Anti-IL-17A therapy protects against bone erosion in experimental models of rheumatoid arthritis

Anti-IL-17A therapy protects against bone erosion in experimentalmodels of rheumatoid arthritis

CHENG-CHI CHAO1,*, SHI-JUANCHEN1,

*, IANNIS E. ADAMOPOULOS1, NICOLEDAVIS2,

KYU HONG2, ANNAVU2, SYLVIA KWAN2, LAURENCE FAYADAT-DILMAN2,

AGELIO ASIO3, & EDWARD P. BOWMAN1

1Department of Immunology, Merck Palo Alto (formerly DNAX Research Institute), Palo Alto, CA 94304, USA, 2Department

of Bioanalytical and Protein Chemistry, Merck Palo Alto, Palo Alto, CA, USA, and 3Department of Experimental Pathology

and Pharmacology, Merck Palo Alto, Palo Alto, CA, USA

(Submitted 7 May 2010; revised 6 August 2010; accepted 19 August 2010)

AbstractInterleukin-17A (IL-17A) is a pro-inflammatory cytokine secreted by a subset of memory T cells and other innate immunecells. It is associated with rheumatoid arthritis (RA) due to IL-17A expression in RA synovial fluid. The severe bone erosive ratadjuvant-induced arthritis (rAIA) and mouse collagen-induced arthritis (mCIA) models were used to address the therapeuticefficacy of anti-IL-17A treatment with a focused investigation on bone protection. In the rAIA model, treatment with anti-IL-17A completely alleviated arthritis, lowered the level of receptor activator of NFkB ligand (RANKL), and inhibited structuraldamage to the bones. In the mCIA model, IL-17A neutralization coincident with arthritis development or in mice withestablished arthritis diminished joint swelling by inhibiting disease initiation and progression. Intriguingly, even the few jointsthat became outwardly severely inflamed in the presence of an anti-IL-17A antagonist had diminished joint histopathologyscores compared to severely inflamed, control-treated mice. The bone-preserving property correlated with decreased RANKLmessage in severely inflamed paws of arthritic mice. These data identify IL-17A as a key factor in inflammation-mediated bonedestruction and support anti-IL-17A therapy for the treatment of inflammatory bone diseases such as RA.

Keywords: IL-17A, rheumatoid arthritis, inflammation, adjuvant-induced arthritis, collagen-induced arthritis

Introduction

Rheumatoid arthritis (RA) is an autoimmune disease

characterized by joint inflammation initially resulting

in pain and swelling. In a majority of patients, there is

progressive bone and cartilage erosion of the joint with

current treatment regimens that aim to control the

inflammation to retard structural damage. Mild to

moderate RA is treated with disease-modifying anti-

rheumatic drugs such as methotrexate and steroids.

Methotrexate is often used in combination with TNF

antagonists, such as Etanercept or Infliximab, for

severe RA.

IL-17A is the signature pro-inflammatory cytokine

produced by Th17 memory T cells [1]. Th17 cells

have been identified as the key T-cell subset implicated

in multiple autoimmune conditions such as multiple

sclerosis, Crohn’s disease, and RA [2,3]. Multiple

lines of evidence point to the association of IL-17 with

RA. IL17A message is present in mononuclear cells

harvested from RA synovial fluid [4], and immuno-

histochemistry demonstrates the presence of IL17A

protein in infiltrating synovial cells [5–7]. A subset of

T-cell lines expanded in vitro from RA synovium

express IL-17A following activation [8]. IL-17A

protein is found in RA synovial fluid, but rarely in

osteoarthritis synovial fluid, consistent with the

inflammatory nature of RA vs. osteoarthritis [9–11].

Two prospective clinical trials not only supported

the association of IL-17A with RA, but importantly

implicated its correlation with poor disease prognosis.

*The first two authors (C. C. Chao and S. J. Chen) contributed equally to this work.

