

bone marrow cells. This might indicate that jaw bone marrow may have a stronger

chemoattracting potential than long bone marrow cells. Therefore, it would be of interest to

investigate whether jaw bone marrow has a stronger chemoattracting potential towards

other bone marrow cells or hematopoietic precursors from the blood. Also, this may indicate

that the egress of hematopoietic precursors from the bone marrow to the blood [7] may

occur preferably from the long bones, and further research is required to investigate

whether the jaw and long bone marrows play a different role in terms of the hematopoietic

stem and progenitor cell pool.

Wright et al. (2005) showed that CXCL12 can, besides attracting osteoclast

precursors, inhibit the pro-apoptotic protein Bim, and increase the anti-apoptotic/pro-

apoptotic mRNA ratio, thereby stimulating osteoclast survival [4]. The high CXCL12

expression in the jaw cells that we found in this study, may therefore explain the higher

expression of the anti-apoptotic genes Bcl-2 and Bcl-xL that we found in jaw than in long-

bone marrow cells in the previous study [20]. Currently, it is unknown through which

receptor the anti-apoptotic effect of CXCL12 may be mediated. Altogether, these results

may indicate that jaw osteoclast precursors express a higher level of survival factors than

long bone osteoclast precursors, and therefore, they may be less sensitive to BP-induced

apoptosis.

In conclusion, we showed that long bone and jaw bone marrow osteoclast

precursors do not differ in proliferation, migration, and fusion rates. However, jaw bone

marrow cells express more CXCL12 than long-bone marrow cells, possibly explaining why

jaw bone marrow cells had a higher anti-apoptotic gene expression than long-bone marrow

cells. Our data provide additional evidence that the osteoclast precursors from long bone

and jaw are different.


119

Acknowledgments

The authors would like to thank Jolanda Hogervorst for technical assistance, Ineke D.C.

Jansen, MSc for technical support with time-lapse experiments and Ana Paula de Souza

Faloni, DDS, MSc, PhD for preparing samples.

Chapter 6

120

References

1. Udagawa N, Takahashi N, Akatsu T, Tanaka H, Sasaki T, Nishihara T, et al. Origin of

osteoclasts: mature monocytes and macrophages are capable of differentiating into

osteoclasts under a suitable microenvironment prepared by bone marrow-derived stromal

cells. Proc Natl Acad Sci U S A. 1990;87(18):7260-4.

2. Ishida N, Hayashi K, Hattori A, Yogo K, Kimura T, Takeya T. CCR1 acts downstream of NFAT2

in osteoclastogenesis and enhances cell migration. J Bone Miner Res. 2006;21(1):48-57.

3. Koizumi K, Saitoh Y, Minami T, Takeno N, Tsuneyama K, Miyahara T, et al. Role of

CX3CL1/fractalkine in osteoclast differentiation and bone resorption. J Immunol.

2009;183(12):7825-31.

4. Wright LM, Maloney W, Yu X, Kindle L, Collin-Osdoby P, Osdoby P. Stromal cell-derived

factor-1 binding to its chemokine receptor CXCR4 on precursor cells promotes the

chemotactic recruitment, development and survival of human osteoclasts. Bone.

2005;36(5):840-53.

5. Yu X, Huang Y, Collin-Osdoby P, Osdoby P. Stromal cell-derived factor-1 (SDF-1) recruits

osteoclast precursors by inducing chemotaxis, matrix metalloproteinase-9 (MMP-9) activity,

and collagen transmigration. J Bone Miner Res. 2003;18(8):1404-18.

6. Sugiyama T, Kohara H, Noda M, Nagasawa T. Maintenance of the hematopoietic stem cell

pool by CXCL12-CXCR4 chemokine signaling in bone marrow stromal cell niches. Immunity.

2006;25(6):977-88.

7. Golan K, Kollet O, Lapidot T. Dynamic Cross Talk between S1P and CXCL12 Regulates

Hematopoietic Stem Cells Migration, Development and Bone Remodeling. Pharmaceuticals

(Basel, Switzerland). 2013;6(9):1145-69.

8. Ishii M, Egen JG, Klauschen F, Meier-Schellersheim M, Saeki Y, Vacher J, et al. Sphingosine-1-

phosphate mobilizes osteoclast precursors and regulates bone homeostasis. Nature.

2009;458(7237):524-8.

9. Puchert M, Engele J. The peculiarities of the SDF-1/CXCL12 system: in some cells, CXCR4 and

CXCR7 sing solos, in others, they sing duets. Cell and tissue research. 2014;355(2):239-53.

10. Sun X, Cheng G, Hao M, Zheng J, Zhou X, Zhang J, et al. CXCL12 / CXCR4 / CXCR7

chemokine axis and cancer progression. Cancer metastasis reviews. 2010;29(4):709-22.

11. Russell RG. Bisphosphonates: the first 40 years. Bone. 2011;49(1):2-19.






Commun. 1999;264(1):108-11.

14. Itzstein C, Coxon FP, Rogers MJ. The regulation of osteoclast function and bone resorption by

small GTPases. Small GTPases. 2011;2(3):117-30.

15. Warshafsky B, Aubin JE, Heersche JN. Cytoskeleton rearrangements during calcitonin-induced

changes in osteoclast motility in vitro. Bone. 1985;6(3):179-85.


121





2011;88(1):63-74.






20. Vermeer JA, Jansen ID, Marthi M, Coxon FP, McKenna CE, Sun S, et al. Jaw bone marrow-

derived osteoclast precursors internalize more bisphosphonate than long-bone marrow

precursors. Bone. 2013;57(1):242-51.



2000;39(4):295-9.

22. Yagi M, Miyamoto T, Sawatani Y, Iwamoto K, Hosogane N, Fujita N, et al. DC-STAMP is

essential for cell-cell fusion in osteoclasts and foreign body giant cells. The Journal of

experimental medicine. 2005;202(3):345-51.

23. Jansen ID, Vermeer JA, Bloemen V, Stap J, Everts V. Osteoclast fusion and fission. Calcif

Tissue Int. 2012;90(6):515-22.

24. Bloemen V. Cell-cell interactions during osteoclastogenesis [PhD Thesis]. Amsterdam: VU

University Amsterdam; 2010.

25. Balabanian K, Lagane B, Infantino S, Chow KY, Harriague J, Moepps B, et al. The chemokine

SDF-1/CXCL12 binds to and signals through the orphan receptor RDC1 in T lymphocytes. The

Journal of biological chemistry. 2005;280(42):35760-6.

26. Naumann U, Cameroni E, Pruenster M, Mahabaleshwar H, Raz E, Zerwes HG, et al. CXCR7

functions as a scavenger for CXCL12 and CXCL11. PloS one. 2010;5(2):e9175.

Chapter 7.

General discussion

Chapter 7

124

Bone resorption by osteoclasts is often seen as the initial step of bone remodeling. Together

with the other bone cells, i.e. osteoblasts, osteocytes, and bone-lining cells, osteoclasts are

essential to maintain bone and mineral homeostasis. The last decades, it has become clear

that osteoclasts are bone-site specific, a phenomenon termed osteoclast heterogeneity

[1,2]. When osteoclastic bone resorption outbalances bone formation, bone is lost and the

skeleton becomes more fragile. In degenerative bone diseases, such as osteoporosis and

bone cancer, bisphosphonates (BPs) are widely used as a treatment to counteract excessive

bone resorption by causing osteoclast apoptosis and inhibiting resorptive activity.

BP use is associated with a rare, though severe side effect in the jaw bone:

osteonecrosis of the jaw (ONJ), which is defined as exposed necrotic bone that does not

heal for at least 8 weeks [3,4]. Depending on the dose, prevalence of BP-related ONJ is

estimated between 0.1 and 10% for people that receive high doses for cancer treatment

[5,6]. Risk factors for this condition are dental extractions and other oral trauma, however it

is not clear why specifically the jaw is negatively affected by BPs. With the increasing life

expectancy these drugs will be subscribed more often, and the side effects will become an

emerging problem. Gaining more insight into the pathogenesis of ONJ is therefore highly

relevant for researchers in the fields of bone and dentistry, the clinicians subscribing BPs,

the patients using BPs, and the dentists and maxillofacial surgeons treating those patients.

In this thesis, we investigated the relation between two bone-site-specific

phenomena: osteonecrosis of the jaw and osteoclast heterogeneity. We hypothesized that

bisphosphonates may have a different effect on osteoclasts and their precursors in the long

bones and in the jaw. We used in vitro and in vivo approaches to investigate this hypothesis

and revealed novel findings that will be discussed in this chapter. The major findings are

summarized in Table 1 (the effect of BPs) and Table 2 (gene expression in long bone and

jaw osteoclast precursors).