Correspondence: E. P. Bowman, Department of Immunology, Merck Palo Alto, 901 California Avenue, Palo Alto, CA 94304, USA.Tel: (650) 496 1249. Fax: (650) 496 1200. E-mail: [email protected]

Autoimmunity, May 2011; 44(3): 243–252q Informa UK, Ltd.ISSN 0891-6934 print/1607-842X onlineDOI: 10.3109/08916934.2010.517815

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Lib

rary

Utr

echt

on

08/0

1/13

For

pers

onal

use

onl

y.

Raza et al. prospectively collected synovial fluid

samples within a few weeks after symptom onset

from patients with early synovitis. Patient outcomes

were subsequently noted 18 months after fluid

acquisition to determine which early factors were

correlated with the progression of synovitis to RA.

Patients were grouped into those who subsequently

progressed to RA, those who subsequently progressed

to nonrheumatoid persistent synovitis, or those whose

synovitis was resolved.

Importantly, IL17A was one of the few factors

whose expression in the earliest stages of disease was

associated with subsequent progression to RA [12].

Kirkham et al. [13] concluded that IL17A gene

expression in synovial membrane biopsies was one

factor that was predictive for subsequent bone erosion

and joint damage as assessed byMRI and radiography.

Collectively, these data support that IL-17A is present

in the inflamed synovium and that IL-17A expression

levels correlate with poor prognosis and greater joint

destruction.

Although previous lines of evidence supported that

IL-17A plays a pathogenic role in arthritis, how

IL-17A may be involved in arthritis-mediated joint

pathology was less specifically addressed. We have

used the bone erosive rat adjuvant-induced arthritis

(rAIA) and mouse collagen-induced arthritis (mCIA)

models to investigate the therapeutic benefits of anti-

IL17A in these models, with a focused analysis of

cartilage and bone protection.

Materials and methods

Reagents and antibodies

The rat anti-rat IL-17A antibody JL8.18E10 used in

the rAIA model was developed using conventional

hybridoma technology to immunize rats with IL-17A

and screen for clones that bound to rat IL-17A.

JL8.18E10 has an affinity of 620 pM for rat IL-17A by

Biacore analysis and neutralizes 0.3 nM rat IL17A-

stimulated IL-6 production with an IC50 of 0.2 nM.

The neutralizing rat anti-mouse IL17A antibody

(1D10) used for the mCIA model was developed at

Schering-Plough Biopharma. It binds tightly to mouse

IL-17A (KD ¼ 10 pM by Kinexa) and neutralized

1 nM mouse IL-17A-stimulated ST2 cell IL-6

production with an IC50 of 0.5 ^ 0.2 nM. 1D10 did

not bind to any other mouse IL17 family member,

including IL17F, in an ELISA format.

Rat adjuvant-induced arthritis and mouse collagen-

induced arthritis

Male Dark Agouti rats were anesthetized with

isoflurane and 100ml of 10mg/ml Mycobacterium

butyricum cell wall suspension in paraffin oil, and

mannide monoleate (Difco, Detroit, MI, USA) was

injected intradermally into the base of the tail. Rats

with severe arthritis were dosed once on day 11 with

20mg/kg JL8.18E10 or isotype control antibody

subcutaneously. Some rats were sacrificed 3 days

post-dosing for gene expression studies, and other rats

were observed until day 23. Arthritis severity was

measured using calipers to measure paw thickness.

Male B10.RIII mice (Jackson Laboratory, Bar

Harbor, ME, USA) were immunized with bovine

type II collagen (CII) (Sigma-Aldrich, St. Louis, MO,

USA) emulsified intradermally with complete

Freund’s adjuvant (DIFCO). Fifty microliters of

1mg/ml of CII/CFA emulsion was injected on both

sides of the tail. Mice were challenged intradermally

21 days later with 1mg/ml CII emulsified in

incomplete Freund’s adjuvant. Mice were observed

daily until the first signs of a severely swollen paw

within the cohort. On that day (designated day 0), the

remaining mice were randomized to different treat-

ment groups, and received 20–30mg/kg anti-IL17A

or control antibody administered subcutaneously.

Subcutaneous dosing was repeated weekly for a total

of five doses, with sacrifice of mice seven days after the

last dose.

Microcomputed tomography imaging

Rat hind paws fixed in 10% neutral buffered formalin

were scanned by GE eXplore Lotus mCT scanner

(GE, Piscataway, NJ, USA). Data were acquired at

46mm isotropic voxel size with 720 projections by 360

degree scan, integration time of 400ms with six

frames, photon energy of 80KeV and current at

450mA. The duration of imaging was approximately

1 h per scan, followed by 40min of projection

correction and volume reconstruction. 3-D image

renderings were generated through original volumetric

reconstructed images using Microview software (GE).

Visual disease severity scoring system(DSS)

The number of swollen paws and the severity of edema

were tracked using a categorical visual disease severity

scoring (DSS) system. DSS ¼ 0 indicates no visual

change in paw appearance. DSS ¼ 1 indicates redness

and swelling in any one distal digit. DSS ¼ 2 indicates

redness and swelling in two or more distal digits or

swelling in the palm of the paw. DSS ¼ 3 indicates

swelling in palm/wrist plus swelling in any distal digit.

DSS ¼ 1 is fairly mild with limited focal inflam-

mation; DSS ¼ 2–3 were grouped together as

“severely inflamed paws”.

Histological analysis of inflamed joints

Paws were fixed in neutral buffered formalin,

de-calcified in EDTA solution, embedded in paraffin,

and stained. Sections were scored by a board-certified

C.-C. Chao et al.244

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Lib

rary

Utr

echt

on

08/0

1/13

For

pers

onal

use

onl

y.

pathologist blinded to treatment arms and

were analyzed by a categorical scoring system

(Supplementary Table S1).

Statistical analysis

The differences in the clinical paw scores and paw

histology scores between isotype control-treated and

anti-IL-17A-treated mice were evaluated using the

non-parametric Mann–Whitney U-test. Note that

after statistical analysis, the original data were

manually modified slightly higher or lower than

original values in order to graphically present the

data in Figures 2C, 3, and 4.

Results

Therapeutic anti-IL-17A protects against bone erosion in

severely inflamed paws in rAIA

Arthritic rats were dosed with anti-IL-17A or control

antibody after severe arthritis had been established to

address the therapeutic potential of anti-IL17A

treatment to provide bone protection in an inflam-

mation-driven bone erosive environment. A single

dose of anti-IL-17A reversed the established arthritis,

and in 9 days the paws of the anti-IL17A-treated rats

reverted back to a practically naı̈ve state (Figure 1A).

mCT analysis of paws from control antibody-treated

rats confirmed massive bone erosion of a subset of

tarsal bones that had rendered them almost unrecog-

nizable in X-ray analysis.

The previously severely inflamed paws treated with

anti-IL-17A, in contrast, had little to no detectable

bone erosion (Figure 1B). Severely arthritic rats were

dosed with anti-IL-17A or control antibody for a short

time (3 days), and paw gene expression analysis was

performed to mechanistically ascertain how IL17A

neutralization gave rise to reduced bone erosion. After

only three days of IL17A neutralization (i.e., at a time

point prior to dramatic changes in paw swelling), the

arthritis-elevated and osteoclast-specific genes, such

as TRAP, cathepsin K (CatK) and MMP-9, were

decreased in the paws (Figure 1C). Anti-IL-17A also

significantly lowered the arthritis-elevated levels of

serum receptor activator of NFkB ligand (RANKL)

with minimal activity on serum osteoprotegrin (OPG)

levels (Figure 1D).

Inhibition of collagen-induced arthritis (CIA)

disease progression by anti-IL-17A therapy in

mCIA

We next investigated whether the bone sparing activity

following IL-17A neutralization would hold true in the

mCIA model. Anti-IL-17A was administered to CII-

immunized mice either after severe arthritis was first

observed in the cohort (Figure 2A and C) or after each

mouse displayed a severely arthritic paw and was

recruited into the dosing phase of the study

(Figure 2D–E). Figure 2A shows that mice treated

with a control antibody, after severe arthritis was

observed in their cohort, had a disease incidence of

approximately 90% over the 5-week observation

period post-boost. In contrast, mice therapeutically

treated with an anti-IL-17A antibody showed delayed

and diminished disease incidence, which gave rise to

decreased total animal DSS in the anti-IL-17A-

treated cohorts, as shown in Figure 2B.