First, we used a fluorescently labeled BP and investigated its internalization by long

bone and jaw osteoclast precursors in vitro (chapter 2). Jaw osteoclast precursors

internalized more BPs than long bone osteoclast precursors, both after 4 and 24 hours. After

24 hours, this was accompanied by the accumulation of more unprenylated Rap1a in jaw

cells. Surprisingly, this did not result in a different effect on osteoclastogenesis. These

findings indicate that jaw osteoclast precursors might be less sensitive to BPs than long

bone osteoclast precursors after BP uptake. A possible explanation for this might be the

lower activity of the apoptosis marker caspase 3/7 in jaw osteoclast precursors than in long

bone osteoclast precursors after BPs. Therefore, we propose that jaw osteoclast precursors

are more resistant to BP-induced apoptosis than those from long bone.

Table 1.

findings in this thesis. + indicate a relative stimulation (or levels of internalization/accumulation),

indicate an inhibition, and +

osteoclast precursors. Jaw osteoclast precursors expressed a higher level of anti

genes B cell lymphoma 2 (Bcl

precursors. Ye

expressed in both cell types, leading to a higher anti

expression ratio in jaw cells

increa

expressed in the bone marrow

precursors expressed more CXCL12, as well as its receptor CXCR4, a similar mechanism m

apply to our findings. The exact mechanism behind higher anti

and lower caspase 3/7 activity in jaw bone marrow cells, and a possible resistance to BPs,

requires further research. Besides exploring the response of jaw and long

to CXCL12 in more detail, the effect of two key regulators of osteoclast formation could be

investigated. Since macrophage

nuclear factor κB ligand (RANKL) were shown to induce B

they are interesting anti

Table 1.





precursors. Ye



increa










Table 1.





precursors. Ye



increased by CXCL12










The effect of bisphosphon



Several mechanisms



precursors. Ye



sed by CXCL12













Several mechanisms



precursors. Yet, the pro



sed by CXCL12













Several mechanisms



t, the pro



sed by CXCL12













Several mechanisms



t, the pro



sed by CXCL12 [7]












indicate an inhibition, and +-

Several mechanisms



t, the pro



[7], a chemoattractant for osteoclast precursors and other cells, highly









they are interesting anti-apoptotic candidates in this regard.



- means that there was no effect.

Several mechanisms



t, the pro-apoptotic gene Bcl



, a chemoattractant for osteoclast precursors and other cells, highly









apoptotic candidates in this regard.

The effect of bisphosphonates (BPs) on long bone and jaw osteoclasts. Overview of the main


means that there was no effect.

Several mechanisms may explain lower apoptotic activity in jaw than in long bone


genes B cell lymphoma 2 (Bcl-

apoptotic gene Bcl


expression ratio in jaw cells (chapter 2








investigated. Since macrophage-



ates (BPs) on long bone and jaw osteoclasts. Overview of the main



may explain lower apoptotic activity in jaw than in long bone


-2) and Bcl

apoptotic gene Bcl


chapter 2


expressed in the bone marrow [8]






-colony stimulating factor (M








2) and Bcl

apoptotic gene Bcl


chapter 2


[8]. Since we showed in






colony stimulating factor (M








2) and Bcl

apoptotic gene Bcl


chapter 2, Table 2). This ratio was previously shown to be


. Since we showed in














2) and Bcl-2

apoptotic gene Bcl


, Table 2). This ratio was previously shown to be
















2-like protein 1 (Bcl

apoptotic gene Bcl-2-


















like protein 1 (Bcl

-associated X protein (Bax) was similarly


















like protein 1 (Bcl

associated X protein (Bax) was similarly

















like protein 1 (Bcl


expressed in both cell types, leading to a higher anti-apoptotic to pro
















like protein 1 (Bcl


apoptotic to pro



. Since we showed in chapter 6






colony stimulating factor (M-CSF) and receptor activator of

nuclear factor κB ligand (RANKL) were shown to induce Bcl-






like protein 1 (Bcl


apoptotic to pro



chapter 6


apply to our findings. The exact mechanism behind higher anti-apoptotic gene expression




CSF) and receptor activator of

-2 and Bcl





like protein 1 (Bcl-xL) than long osteoclast


apoptotic to pro



chapter 6


apoptotic gene expression





2 and Bcl





xL) than long osteoclast


apoptotic to pro



chapter 6







2 and Bcl

General







apoptotic to pro



that jaw osteoclast




requires further research. Besides exploring the response of jaw and long-bone marrow cells



2 and Bcl-xL expression

General







apoptotic to pro-apoptotic gene



that jaw osteoclast




bone marrow cells



xL expression

General d




osteoclast precursors. Jaw osteoclast precursors expressed a higher level of anti-



apoptotic gene



that jaw osteoclast




bone marrow cells



xL expression

discussion




-apoptotic



apoptotic gene



that jaw osteoclast




bone marrow cells



xL expression

iscussion

125




apoptotic



apoptotic gene



that jaw osteoclast




bone marrow cells



xL expression [9]

iscussion

125


findings in this thesis. + indicate a relative stimulation (or levels of internalization/accumulation), -


apoptotic



apoptotic gene



that jaw osteoclast

precursors expressed more CXCL12, as well as its receptor CXCR4, a similar mechanism may



bone marrow cells



[9],



apoptotic



apoptotic gene



that jaw osteoclast

ay



bone marrow cells



,

Chapter 7

126

Table 2.

Overview of the main findings in this thesis. Asterisks indi

within the bone marrow internalized more BPs. Long bone and jaw bone marrow cells were

shown to have a different composition and freshly isolated

more osteoclast precursors than jaw bone marrow

long bone and jaw osteoclast precursors at day 3 of culture is different

results in a different cellular composition

than osteoclast precursors may be another explanation why despite a different uptake,

osteoclastogenesis from jaw and long

chapter 2

macrophage markers F4/80, MafB, and Irf8 in long bone and jaw bone marrow cultures at

the end of the culture period of 6 days with M

differentiation, after 3 days of

studied, jaw bone marrow cultures expressed more MafB and F4/80 than long

(chapter 6)

than long bone osteoclast cultu

by macrophages might explain why jaw bone marrow cultures internalize more BP than

those from long bone, and also explains why the actual precursors were similarly affected

by BPs. Yet, in

populations internalized more BPs in jaw cultures. Since macrophages likely belong to the

small cell population, strong BP uptake by macrophages cannot fully explain higher BP

uptake by jaw oste

dextran

Chapter 7

126

Table 2.









chapter 2





(chapter 6)




by BPs. Yet, in



uptake by jaw oste

dextran

Chapter 7

Table 2.









chapter 2





(chapter 6)




by BPs. Yet, in



uptake by jaw oste

dextran [11]

Chapter 7

Gene expression in long bone and jaw osteoclast precursors (unless mentioned otherwise).


Another question remaining after our internalization studies is which cell type








chapter 2, we did not detect differen





(chapter 6)




by BPs. Yet, in



uptake by jaw oste

[11]











, we did not detect differen





(chapter 6), which could imply that jaw cultures have a more macrophage




by BPs. Yet, in



uptake by jaw oste

[11], was similar in long bone and jaw cells indicating that they have similar rates of
















, which could imply that jaw cultures have a more macrophage




by BPs. Yet, in chapte



uptake by jaw oste

, was similar in long bone and jaw cells indicating that they have similar rates of




















chapte



uptake by jaw osteoclast cultures. Moreover, uptake of a fluid





















chapter 2



oclast cultures. Moreover, uptake of a fluid





















r 2, we distinguished between small and large cells, and both





















than long bone osteoclast cultures. Since macrophages are highly endocytic cells, BP uptake



, we distinguished between small and large cells, and both


















differentiation, after 3 days of culture with M



res. Since macrophages are highly endocytic cells, BP uptake





















culture with M




















osteoclastogenesis from jaw and long-bone marrow was similarly affected by BPs. In

, we did not detect different levels of the osteoblast marker ALP, and the



culture with M


















results in a different cellular composition of the osteoclast cultures. Uptake by cells other


bone marrow was similarly affected by BPs. In

t levels of the osteoblast marker ALP, and the



culture with M


















of the osteoclast cultures. Uptake by cells other





the end of the culture period of 6 days with M-CSF and RANKL. Yet, in an earlier stage of

culture with M-CSF and RANKL and when BP uptake was












Overview of the main findings in this thesis. Asterisks indicate anti




more osteoclast precursors than jaw bone marrow [10]







CSF and RANKL. Yet, in an earlier stage of

CSF and RANKL and when BP uptake was












cate anti




[10]. Also, the differentiation stage of the




















cate anti-




. Also, the differentiation stage of the




















-apoptotic genes.
























apoptotic genes.



shown to have a different composition and freshly isolated long


















oclast cultures. Moreover, uptake of a fluid-phase endocytosis marker,



apoptotic genes.



long-bone marrow contained


















phase endocytosis marker,



apoptotic genes.



bone marrow contained





















apoptotic genes.





long bone and jaw osteoclast precursors at day 3 of culture is different [10]























[10], which likely








studied, jaw bone marrow cultures expressed more MafB and F4/80 than long-bone marrow

, which could imply that jaw cultures have a more macrophage-like phenotype














, which likely








bone marrow

like phenotype














, which likely








bone marrow

like phenotype














, which likely








bone marrow

like phenotype














, which likely








bone marrow

like phenotype














, which likely








bone marrow

like phenotype









General discussion

127

fluid phase endocytosis. Therefore, our findings suggest that long bone and jaw osteoclast

precursors have distinct mechanisms of BP uptake specifically and therefore, these data

provide more evidence that osteoclast precursors from long bone and jaw are different.

A third explanation why jaw osteoclasts might be less sensitive to BPs, may be a

compensation mechanism during the process of osteoclastogenesis. Osteoclast formation

consists of several steps, starting with the attraction and migration of osteoclast precursors

towards the target site. There the precursors interact with bone-lining cells and other bone-

associated cells, and they fuse to become multinucleated osteoclasts. In chapter 5, we

used time-lapse microscopy and proved it to be a useful tool to study osteoclastogenesis.

We showed that next to fusion, osteoclasts can undergo fission and may be able to get rid

of apoptotic nuclei. Possibly, using this mechanism, osteoclasts can also get rid of their

cellular compartments that became apoptotic due to BP treatment and therefore

compensate for the induced apoptotic pathways.

Using time-lapse microscopy, we also studied other steps during osteoclastogenesis

and investigated whether long bone and jaw osteoclast precursors differ in their rates of

proliferation, migration, and fusion; three important steps during osteoclastogenesis

(chapter 6). All rates were similar for long bone and jaw osteoclast precursors and

surprisingly, in neither of the populations, the migration velocity was affected by the BP

pamidronate. The used nitrogen-containing BP inhibits the prenylation, and thereby the

cellular location and function of small GTPases. Those proteins are important for cytoskeletal

rearrangements, and therefore play a role in cell adhesion and migration. In our system

using time-lapse microscopy and a non-toxic BP concentration (10 µM), we did not find an

effect on migration velocity. A drawback of this system is the substrate that was used.

Carbon-coated chamber slides that were used for time-lapse microscopy were shown to

improve fibroblast cell adhesion [12]. Adhesion to this substrate by osteoclast precursors

and culture conditions for those cells, may not have been optimal. Although

osteoclastogenesis did take place, the number of fusion events varied between experiments.

Therefore, it would be worthwhile to optimize these osteoclastogenesis cultures on carbon-

coated chamber slides. Another interesting improvement, resembling the in vivo situation

more closely, would be to coat the slides with hydroxyapatite and image the cells on this

substrate. To circumvent culturing on chamber slides, the effect of BPs on directional

migration towards CXCL12 could also be investigated in the transwell culture system that we

used. Yet, as far as we know, we are the first to study the effect of nitrogen-containing BPs

on migration velocity of primary bone marrow cells. We did not find a BP concentration that

inhibited migration without being toxic to the cells. This may indicate that BPs do not affect

osteoclast precursors migration, and do therefore not interfere with osteoclastogenesis at

this differentiation stage.

Chapter 7

128

Altogether, our results showed that jaw osteoclasts or precursors may be less

sensitive to BPs than long bone osteoclasts or their precursors. Most likely, this was caused

by higher anti-apoptotic gene expression and lower caspase 3 activity in jaw than in long

bone osteoclast precursors (chapter 2), which might have been the result of higher

CXCL12 expression in the jaw than in long bone osteoclast cultures (chapter 6).

In chapter 3, we investigated whether BPs differently affected osteoclasts in the

long bones and jaw in vivo. Zoledronic acid (ZA) was administered to 3-month-old female

C57BL/6J mice for up to 6 months. At baseline, and after 1, 3, and 6 months of treatment,

mice were sacrificed and bone marrow cells were isolated from the mandible, femur, and

tibia. Interestingly, after 6 months, significantly fewer cells were isolated from the jaw from

BP-treated mice than from controls, whereas the number of long-bone marrow cells was

unaffected by BP treatment. This might have been the result of an increased jaw bone

volume to over 90% after BPs, which was highly likely accompanied by a decreased bone

marrow space. Another explanation could be that BPs are more toxic to the bone marrow

cells in the jaw than in the long bones. Although we showed in chapter 2 that jaw bone

marrow cells might be less sensitive to BPs than long-bone marrow cells, they may be

exposed to higher BP concentrations in vivo. Jaw and long bones from rats were shown to

contain a similar amount of BPs per unit of dry weight or calcium [13,14]. Since the jaw

contains more mineral per bone volume ([14], chapter 3), it likely adsorbed more BPs than

long bones. Together with the finding that jaw osteoclasts, in vitro, resorb at least as much

bone or dentin as long bone osteoclasts ([10], chapter 2), more BP is likely released from

the jaw than from long bone, making it available for the surrounding cells in the bone

marrow [15,16]. The resulting high local BP concentration might cause apoptosis of bone

marrow cells, leading to a reduced number of jaw bone marrow cells.

Which subset of bone marrow cells internalized BPs and which subset was affected

by 6 months of ZA requires further investigation. This could be done by staining of specific

markers for different bone marrow cells in decalcified sections. Monocytes are promising

candidates in this regard, as they were shown to internalize BPs in vivo [17]. Furthermore, a

role for diminished macrophage activity after BP treatment was previously proposed to play

a role in ONJ [18]. This hypothesis, together with our findings, may provide more insight

into the pathogenesis of ONJ, and suggest a combination of hypotheses to play a role in

ONJ (chapter 1), as illustrated in Figure 1. We suggest that BP release and the resulting

high local BP concentration due to low sensitivity of jaw osteoclasts to BPs, play a more

central role in the pathogenesis of ONJ. This hypothesis provides a strong connection

between the previously suggested hypotheses of microbial infection and the toxicity of BPs

to cells other than osteoclasts (Figure 1).

General discussion

129

Another hypothesis regarding the etiology of ONJ is the oversuppression of bone

turnover. A reduced bone remodeling capacity and reduced ability to repair bone damage

can lead to the accumulation of microdamage [19,20]. Our results however, do not support

a long-term, reducing effect on bone turnover markers in the jaw, whereas bone turnover

was inhibited in the long bones (chapter 3). Also, the bone surface covered with

osteoclasts and osteoclast numbers increased with age in mice only in long bone, to levels

that were 2 to almost 4 times higher in the long bones than in the jaw. Therefore, our

results suggest that suppression of bone turnover and the resulting microdamage would

more likely cause side effects in the long bones than in the jaw. Indeed, BPs were recently

associated with atypical femur fractures that are probably caused by the inability to heal

microcracks [21].

Figure 1. New hypothesis for BP-related ONJ (squares), where the BP release (purple square) due to

the mild effect of BPs on jaw osteoclasts specifically plays a central role. A possible toxicity to immune

cells and a reduced ability to clear microbial infections provides a link between the 2 previously

proposed hypotheses (black boxes). ONJ (in the blue square) consists of 3 events shown in the squares

within the blue box. Green arrows represent positive effects, red blunted arrows show inhibitions. The

blue arrows point out a result of the inhibitions indicated in red. The dashed arrows with question

marks are speculative, since positive effects of BPs on osteoblasts and osteocytes have also been

shown.

Chapter 7

130

Dental trauma such as a tooth extraction was reported to be a risk factor for ONJ,

however, a causal relationship has never been shown [22]. In our model, we did not

perform a tooth extraction and the mice did not develop any signs of ONJ. However, we did

find BP-induced osteoclast activity at the molar roots, despite the presence of

bisphosphonates. An explanation for this may be that, in contrast to BP adsorbed to the

alveolar bone, little BP seems to accumulate at the root cementum and therefore osteoclasts

formed at this site were hardly or not exposed to BPs [23]. The tooth loss that might result

from osteoclast activity at the root, may be related to the reported relation between tooth

extraction and ONJ. It would be of interest to investigate a possible causative role for root

resorption and tooth loss in the pathogenesis of ONJ. Therefore, it requires further

investigation whether BP use, for instance by applying a longer administration or a higher

dose, can result in tooth loss caused by root resorption, and whether this could eventually

lead to osteonecrosis of the jaw. Altogether, we reported a stimulating, rather than an

inhibiting effect of BPs on osteoclasts at the molar root, thereby revealing another bone-

site-specific effect of BPs.