Unlike joint inflammation in rAIA, the joint

inflammation in a collagen-induced arthritis (CIA)

mouse is asynchronous and of variable intensity. For

reasons outlined below, it was important to view the

DSS of each paw individually vs. the conventional

method of summing the DSS of four paws to generate

an animal’s total DSS (Figure 2C). A high percentage

of paws in the control-treated group became severely

swollen (DSS ¼ 2–3), but some paws showed no signs

of swelling (i.e. joint swelling was neither synchronous

nor uniform within an isotype-treated CIA mouse). In

a cohort of mice, anti-IL17A treatment at disease

onset inhibited paws from progressing to a severely

swollen state over the 5-week observation period. The

few inflamed paws in the anti-IL-17A-treated

group would ultimately allow higher level analysis of

how IL-17A neutralization affects the natural disease

progression (cartilage and bone destruction) in a

severely inflamed environment, as reported next.

Alternatively, individual mice were recruited into

the dosing phase of the study, once each exhibited a

severely arthritic paw, and they were dosed weekly for

3 weeks. Control-treated mice demonstrated pro-

gressive, severe arthritis development in the “other”

paws, giving rise to increasing disease severity over

time (Figure 2E). In contrast, IL-17A neutralization

in a mouse with established arthritis prevented further

arthritis development in the remaining paws. None of

the recruited arthritic mice advanced more than one

severity score when dosed with anti-IL-17A

(Figure 2D). The outward clinical score of the

already arthritic paw that recruited the mouse into the

study was not changed by isotype or by anti-IL-17A

therapy.

Inhibition of joint structural damage by

anti-IL-17A therapy in the mCIA model

Figure 3 combines results from multiple studies to

outline the benefit of anti-IL-17A neutralization

on various histopathological changes that occur as a

consequence of autoimmune inflammation in the

articular joint. Anti-IL17A significantly decreased

leukocyte recruitment, pannus formation, articular

cartilage destruction, and bone erosion in the

cohort of treated mice. The effect of anti-IL17A on

all histological parameters was consistent with its

Anti-IL-17A therapy and bone erosion in RA 245

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Lib

rary

Utr

echt

on

08/0

1/13

For

pers

onal

use

onl

y.

beneficial effect on swollen paw incidence. Impor-

tantly, all histology parameters (as reflections of

multiple components of the disease pathobiology)

were similarly sensitive to anti-IL-17A treatment.

Although there was statistically significant inhi-

bition of swollen paw incidence by anti-IL-17A

treatment, there were a few paws that became visually

severely swollen. Importantly, these few DSS ¼ 2–3

swollen paws then allowed investigation into the

impact of therapeutic IL-17A neutralization on the

natural disease progression that occurs in this

explosively erosive model even when paws are severely

Figure 1. IL-17A neutralization inhibits arthritis progression and protects against structural joint damage in rAIA. A, Day 24 photographs of

front and hind paws of naı̈ve rats and arthritic rats treated with isotype or anti-IL-17A on day 11. Time course of swollen paw resolution by

20mg/kg anti-IL-17A dosed once at day 11 (data representative of three experiments). B, micro-CTanalysis of day 24 hind paw tarsal bones

(data representative of two experiments). C, gene expression of osteoclast markers from rAIA paws at day 14. D, serum RANKL/OPG levels

after treatment with isotype or anti-IL-17A antibody (data representative of two experiments). *p , 0.05, **p and ***p , 0.005 analyzed by

unpaired t-tests. ns, non-significant difference.


Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Lib

rary

Utr

echt

on

08/0

1/13

For

pers

onal

use

onl

y.

swollen. Figure 4 shows the analysis of the various

histology parameters only in severely swollen

DSS ¼ 2–3 paws. By focusing on only severely

swollen paws, we have removed the prominent effect

of therapeutic anti-IL17A treatment on disease

initiation and have allowed focused analysis on the

mechanisms of joint destruction, and on whether

IL-17A neutralization changes the natural progression

of joint destruction once it occurs in this erosive

disease model.

As expected from the results shown in Supplemen-

tary Figure S1 where the magnitude of paw swelling

was highly correlated with underlying joint histo-

pathology, control-treated paws that were severely

swollen had high histology scores for leukocyte

recruitment, pannus formation, cartilage destruction,

0 7 14 21 28 350

25

50

75

100

**

Time after mAb treatment (days)

Dis

ease

inci

denc

e

0 7 14 21 28 350

2

4

6

8

**

**

**

**

Time after mAb treatment (days)

Mea

n di

seas

e se

verit

y (0

-12)

A

B

0 7 14 21–2

0

2

4

6

8

10D

Time after anti-IL-17A treatment (days)

Cha

nge

in D

SS

pos

t-re

crui

tmen

t

0 7 14 21–2

0

2

4

6

8

10E

Time after isotype treatment (days)

Cha

nge

in D

SS

pos

t-re

crui

tmen

tIsotype Anti-IL-17A

0

1

2

3

C

p < 0.0001

D35

DS

S (

0-3)

/paw

Figure 2. Anti-IL-17A therapy inhibits arthritis progression in mCIA. Mice were immunized, challenged with bovine CII, and randomized

to different treatment groups. When the first sign of severe paw swelling was seen in the cohort, the rest of the mice were then treated

subcutaneously with 20–30mg/kg anti-IL-17A (filled triangles) or an isotype control antibody (open diamonds) weekly for 5 weeks. Seven

independent experiments were combined to give mean ^ SEM for disease incidence (A) and disease severity (B). Seventy-two mice were

treated with isotype control antibody, and 52 mice were treated with anti-IL17A in total among the seven experiments. C, individual paw

scores are presented instead of summing the four paw scores, as shown in Figure 2A–B. Values from seven independent experiments have been

combined (288 paws from isotype control-treated mice and 208 paws from anti-IL-17A-treated mice) and presented slightly higher or lower

than the standard scores of 0, 1, 2, and 3 for graphical purposes. D and E, samemCIA protocol as panel A and B, but individual mice recruited

into study when any paw reached DSS . 2. The recruited mice (n ¼ 5) were dosed after arthritis established with anti-IL-17A (D) or control

antibody (n ¼ 5) (E). *p , 0.05 and **p , 0.01 vs. controls analyzed by the Mann–Whitney U-test.


Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Lib

rary

Utr

echt

on

08/0

1/13

For

pers

onal

use

onl

y.

and bone erosion. Intriguingly, severely swollen paws

from anti-IL-17A-treated mice had statistically lower

histopathology scores for pannus formation and bone

erosion (Figure 4A–D). The inhibition of joint

destruction was demonstrated in severely swollen

paws from anti-IL17A-treated mice. The mechanism

of joint protection could not be solely attributed to

modulation of leukocyte recruitment, since anti-

IL17A did not have a significant effect compared to

isotype control on the magnitude of the leukocyte

infiltrate (including neutrophils) in the severely

swollen joints.

Additionally, anti-IL-17A-treated mice had lower

bone erosion and pannus formation scores vs. isotype-

treated mice, even when severely swollen joints with

equal leukocyte infiltrate histology score (histology

score 2, 3, and 4) were compared. Potentially, anti-IL-

17A modulation of cartilage damage is more closely

tied to leukocyte infiltratem, since no significant

difference or trends in cartilage damage were observed

once severely swollen joints with equal leukocyte

infiltrate histology score (histology score 2, 3, and 4)

were compared between the two treatment groups.

Representative histology pictures from severely

inflamed paws treated with isotype or anti-IL-17A

are shown in Figure 4E–G.