The osteoclasts that were formed at the molar roots of BP-treated mice were

located in the periodontal ligament (PDL). Previously, PDL fibroblasts were shown to induce

osteoclast formation in a co-culture with osteoclast precursors from peripheral blood

mononuclear cells (PBMCs) [24,25]. Also, PDL fibroblast viability was reduced by BPs in vitro

[26,27] and in our in vivo study, the periodontal area was only mildly affected by BPs

(chapter 3). Therefore, we investigated whether PDL fibroblasts may have played a role in

the observed osteoclast formation at the molar root after BP treatment (chapter 4). Molar-

root-associated human PDL fibroblasts were allowed to attach to 48-well plates and

subsequently pre-treated with pamidronate (1-100 µM) or vehicle for 24 hours. After this

pre-incubation, the number of PDL fibroblasts and their viability seemed to be unaffected.

BPs were washed away with PBS, and freshly isolated PBMCs were added. After 21 days,

the number of osteoclasts was counted. Although directly after a 24-hour treatment the PDL

fibroblasts were not affected, after long-term culturing, pre-treatment with 100 µM

pamidronate induced PDL fibroblast death. This resulted in a completely abolished

osteoclastogenesis in co-culture with PBMCs. Lower concentrations did not affect PDL

fibroblasts or their effect on osteoclastogenesis. These data may indicate that non-toxic BP

concentrations do not affect osteoclastogenesis through PDL fibroblasts in vitro.

In vivo, BP administration increases bone volume and therefore likely increases

compressive loading on the PDL (chapter 3). Previously, PDL fibroblasts under

compressive loading were shown to induce osteoclastogenesis [28] and a similar mechanism

might play a role in BP-induced root resorption. Studying the effect of BPs on mechanically

General discussion

131

loaded PDL fibroblasts, might therefore give us more insight into the mechanism behind BP-

induced root resorption. It also requires further investigation whether in vivo, PDL

fibroblasts are exposed to BP concentrations that may affect PDL-mediated

osteoclastogenesis.

Conclusion

Taken together, the data presented in this thesis suggest that bisphosphonates similarly

affect long bone and jaw osteoclasts in vitro and in vivo. Interestingly, this similarity was

likely the result of different mechanisms of action (summarized in Table 1). More BPs

accumulated in jaw osteoclasts and their precursors than in those from long bone. Yet, jaw

osteoclast precursors likely compensate for the higher internal BP concentration by using

anti-apoptotic pathways. We provided mechanistic insight into this hypothesis by showing

higher expression of CXCL12 in jaw osteoclast precursors, which was previously shown to

inhibit apoptosis in osteoclasts. These data provide further evidence that the osteoclasts and

their precursors are bone-site specific. How these differences might provide more insight

into the pathogenesis of osteonecrosis of the jaw is explained below.

We hypothesize that the more apoptosis-resistant jaw osteoclasts can survive

longer than long bone osteoclasts in the presence of BPs. Therefore, jaw osteoclasts might

absorb more BPs from the bone, thereby making them available for uptake by other

surrounding cells. BP toxicity to monocytes and macrophages, epithelial cells, endothelial

cells, osteocytes, and osteoblasts may all contribute to the pathogenesis of ONJ (Figure 1).

Thus, the previously postulated hypothesis that BPs may be toxic to other cells, may be

explained by the resistance of jaw osteoclasts to BP-induced apoptosis that we reported

here. This resistance also explains the paradoxical finding of bone loss in osteonecrosis of

the jaw, despite the presence of BPs and subsequent inhibition of osteoclasts. Furthermore,

the toxicity to immune cells may reduce the ability to clear infections, thereby supporting

the other hypothesis that an infection may cause ONJ. Therefore, this multifactorial disease

consisting of soft tissue damage, bone necrosis, and bone loss, probably originates from

multifactorial causes (Figure 1). However, our findings contradict the hypothesis that

oversuppression of bone turnover may explain its etiology, since bone turnover was

inhibited in the long bones, rather than in the jaw.

Our data imply that further research is required to fully elucidate the effect of BPs

on osteoclasts and surrounding cells at different bone sites. Uptake by those cells in vivo

should be assessed to gain more insight into our hypothesis that a high local BP

concentration in the jaw microenvironment plays a more central role in the pathogenesis of

ONJ. Also, the effect of RANKL inhibitors on bone-site-specific cells should be investigated,

as denosumab was also shown to increase the risk for ONJ. Moreover, the finding that jaw

Chapter 7

132

bone marrow cells may contain more CXCL12 is intriguing, and suggests that more research

could focus on the role of bone-site-specific bone marrow cells and osteoclasts in terms of

the hematopoietic stem and progenitor cell pool.

This thesis provides additional evidence for bone-site specificity of osteoclasts and

their precursors in the bone marrow. Therefore, these studies may contribute to the

development of more specific anti-resorptive drugs. To achieve this, more research is

necessary to gain more insight into the differences between bone-site-specific osteoclasts

and other bone cells.

General discussion

133

References





2011;32(1):31-63.




4. Reid IR, Cornish J. Epidemiology and pathogenesis of osteonecrosis of the jaw. Nat Rev

Rheumatol. 2012;8(2):90-6.

5. Abu-Id MH, Warnke PH, Gottschalk J, Springer I, Wiltfang J, Acil Y, et al. "Bis-phossy jaws" -

high and low risk factors for bisphosphonate-induced osteonecrosis of the jaw. J

Craniomaxillofac Surg. 2008;36(2):95-103.

6. Reid IR. Osteonecrosis of the jaw: who gets it, and why? Bone. 2009;44(1):4-10.

7. Wright LM, Maloney W, Yu X, Kindle L, Collin-Osdoby P, Osdoby P. Stromal cell-derived

factor-1 binding to its chemokine receptor CXCR4 on precursor cells promotes the

chemotactic recruitment, development and survival of human osteoclasts. Bone.

2005;36(5):840-53.

8. Yu X, Huang Y, Collin-Osdoby P, Osdoby P. Stromal cell-derived factor-1 (SDF-1) recruits

osteoclast precursors by inducing chemotaxis, matrix metalloproteinase-9 (MMP-9) activity,

and collagen transmigration. J Bone Miner Res. 2003;18(8):1404-18.

9. Lacey DL, Tan HL, Lu J, Kaufman S, Van G, Qiu W, et al. Osteoprotegerin ligand modulates

murine osteoclast survival in vitro and in vivo. Am J Pathol. 2000;157(2):435-48.



2011;88(1):63-74.

11. Thompson K, Rogers MJ, Coxon FP, Crockett JC. Cytosolic entry of bisphosphonate drugs

requires acidification of vesicles after fluid-phase endocytosis. Mol Pharmacol.

2006;69(5):1624-32.



2000;39(4):295-9.

13. Bauss F, Pfister T, Papapoulos S. Ibandronate uptake in the jaw is similar to long bones and

vertebrae in the rat. Journal of bone and mineral metabolism. 2008;26(4):406-8.

14. Wen D, Qing L, Harrison G, Golub E, Akintoye SO. Anatomic site variability in rat skeletal

uptake and desorption of fluorescently labeled bisphosphonate. Oral diseases.

2011;17(4):427-32.



Chapter 7

134







2011;103(3):232-40.

19. Hoefert S, Schmitz I, Tannapfel A, Eufinger H. Importance of microcracks in etiology of

bisphosphonate-related osteonecrosis of the jaw: a possible pathogenetic model of

symptomatic and non-symptomatic osteonecrosis of the jaw based on scanning electron

microscopy findings. Clinical oral investigations. 2010;14(3):271-84.

20. Mashiba T, Hirano T, Turner CH, Forwood MR, Johnston CC, Burr DB. Suppressed bone

turnover by bisphosphonates increases microdamage accumulation and reduces some

biomechanical properties in dog rib. J Bone Miner Res. 2000;15(4):613-20.

21. Shane E, Burr D, Abrahamsen B, Adler RA, Brown TD, Cheung AM, et al. Atypical

subtrochanteric and diaphyseal femoral fractures: second report of a task force of the

american society for bone and mineral research. J Bone Miner Res. 2014;29(1):1-23.



24.



24. Bloemen V, Schoenmaker T, de Vries TJ, Everts V. Direct cell-cell contact between periodontal

ligament fibroblasts and osteoclast precursors synergistically increases the expression of

genes related to osteoclastogenesis. Journal of cellular physiology. 2010;222(3):565-73.



research. 2001;80(3):887-91.






28. Kanzaki H, Chiba M, Shimizu Y, Mitani H. Periodontal ligament cells under mechanical stress

induce osteoclastogenesis by receptor activator of nuclear factor kappaB ligand up-regulation

via prostaglandin E2 synthesis. J Bone Miner Res. 2002;17(2):210-20.

General summary

136

Bone remodeling is a life-long process, and important to maintain a well-functioning

skeleton and mineral homeostasis. Multinucleated osteoclasts degrade bone, whereas

osteoblasts deposit new bone. Osteocytes experience and respond to mechanical loading

and regulate osteoblast and osteoclast activity. When osteoclasts are overactive, for

instance in diseases such as osteoporosis and bone cancer, bisphosphonates (BPs) can be

used to inhibit osteoclast activity. With the increasing life-expectancy, those bone diseases

are expected to become even more common. Therefore, possible side effects of anti-

resorptive drugs will also become an emerging problem. Osteonecrosis of the jaw is such a

rare, though severe side effect of BPs, and it is not clear how and why specifically the jaw is

affected. We hypothesized that osteonecrosis of the jaw is caused by a different response of

bone-site-specific osteoclasts and their precursors to BPs. We used several approaches to

investigate whether BPs might have a different effect on long bone and jaw osteoclasts and

their precursors. With these studies, which are summarized below, we aimed to get more

insight into (i) differences between bone-site-specific osteoclasts and precursors, and (ii) the

pathogenesis of BP-related osteonecrosis of the jaw.

In chapter 2, we investigated the internalization of BPs by long bone and jaw

osteoclast precursors and studied the effect of BPs on osteoclastogenesis and apoptosis of

long bone and jaw bone marrow cells isolated from healthy mice. Jaw osteoclast precursors

internalized more BPs than long bone osteoclast precursors. This was accompanied by an

accumulation of more unprenylated Rap1a in jaw osteoclast precursors. Intriguingly, a

higher intracellular BP concentration did not differently affect osteoclastogenesis of long

bone and jaw bone marrow cells. This could be explained by a higher anti-apoptotic gene

expression and lower caspase 3/7 activity in jaw than in long bone osteoclast cultures.

Therefore, these data show that jaw osteoclasts or precursors may be less sensitive to

bisphosphonates than those from long bone.

In chapter 3, the effect of bisphosphonates on long bone and jaw osteoclasts and

bone remodeling was assessed in vivo. Female C57BL/6J mice were subjected to weekly

injections of zoledronic acid for 1, 3, or 6 months. BPs did not significantly affect the

number of osteoclasts in both long bone and jaw, thereby confirming the results obtained in

vitro that BPs did not have a different effect on those different osteoclasts. Also, bone

volume fraction and tissue mineral density were similarly increased by BPs in long bone and

jaw. Yet, the number of bone marrow cells isolated from the jaw was 4 times lower in the

BP-treated animals than in controls. The number of long-bone marrow cells, however, was

not affected. These results indicate that BPs may affect osteoclast precursors and/or other

bone marrow cells in the jaw specifically. On the other hand, bone formation was inhibited

in the long bones on the long term, whereas bone formation in the jaw was not affected by

six months of BPs. Furthermore, this study revealed another bone-site-specific effect of BPs;

General summary

137

the drugs were able to induce osteoclast formation at the molar root. These cells also

proved to be actively resorbing. Therefore, in chapter 4, we used human cells to

investigate whether this induction and/or stimulation may be induced by periodontal

ligament (PDL) fibroblasts that were treated with BPs. Pre-treatment of PDL fibroblasts with

a high concentration of pamidronate was toxic, not only to the fibroblasts, but also to the

PBMCs that were co-cultured with the pre-treated fibroblasts. BP concentrations that were

not toxic to PDL fibroblasts did not affect osteoclastogenesis in co-culture with PBMCs.

In order to study earlier steps during osteoclast formation, time-lapse microscopy

was used to visualize the fusion of long bone osteoclast precursors and multinucleated

osteoclasts (chapter 5). Fusion was seen (i) between mononuclear cells, (ii) between

multinucleated cells, and (iii) between a multinucleated and a mononuclear cell.

Interestingly, cells were also shown to undergo fission, a process that be mediated by small

mononuclear cells present in the bone marrow. The cellular compartments that arise as a

result of fission may contain apoptotic nuclei. Therefore, we hypothesize that osteoclasts

can use the unique process of fission to get rid of apoptotic nuclei.

In chapter 6, we used time-lapse microscopy to study early steps of

osteoclastogenesis, i.e. proliferation, migration, and fusion, of long bone and jaw bone

marrow cells. We also analyzed the expression of genes involved in these processes and the

effect of BPs on osteoclast precursor migration. Long bone and jaw osteoclast precursors

proliferated, migrated, and fused with similar rates. BPs did not affect the migration of both

sets of precursors. Interestingly, jaw osteoclast precursors expressed more CXCL12, CXCR4,

and CXCR7, than long bone osteoclast precursors. Those genes are involved in directional

migration of osteoclast precursors into the bone marrow and CXCL12 was also shown to

inhibit apoptotic pathways in osteoclasts. By showing higher CXCL12 expression in jaw than

in long bone osteoclast precursors, we provide more mechanistic insight into the higher

anti-apoptotic capacity of jaw than of long bone osteoclasts or precursors.

Collectively, we provide more evidence for the existence of bone-site-specific

differences in osteoclasts and their precursors in the bone marrow. The finding that jaw

osteoclasts were less sensitive to bisphosphonates may support our hypothesis that those

bone-site-specific osteoclasts contribute to the development of osteonecrosis of the jaw.

Therefore, we propose a new hypothesis for the pathogenesis of osteonecrosis of the jaw,

which includes two hypotheses previously proposed in literature.

Despite the presence of BPs, the more apoptosis-resistant jaw osteoclasts survive

longer than their long bone counterparts, and therefore they resorb more bone.

Consequently, those jaw osteoclasts are able to release more BPs into the

microenvironment, making them available for uptake by surrounding cells. The resulting

toxicity to other cells contributes to the onset of osteonecrosis of the jaw. The lethality to

138

immune cells, and the inability to tackle infections, followed by a decrease in pH, might lead

to the release of even more BPs, thereby activating a vicious cycle.

This hypothesis proposes that jaw osteoclasts specifically may play a role in the

pathogenesis of osteonecrosis of the jaw, and therefore emphasizes the need for the

development of more specific anti-resorptive drugs. Gaining more insight into the

differences between bone-site-specific osteoclasts is therefore crucial and requires further

research.

Algemene samenvatting

140

Botremodellering is een levenslang durend proces dat belangrijk is voor het behoud van een

goed functionerend skelet en voor mineraalhomeostase. Meerkernige osteoclasten breken

bot af, terwijl osteoblasten het opnieuw opbouwen. Osteocyten ondervinden mechanische

belasting, reageren hierop en reguleren osteoblasten osteoclastactiviteit. Als osteoclasten

te actief zijn, bijvoorbeeld bij ziektes zoals osteoporose en naar bot uitgezaaide kanker,

kunnen bisfosfonaten (BP) gebruikt worden om de activiteit te remmen. Met een

toenemende levensverwachting zullen deze ziektes waarschijnlijk vaker voorkomen en

daarom zullen mogelijke bijwerkingen van anti-resorberende behandelingen een groter

probleem worden. Osteonecrose van de kaak is een zeldzame, maar ernstige bijwerking van

BP en het is niet duidelijk hoe en waarom juist de kaak hierdoor wordt aangedaan. Onze

hypothese was dat osteonecrose van de kaak veroorzaakt wordt door een verschillende

reactie van botspecifieke osteoclasten en hun voorlopers op BP. We hebben verschillende

benaderingen toegepast om te onderzoeken hoe BP pijpbeen- en kaakosteoclasten en hun

voorlopers verschillend kunnen beïnvloeden. Met deze studies, die in dit hoofdstuk zijn

samengevat, trachtten we meer inzicht te krijgen in (i) verschillend tussen botspecifieke

osteoclasten en hun voorlopers en (ii) de pathogenese van bisfosfonaat gerelateerde

osteonecrose van de kaak.

In hoofdstuk 2 hebben we de opname van BP door pijpbeen- en

kaakosteoclastvoorlopercellen onderzocht en het effect op osteoclastogenese en apoptose

bestudeerd in pijpbeen en kaak beenmergcellen geïsoleerd uit gezonde muizen.

Kaakosteoclastvoorlopercellen namen meer BP op dan pijpbeenosteoclastvoorlopers. Dit

ging gepaard met een ophoping van meer ongeprenyleerd Rap1a in osteoclast-

voorlopercellen uit de kaak. Fascinerend was dat een hogere intracellulaire BP concentratie

niet leidde tot een ander effect op osteoclastogenese van pijpbeen- en kaakbeenmergcellen.

Dit zou veroorzaakt kunnen zijn door een hogere expressie van anti-apoptotische genen en

een lagere caspase 3/7 activiteit in de kaak- dan in pijpbeenosteoclastkweken. We laten met

deze data zien dat kaakosteoclasten of hun voorlopercellen wellicht minder gevoelig zijn

voor bisfosfonaten dan deze cellen uit pijpbeenderen.