Modulation of RANKL expression by

anti-IL-17A therapy

RANKL is critical in promoting pre-osteoclast

differentiation into mature osteoclasts. RANKL is

elevated in rodent arthritis models [14] and in human

RA [15] and has been used as an indicator of bone

erosion. Figure 5 shows that RANKL mRNA was

elevated in severely inflamed paws (DSS ¼ 3) and that

anti-IL-17A therapy normalized RANKL expression

even in the few DSS ¼ 3 paws that were outwardly

swollen. RANKL message was not elevated in paws

with only mild focal inflammation (DSS ¼ 1). Like-

wise, arthritis-induced elevated serum RANKL

protein was decreased by anti-IL17A treatment (data

not shown). The decreased RANKL in gene and

protein levels was also demonstrated in anti-IL-17A

treated adjuvant-induced arthritic rats, shown in

Figure 1C and D.

Discussion

IL-17A, present at the earliest stages of RA, is one of a

few proteins that retrospectively predicted patients

whose initial synovitis would ultimately progress to

RA [12,13]. Importantly, the validated RA targets

TNF and IL-1b were hardly present, much less

predictive, at this early stage of disease. We have

01234

Pan

nus

form

atio

n (0

-4)

01234

Car

tilag

e da

mag

e (0

-4)

01234

Bon

e er

osio

n (0

-4)

01234

Isotype Anti-IL-17A

Isotype Anti-IL-17A Isotype Anti-IL-17A

Isotype Anti-IL-17A

4

p < 0.001A B

C D

p < 0.001

p < 0.001p < 0.001

Leuk

ocyt

e in

filtr

ate

(0-4

)

Figure 3. Inhibition of histopathologic damage to joints by anti-IL-17A.Mice were immunized, challenged with bovine CII, and treated with

anti-IL-17A or an isotype control antibody at the first sign of severe paw swelling weekly for 5 weeks. After 35 days of treatment, front and back

paws were removed, fixed in neutral-buffered formalin, paraffin embedded, sectioned, stained with H and E, and scored by a board-certified

pathologist. The examined histological parameters were leukocyte infiltrate (A), pannus formation (B), cartilage damage (C), and bone

erosion (D). Horizontal bars indicate median values of the histological scores from 203 paws from isotype-treated mice and 208 paws from

anti-IL-17A-treated mice. Statistical analysis was performed by non-parametric Mann–Whitney U-analysis. Note that after statistical

analysis, the original data were manually modified slightly higher or lower than original values in order to graphically present the data (see

Materials and methods). The thick line represents the median value of histology scores prior to manual modification.


Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Lib

rary

Utr

echt

on

08/0

1/13

For

pers

onal

use

onl

y.

confirmed that IL-17A is present in a subset of early

RA patients and that RA patients have the highest

frequency of detectable IL-17A compared to other

nonautoimmune arthritides. It was noteworthy that

most early RA patients who had detectable IL-17A

had also started to show detectable TNF and IL-1b

expression (data not shown), which had not been

observed by Raza et al. in very early RA patients. IL-6

was found at high levels in all synovitis condition,

regardless of whether they ultimately progressed to

RA [12], and IL-6 was present in all synovial fluids we

obtained regardless of disease type.

Several IL-17A antagonists have been used to

address the efficacy of therapeutic IL17A neutraliz-

ation in a variety of arthritis models. IL-17RA-Ig

therapy started after CII immunization, but prior to

challenge, decreased disease incidence and ankle

radiographic scores in a CIA model [16]. Polyclonal

E F G

Tb T

N C

0

1

2

3

4

p < 0.001

p < 0.001

Sev

erel

y in

flam

ed p

awpa

nnus

for

mat

ion

(0-4

)S

ever

ely

infla

med

paw

bone

ero

sion

(0-

4)

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4



n.s.A B

C D

Sev

erel

y in

flam

ed p

awle

ukoc

yte

infil

trat

e (0

-4)