In hoofdstuk 3 werd het effect van bisfosfonaten op pijpbeen- en

kaakosteoclasten en botremodellering onderzocht in vivo. Vrouwelijke C57BL/6J muizen

kregen gedurende 1, 3 of 6 maanden wekelijks een toediening van zoledroninezuur. Dit

leidde niet tot een significant effect op het aantal osteoclasten, zowel in pijpbeenderen als in

de kaak. Daarmee bevestigen we de resultaten van het in vitro werk dat BP geen ander

effect hebben op osteoclastogenese in de verschillende botten. Bovendien waren het

botvolume en de mineraaldichtheid vergelijkbaar hoger in de pijpbeenderen en in de kaak

na behandeling met bisfosfonaten. Echter, het aantal beenmergcellen dat geïsoleerd kon

worden uit de kaak was 4 keer lager in de BP behandelde groep dan in de controles. Het

Algemene samenvatting

141

aantal beenmergcellen in de pijpbeenderen daarentegen was niet beïnvloed door BP. Deze

resultaten geven aan dat BP de osteoclastvoorlopercellen of andere beenmergcellen alleen

in de kaak aantasten. Botvorming daarentegen was juist geremd op de langere termijn in de

pijpbeenderen, terwijl botvorming in de kaak niet was beïnvloed door een zesmaands

behandeling met BP. Tot slot hebben we nóg een botafhankelijk effect van BP aangetoond;

zoledroninezuur stimuleerde osteoclastvorming aan de wortels van de bestudeerde kies. Ook

resorbeerden deze cellen de wortels. Daarom hebben we in hoofdstuk 4 humane cellen

gebruikt om te onderzoeken of deze inductie of stimulering gemedieerd kan zijn door

parodontaal ligament (PDL) fibroblasten behandeld met BP. Voorbehandeling van PDL

fibroblasten met een hoge concentratie pamidronaat was toxisch, niet alleen voor de

fibroblasten zelf, maar ook voor de PBMCs (die de osteoclastvoorlopercellen) in cokweek

met de voorbehandelde fibroblasten. BP concentraties die niet toxisch waren voor PDL

fibroblasten hadden geen invloed op osteoclastogenese in cokweek met PBMCs.

Om eerdere fases van osteoclastvorming te bestuderen hebben we time-lapse

microscopie gebruikt om de fusie tussen pijpbeenosteoclastvoorlopers en meerkernige

osteoclasten te visualiseren (hoofdstuk 5). We zagen fusie tussen (i) eenkernige cellen,

(ii) meerkernige cellen en (iii) tussen een meerkernige en een eenkernige cel. Uiterst

boeiend was de bevinding dat meerkernige cellen zich konden opsplitsen, een proces dat

gemedieerd zou kunnen zijn door kleine cellen met één kern in het beenmerg. In de

cellulaire compartimenten die ontstonden als gevolg van deze splitsingen zagen we soms

apoptotische kernen. Daarom vermoeden we dat osteoclasten door dit unieke

splitsingsproces apoptotische kernen kunnen afstoten.

In hoofdstuk 6 hebben we time-lapse microscopie gebruikt om vroege processen

gedurende osteoclastogenese (proliferatie, migratie en fusie) van pijpbeen- en

kaakbeenmergcellen te bestuderen. Ook hebben we de expressie geanalyseerd van genen

die betrokken zijn bij deze processen en hebben we het effect onderzocht van BP op

osteoclastvoorloper migratie. Pijpbeen- en kaakosteoclastvoorlopers deelden, migreerden en

fuseerden met gelijke snelheden. BPs hadden geen effect op de migratie van beide

voorlopers. Osteoclastvoorlopercellen geïsoleerd uit de kaak hadden een hogere

genexpressie van CXCL12, CXCR4 en CXCR7 dan deze cellen uit pijpbeenderen. Deze genen

zijn betrokken bij de directionele migratie van cellen richting het beenmerg en het is ook

aangetoond dat CXCL12 betrokken is bij het remmen van apoptotische genexpressie in

osteoclast voorlopers. Doordat wij hier aantonen dat CXCL12 expressie hoger is in kaak- dan

in pijpbeenosteoclastvoorlopers, bieden wij hier meer mechanistisch inzicht in de grotere

anti-apoptotische capaciteit van kaakosteoclasten of hun voorlopers ten opzichte van deze in

de pijpbeenderen.

142

Alles tezamen leveren wij extra bewijs voor het bestaan voor botspecifieke

osteoclasten en hun voorlopercellen in het beenmerg. De bevinding dat kaakosteoclasten

minder gevoelig zijn voor bisfosfonaten onderschrijft onze hypothese dat botspecifieke

osteoclasten een bijdrage kunnen leveren aan het ontstaan van osteonecrose van de kaak.

Mede op basis hiervan poneren wij een nieuwe hypothese voor de pathogenese van

osteonecrose van de kaak, waarbij twee eerder in de literatuur geopperde hypotheses

samen worden gevoegd: ondanks de aanwezigheid van bisfosfonaten overleven de meer

apoptose resistente kaakosteoclasten langer dan de cellen in de pijpbeenderen. Als gevolg

daarvan wordt meer kaakbot geresorbeerd waardoor er meer BP vrijkomt van het bot. Dit

leidt ertoe dat lokaal meer BP beschikbaar zijn voor opname door omliggende cellen. Het

toxisch effect op andere cellen kan een aanleiding zijn voor het ontstaan van osteonecrose

van de kaak. Namelijk, een dodelijk effect op bijvoorbeeld immuuncellen, de verminderde

capaciteit om een infectie te boven te komen en de resulterende verlaging van de pH zou

kunnen leiden tot het vrijkomen van nog meer bisfosfonaten waardoor een vicieuze cirkel

ontstaat.

Deze hypothese impliceert dat specifiek de kaak osteoclasten een rol kunnen

spelen in de pathogenese van osteonecrose van de kaak en benadrukt daarmee de behoefte

aan de ontwikkeling van specifiekere resorptieremmende medicijnen. Hiervoor is het

verrichten van meer onderzoek cruciaal zodat we beter inzicht krijgen in de verschillen

tussen botspecifieke osteoclasten.

Acknowledgments

Dankwoord

Acknowledgments

144

Mijn promotietraject was een fantastische tijd en het is moeilijk om afscheid te nemen van

OCB. Het werken aan mijn promotie heeft eigenlijk nooit als werk aangevoeld. Dat klinkt

misschien ongeloofwaardig, maar ik realiseer me nu dat dat echt zo is en dat ik weinig

anders dan plezier heb ervaren. Natuurlijk waren er ook mindere momenten, maar die zijn

werkelijk waar op één hand te tellen. Daarom wil ik nu iedereen bedanken die heeft

bijgedragen aan deze mooie periode.

Mijn promotor, beste Vincent, jij weet zoveel van biologie, daar zou je jaloers op

worden. Bedankt, in het bijzonder voor je altijd supersnelle reacties op mijn verzoeken en

manuscripten en voor de onmisbare tweewekelijkse besprekingen. Jij wist altijd de positieve

kant van de resultaten te belichten en dat was soms hard nodig. Daardoor kwam ik telkens

weer vol enthousiasme van de besprekingen vandaan. Bedankt voor alles.

Mijn co-promotor, beste Teun, ook jouw input was immer positief, stimulerend en ook

gezellig. Je was heel erg betrokken bij alles en maakte altijd tijd voor mij vrij. Jij was vaak

zó positief dat ik er verlegen van werd, maar dit heeft er wel toe geleid dat ik trots ben op

het eindresultaat. Je bent geweldig.

Geachte leden van de leescommissie, bedankt voor het lezen en beoordelen van mijn

proefschrift en ik kijk uit naar de verdediging hiervan. Beste Bram, ook bedankt voor het

‘adopteren’ van mij tijdens congressen door jullie leuke groep. Beste Sue, succes en veel

plezier met je nieuwe rol bij OCB.

Dan, alle andere betrokken bij het lab van OCB. Beste Jenneke, bedankt voor je leuke

reacties op mijn onderzoek. We hebben de werkdag vaak afgesloten met een gezellig

gesprek of een borrel. Ook tijdens de vergadering van de DUC hebben we ontzettend veel

lol gehad.

Ineke, omdat jij met het onderzoek naar bisfosfonaten begon, ben je van het begin

af aan erg betrokken geweest bij mijn project. Het was leuk om samen te werken en heel

speciaal om tegelijkertijd de afrondingsfase in te gaan. Succes met de laatste loodjes! Ton

S., je bent pas in de laatste fase praktisch betrokken geraakt bij mijn project, maar hebt de

projecten van de studenten ontzettend vooruit geholpen. Dank voor je hulp, je eerlijkheid

en de gezelligheid, met name tijdens de vergaderingen met de DUC. Marion, je bent

geweldig goed en precies in het snijden van MMA coupes en daarom was je inzet hiervoor

onmisbaar. Daarnaast was het samenwerken ook heel gezellig, wat deze taak naast lastig

ook leuk maakte, dankjewel daarvoor. Jolanda, bedankt voor je enorme behulpzaamheid in

het lab. Wat jou verder nog bijzonder maakt, is dat je afwezigheid bijna net zo opvallend is

als je aanwezigheid, omdat je het meteen merkt als je lach een dag ontbreekt. Dirk-Jan,

dank dat je ondanks mijn vaak roze verschijning toch ook altijd een glimlach wist te

Dankwoord

145

produceren. Cor, ik heb genoten van onze e-mails die vaak volgden op jouw notulen van de

culture meetings en van de fietstochten na feestjes, samen met jou en Maybritt.

Ton B., hartelijk dank voor al je kennis en adviezen over histologie en voor de

gezellige liften naar Eindhoven. Don, je hebt vaak geprobeerd mij sneller te laten werken

door te vragen of mijn proefschrift nog niet onderhand af was. Helaas is het niet zo snel

gelukt en zit jij waarschijnlijk alweer in de zon. Astrid, ik vind het altijd mooi om te zien hoe

jij overduidelijk diep na kunt denken over resultaten. Bedankt voor de gezelligheid tijdens de

congressen van de ECTS. Clara, bedankt voor het ontwerpen van het protocol voor de

analyse van de mineralisatielabels. Jack, bedankt voor het me er telkens aan herinneren dat

ik geen leven heb. Ik kan je humor en directheid enorm waarderen.

Lieve Marjolein, ook al was je er niet op mijn eerste werkdag bij OCB, ik voelde me

meteen enorm welkom bij jou op de kamer. Ik heb heel veel bewondering voor hoe

behulpzaam en betrokken je bent, vooral voor ‘nieuwe’ mensen en zowel privé- als

werkgerelateerd. Dit heeft ertoe geleid dat er een mooie vriendschap tussen ons is ontstaan

en dat ik ook jouw gezin heb leren kennen. Alejandro, I admire your passion for many

things, especially your work. David, wat was het gezellig om op je te mogen passen en

samen te spelen.

Dear Alejandra, my paranymph, thanks a million times for all the lovely moments

at the gym, the swimming pool, and the fruit breaks. Thank you for the conversations about

everything and all the fun at drinks, parties, and dinners. Of course, a lot of thanks for great

cooking and proofreading goes to Alan. Thanks to you both for inviting us to your wedding

and organizing an amazing and unforgettable trip to Colombia.

Behrouz, je hebt me vaak enorm vooruit geholpen met je goede adviezen, je

antwoorden op mijn vragen en de hulp met de Western Blots. Bedankt ook voor je

gastvrijheid en je enthousiasme en natuurlijk de gezelligheid samen met de andere

kamergenoten. Henk-Jan, wat was het jammer dat je ons moest verlaten, maar gelukkig

mogen we je nog af en toe begroeten op onze befaamde roomies-drinks. Hartelijk dank ook

dat je nog dacht aan het doorsturen van vele vacatures. Janak, thanks for being a great

roomie and for the very nice weekend in Leuven. I am very happy that you got an

extension, of course for you to finish your work, but also because I would not want to miss

you and Jane at the defense.

Greetje, ik heb een heel hoofdstuk in dit proefschrift te danken aan jouw inzet voor

het project en het aanvragen van het DEC protocol. Het was leerzaam én gezellig om met je

samen te werken aan het onderzoek. Bedankt ook dat je altijd zo ontzettend attent bent.

Hessam, bedankt voor je betrokkenheid en enthousiasme. Bas, je hebt een

geweldig project dat ik leuk vond om te volgen, maak ervan wat je kunt! Bedankt voor je

interesse in mijn project. Thijs, jouw komst naar OCB bracht veel gezelligheid met zich mee,

Acknowledgments

146

bedankt daarvoor. Angela, thank you for your kindness and good luck with your project.

Sara en Yixuan, osteoclasts and their heterogeneity are great fun. Enjoy them and I hope

we will meet again in the future. Janice, succes straks met je verdediging en ik hoop dat je

het naar je zin hebt met het werken in de kliniek.

Jan Harm, bedankt voor de gezelligheid tijdens de lunch. Geerling, jouw kamer

bleek een inspirerende omgeving voor onze hilarische DUC vergaderingen, bedankt. Leo,

bedankt voor het maken van vele backups van de microCT scans. Hans en Peter, bedankt

voor de gezelligheid op het lab en de gezelligheid tijdens de borrels en dagjes uit.

Matthijs, ik ben benieuwd naar je proefschrift en dan natuurlijk vooral naar de resultaten

van de studie waar ik proefpersoon was. Sepanta, thanks for lots of fun and for introducing

me to the jenever bar.

Gang and Tie, it was fun to do experiments with you guys in the Medical Faculty.

Tie, thank you for all the pictures you took during events. Ceylin, it was very nice to work

with you on the article for NTvT. Nawal, you have a great personality and I wish you all the

best. Xingnan, thanks for your humor and the rides on the bike. Dongyun, I hope you enjoy

working for ACTAPro as much as I did. Francis en Berend-Jan, door jullie zijn de promovendi

van OCB en de Material Girls vaker samen gaan borrelen, so thank you guys, and Ana and

Ana for all the fun. René and Hong, I bet you will be fun parents for your daughter and I

wish you all the best. Jing and Gang, good luck in finding your way between Chinese and

Dutch habits, especially with your cute boy! Mahshid, good luck with the meshes. Rozita,

wat heb jij snel Nederlands geleerd, ik wens je een succesvolle carrière hier. Patrick,

bedankt voor je vrolijkheid in en om het lab. Britt, jammer dat je niet vaker in Amsterdam

was, want als je er was was het gezellig.

Veerle, bedankt voor je goede advies dat ik maar al te graag heb opgevolgd. Nina,

wat fijn dat jij mensen zo goed kunt overtuigen om altijd net iets langer op een borrel te

blijven hangen. Bedankt voor de gezelligheid, gelukkig ook nog nadat je OCB officieel had

verlaten. Rishikesh, I wish you and Rashmi all the best in Australia and in all the places you

may end up. Marjoleine, bedankt voor je humor. Dwayne, good luck with your career. Anna,

ik hoop dat je in goede gezondheid aan je sportcarriere kan werken, succes! Daan, wat

jammer voor ons dat je op een gegeven moment minder met OCB te maken hebt gekregen,

want het was leuk om jou erbij te hebben. Petra, bedankt voor de gezellige gesprekken. Ik

wens je veel geluk.

Alle studenten die hebben meegewerkt aan mijn project, Manon, Emine, Matangi,

Mei-Ling, Jeroen, Sophie en Lieneke, hartelijk dank voor jullie inzet.

Finally, thanks to all the guests that temporarily joined OCB, Eli, Jesus, Pedro,

Nilufer, Anna, and Sreeda, it was great to meet you. Dear Marina, thank you for so many

hilarious moments. One year was too short and I hope we will stay in touch. Qilong, I really

Dankwoord

147

enjoyed our conversations during bone marrow isolations. Noel, thank you for the nice chats

about science and careers. Suzanne, dank je voor de gezelligheid in en vooral ook buiten

het lab. Mirte, ik hoop van harte dat je project verlengd wordt, want je verdient het. Ben, ik

maakte kennis met jouw spontaniteit tijdens de cursus microscopie. Des te leuk was het dat

je later bij OCB experimenten kwam doen. I wish you, Manuel, and Samaneh good luck with

your challenging and interesting project.

Dear Fraser, thank you for giving me the opportunity to start working with fluorescently-

labeled bisphosphonates. Your critical comments on the manuscript were highly

appreciated. Beste Jan Stap en Ron Hoebe, hartelijk dank voor jullie hulp met de uitvoering

en uitwerking van de time-lapse microscopie experimenten. Jan, ik vond het samenwerken

op het AMC altijd supergezellig.

Ook hartelijk dank aan alle medewerkers van orale biochemie. Het begon ten slotte allemaal

bij jullie op het lab in de medische faculteit. Menno, Heinze, Kamran, Jan, Toon, Henk,

Petra, Marianne, Floris en Enno, bedankt. Enno, hartelijk dank voor het meten van de

bisfosfonaatconcentraties en voor het plezier dat je daarin had. Jan en Kamran, bedankt

voor de hulp met de flow cytometer. Nivedita, you must have been very focused, finishing

your PhD thesis so quickly, admirable! Irene, wat leuk dat je ondanks dat je grotendeels op

het AMC aan je project werkt toch nog vaak langs kwam bij borrels en activiteiten. Mireille,

leuk dat ook jij een trouwe bezoeker was van deze activiteiten. Tjitske en Sabrina, het was

prettig en leuk om met jullie samen te werken in ACTAPro en jullie zo beter te leren kennen.