Sev

erel

y in

flam

ed p

awca

rtila

ge d

amag

e (0

-4) p < 0.05

0

1

2

3

4

5

DSS = 0 DSS = 3, Isotype DSS = 3, Anti-IL-17A

Figure 4. Diminished pannus formation, cartilage damage, and bone erosion in severely swollen paws by anti-IL-17A inhibition. Mice were

immunized, challenged with bovine CII, and treated with anti-IL-17A or an isotype-control antibody at the first sign of severe paw swelling

weekly for 5 weeks. After 35 days of treatment, the severely inflamed front and back paws were removed, fixed in neutral-buffered formalin,

paraffin embedded, sectioned, stained with H and E, and scored by a pathologist. Histological parameters examined were leukocyte infiltrate

(A), pannus formation (B), cartilage damage (C), and bone erosion (D) in severely swollen DSS ¼ 2–3 paws. Horizontal bars indicate median

histological values from 119 DSS ¼ 2–3 paws from isotype-treated mice paws and 50 DSS ¼ 2–3 paws from anti-IL-17A–treated mice in

each histopathologic parameter. Joint histology in control and severely swollen paws (E–G). E, Representative of a DSS ¼ 0 hind paw from an

anti-IL-17A-treated animal. It was never swollen at any time during the 5 weeks and was graded as “0” (no damage) for its bone erosion

histology. F, representative of DSS ¼ 3 hind paw from an isotype-treated animal. It was severely inflamed for 3 of the 5 weeks and received a

“4” (severe) for its bone erosion histology. A bone erosion score ¼ 4 was the median value for the many severely inflamed isotype-treated paws

in panel D. G, representative of DSS ¼ 3 hind paw from an anti-IL-17A treated animal. It was severely inflamed for 4 of the 5 weeks and

received a “3” (moderate) for its bone erosion histology. A bone erosion score ¼ 3 was the median value for the severely inflamed anti-IL-17A

treated paws in panel D. Tb, tibia; T, talus; N, navicular bone; C, cuneiform bone.


Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Lib

rary

Utr

echt

on

08/0

1/13

For

pers

onal

use

onl

y.

anti-IL-17A decreased clinical scores, resulting in

reduced synovitis, cartilage damage, chondrocyte

death, PG depletion, and bone erosion in collagen-

induced arthritis [17,18]. Polyclonal anti-IL17A

treatment inhibited antigen (mBSA) induced knee

swelling, PG depletion, and bone erosion in the

exacerbated smoldering knee [18]. Mice immunized

with formalin-fixed Borrelia burgdorferi and challenged

with live B. burgdorferi displayed transient joint

inflammation.

Knee swelling was reduced with either monoclonal

anti-IL17A or polyclonal anti-IL-17RA therapy;

ankles and knees were free of histopathological

changes following either therapy [19]. Lastly, rat

adjuvant arthritis models treated with IL17RA-Ig

demonstrated decreased paw swelling, joint histo-

pathology scores, and bone radiographic scores [20].

The above reports describe the effect of IL-17A

antagonists on reducing swollen joint incidence and

the resulting joint histopathology. Since outward joint

swelling is directly correlated with joint histology

(at least in the CIA model), it was unclear whether the

effect of the IL-17A antagonist on joint histology was

due solely to reducing the number of swollen paws or

to some aspect of disease-modifying, anti-rheumatic

effect. Our studies were undertaken to specifically

address this question by using two models of severe

arthritis that cause severe bone pathology (rAIA and

mCIA).

IL-17A neutralization was sufficient to alleviate the

disease severity measured by paw swelling and it

completely blocked bone erosion in the rAIA model

(Figure 1). The joint-preserving efficacy correlated

mechanistically with decreased RANKL expression in

the arthritic paw following short-term anti-IL-17A

treatment. Similarly, markers of osteoclast numbers

such as TRAP and CatK were reduced by IL17A

neutralization. These reduced markers were still

elevated in relation to naive mice as expected, since

other proinflammatory cytokines such as TNF and

IL1a/b induce TRAPþ cells that exhibit osteoclastic

characteristics in inflammatory arthritis [21–23].