Ook alle anderen waarmee ik in het promovendioverleg samen heb gewerkt, Nina, Wilco,

Jacobien, Janice, Greetje, Catherine, Irshad en Denise, bedankt voor een leuke en leerzame

tijd. Vincent en Martijn, bedankt dat jullie openstonden voor onze suggesties en dat we zo

veel leuke activiteiten mochten organiseren. Martine, Peter en Moniek, bedankt voor de

ondersteuning hiervan.

Natuurlijk zijn er tussen het onderzoek door ook veel weekenden geweest waarin ik veel tijd

buiten Amsterdam heb besteed. Martijn, Katelijne, Dirk-Jan, Anne-Lotte, Matthijs, Ruud,

Carlien en Janneke, de weekendjes in Bad Bentheim, Den Bosch, Wroclaw en Florenville

waren natuurlijk superkut, dank jullie wel Dolly. Ook alvast bedankt voor de aanstaande

weekendjes in Volendam, straks weer in Gent en in Oxford.

Dank aan alle Zwevelpinners voor het Seloyencieweekend bij ons in Amsterdam, de

vurige zeilweekenden en de gezellige etentjes in Utrecht.

Kasper, Katelijne, Lisanne S., Lisanne de W., Mark, Niels, Nora, Reinout en alle

anderen, dank voor het gezelschap tijdens mijn eerste festivals: Lowlands en Pukkelpop. Ik

Acknowledgments

148

hoop dat er nog meer zullen volgen. Katelijne, Kasper en Nora, ook bedankt voor het

afreizen naar Amsterdam voor de vaak minder succesvolle, maar wel leuke quizavonden.

Lieve Flügelmeisjes, Carlien, Daniëlle, Ellen, Marjolein en Martje, jullie vriendschap

is onmisbaar. Gelukkig is deze gemakkelijk te onderhouden en zal die altijd blijven bestaan,

want hoe lang het soms ook duurt voordat we weer wat af kunnen spreken, het is altijd

weer een feest om jullie te zien.

Verder waren de weekenden ook vaak gevuld met familieactiviteiten en feesten en

daarvoor wil ik oma Vermeer, oma van den Broek en de hele familie heel erg bedanken.

Dinie, ontzettend bedankt voor je onuitputtelijke interesse. Ik voel me altijd welkom bij jou

en probeer nog steeds te leren om net zoveel vragen te kunnen stellen als jij. Wil, ook jij

bedankt voor je interesse en gezelligheid. Jeroen, het was erg leuk om betrokken te zijn bij

je verhuizing naar Södertälje. Hopelijk kunnen we elkaar ondanks de afstand regelmatig

bezoeken.

Steffi, als grote zus is het geweldig om te zien dat jij steeds zelfstandiger en steeds minder

mijn kleine zusje wordt. Ik ben supertrots op jouw doorzettingsvermogen! Irma, bedankt

dat je mijn paranimf wilt zijn en voor je interesse. Vooral de laatste jaren waren de avondjes

in de Zwaan me dierbaar. Sebastian, ook jij bedankt hiervoor en voor de fijne discussies.

Ankie, het is heerlijk om altijd zo enthousiast door jou begroet te worden en lekker met je te

knuffelen. Rob, blijf vooral je gezin verblijden met je geweldige gevoel voor humor. Harm, jij

bent een extra goede reden om regelmatig naar het zuiden van het land af te reizen. Voor

jouw lach wringen we ons letterlijk in de gekste bochten. Lief zijn straks voor de kleine

Ieniemienie. Ellen, van kleins af aan hebben we een kamer gedeeld en al vlug zag ik je

naast als zus ook als vriendin. Bedankt voor al die fijne momenten. Martijn, bedankt voor de

gezelligheid en gastvrijheid, vooral tijdens de Nijmeegse Vierdaagse.

Lieve pap en mam, het is een genot om nog regelmatig op de boerderij in Heeze te

komen waar ik me nog steeds zo ontzettend thuisvoel. De nuchtere en relativerende kijk op

het leven die jullie op mij hebben overgebracht hebben mij enorm geholpen in het dagelijks

leven en tijdens mijn promotie. Ik ben heel erg blij met jullie. Bedankt voor alles.

Tot slot, lieve lieve Martijn, wat is het geweldig om alles in het leven samen met jou te

beleven. We hadden van tevoren niet gedacht dat we ooit in Amsterdam zouden gaan

wonen, maar samen met jou in ons appartement (met akker, dat wel) is het alleen maar fijn

geweest. Ik hoop dat je snel zult promoveren, zodat we weer gauw samen kunnen zijn.

Bedankt voor al je steun, maar vooral voor het samen met mij genieten van het leven.

About the author

About the author

151

Jenny Vermeer was born on the 13th of March 1984 in Heeze, the Netherlands. After

receiving her VWO diploma at SG Augustinianum in Eindhoven, she started her BSc studies

in Biotechnology at Wageningen University and Research Centre in 2002. In 2006 she took

one gap year to fulfil a full time post in the board of her student association, SSR-W. After

that and as part of her Masters Medical Biotechnology, she did research for her thesis at the

department of Metabolism and Nutrigenomics at Wageningen UR under the supervision of

dr. Bart L.T. Vaes. Subsequently, she did a research project at the department of

Experimental Immunology at the KU Leuven in Belgium under the supervision of prof.dr.

Peter W. Hellings. She finished her MSc in 2009 with an internship at the department of

Molecular Cell Biology at Sanquin Research in Amsterdam under the supervision of dr. Paula

B. van Hennik. In 2010, she started her PhD research at the department of Oral Cell Biology

at the Academic Centre for Dentistry Amsterdam (ACTA) under the supervision of her

promotor, prof.dr. Vincent Everts and copromotor, dr.ir. Teun J. de Vries. The results of this

project are described in this thesis.

152

List of publications

Vermeer JAF, Jansen IDC, Marthi M, Coxon FP, McKenna CE, Sun S, de Vries TJ, Everts V.

Jaw bone marrow-derived osteoclast precursors internalize more bisphosphonate than long-

bone marrow precursors. Bone. 2013 Nov;57(1):242-51

Jansen IDC, Vermeer JAF, Bloemen V, Stap J, Everts V. Osteoclast fusion and fission.

Calcif Tissue Int. 2012 Jun;90(6):515-22

Vermeer JAF, Renders GAP, van Duin MA, Jansen IDC, Bakker LF, Kroon SA, de Vries TJ,

Everts V. Zoledronic acid differently affects long bone and jaw bone turnover and induces

molar root resorption in female mice. (submitted for publication)

Vermeer JAF, Schoenmaker T, Tsui ML, Stap J, de Vries TJ, Everts V. Migration, fusion and

CXCL12-CXCR4-mediated chemoattraction of long bone and jaw osteoclast precursors.

(submitted for publication)

Renders GAP, Vermeer JAF, Leung PM, Reynaert F, Prins C, Langenbach GEJ, Everts V.

Implications of high-dosage bisphosphonate treatment on bone tissue in the jaw and knee

joint. (accepted for publication in Calcif Tissue Int.)

Vermeer JAF, Langeveld J, Schoenmaker T, Everts V, de Vries TJ. The effect of

bisphosphonates on periodontal-ligament-fibroblast-mediated osteoclastogenesis.

(manuscript in preparation)

Vaes BLT, Lute C, van der Woning SP, Piek E, Vermeer J, Blom HJ, Mathers JC, Müller M,

de Groot LCPGM, Steegenga WT. Inhibition of methylation decreases osteoblast

differentiation via a non-DNA-dependent methylation mechanism. Bone. 2010

Feb;46(2):514-23

About the author

153

International meetings attended

2014 Annual meeting of the European Calcified Tissue Society (ECTS), Prague,

Czech Republic (poster presentation)

2013 Annual meeting of the American Society for Bone and Mineral Research

(ASBMR), Baltimore, MD, USA (poster presentation)

2013 Annual meeting of the European Calcified Tissue Society (ECTS), Lisbon,

Portugal (poster presentation)

2011 Gordon Research Conference Cell-Cell Fusion (poster presentation) and

Gordon Research Seminar Cell-Cell Fusion, Biddeford, ME, USA (oral

presentation)

2010 PhD training course of the European Calcified Tissue Society (ECTS),

Berlin, Germany (oral presentation)

National meetings attended

2010-2013 Annual meeting of the Dutch Society for Calcium and Bone Metabolism

(NVCB), Zeist, the Netherlands. 2010 (oral presentation), 2011 (oral

presentation), 2012 (oral presentation), 2013 (oral presentation, chair)

2010-2013 Annual meeting of the Netherlands Institute of Dental Sciences, Lunteren,

the Netherlands. 2010, 2011, 2012, 2013 (oral presentation)

2010-2014 Annual meeting of the MOVE Research Institute Amsterdam, Amsterdam,

the Netherlands. 2010 (poster presentation), 2011 (poster presentation),

2013, 2014 (oral presentation)

8mm

8mm

Documents

Proefschrift Vermeer