The B10.RIII mCIA model was also chosen due to

its high disease penetrance and to its high cartilage and

bone pathology properties. One property of the model

was that joint inflammation was asynchronous within

each mouse, which necessitated tracking disease in a

paw-specific (Figure 2C) vs. animal-specific

(Figure 2B) manner in order to delve into the details

of the therapeutic potential of anti-IL-17A to modify

the joint erosive processes.

IL-17A blockade slightly delayed the onset of joint

swelling (Figure 2A), potentially due to the late

therapeutic dosing scheme employed in this report.

Collagen-specific T cell responses and T-cell depen-

dent anti-CII antibody responses had already been

developed and were interacting to initiate disease at

the time of starting anti-IL-17A treatment. As

described previously, the number of paws that became

involved and the severity of the swelling response were

reduced by IL17A blockade.

Since the magnitude of paw swelling was directly

correlated with the underlying joint pathology

(Supplementary Figure S1), the anti-IL-17A therapy,

which reduced the frequency and severity of paw

swelling, should be reflected in reduced joint histology

scores. The expected score reduction was obtained for

leukocyte recruitment, pannus formation, cartilage

destruction, and bone erosion (Figure 3). Previous

reports that analyzed IL-17A blockade in a conven-

tional manner, however, did not allow further

investigation of anti-IL-17A’s disease-modifying

potential within an inflamed joint.

The severely inflamed joints from control-treated

mice showed fairly uniform high histology scores for

pannus formation, cartilage destruction, and bone

erosion (Figure 4). Conversely, mice with anti-IL-17A

therapy that had the same “clinical” risk factor for

joint destruction (i.e., the few severely swollen paws)

fared much better. Anti-IL-17A’s disease-modifying

effect was apparent regardless of which histology

parameter was examined: pannus formation, cartilage

destruction, or bone erosion.

The joint-protective effect could not be ascribed to

modulation of leukocyte recruitment, since non-

statistical significant effect of anti-IL17A on this

parameter in severely swollen paws was uncovered. To

our knowledge, these data are the first description of

anti-IL-17A therapy that demonstrated the potential

to modify joint disease. Previous reports had failed to

separate the ability of various IL-17A antagonists to

delay the onset of paw swelling with any joint-

preserving function, if and when a paw developed

outward signs of arthritis. We have removed the

prominent effect of anti-IL-17A on arthritis incidence

by post hoc analysis of the few severely swollen paws

that developed in the presence of anti-IL-17A.

– + – + – +0

100

200

300

DSS = 0 DSS = 1 DSS = 3

p = 0.001

Anti-IL-17A

RA

NK

-L (

norm

aliz

ed to

Ub)

Figure 5. Modulation of RANKL expression in arthritic paws in

mCIA mice after IL-17 therapy. DSS ¼ 0, DSS ¼ 1, and DSS ¼ 3

paws from anti-IL-17A and isotype-treated mice were harvested,

frozen, and mRNA isolated. RANKL message normalized to

ubiquitin levels was determined. Statistical analysis was performed

by non-parametric Mann–Whitney analysis. “ 2 ” and “ þ ”

symbols represent isotype and anti-IL-17A-treated groups,

respectively.


Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Lib

rary

Utr

echt

on

08/0

1/13

For

pers

onal

use

onl

y.

In addition to using visual scores, we have

compared similarly infiltrated joints with bone

erosion, cartilage destruction, and pannus formation

by histology scores (Supplementary Figure S2). Our

analysis shows that anti-IL-17A greatly protects bone

erosion where the paws experience different degrees of

cellular infiltration from mild “score 1” to severe

“score 4” (Supplementary Figure S2A). Anti-IL-17A

inhibits cartilage destruction where paws have a mild

score (2), but do not have severe cellular infiltration.

The beneficial effect of anti-IL-17A also inhibits

pannus formation where mild (score 2) and severe

(score 3) infiltrates occur.

The therapeutic benefit of anti-IL-17A therapy was

observed when arthritic mice were dosed coincident

with arthritis development (Figure 2A–C). Further-

more, anti-IL-17A therapy prevented arthritis pro-

gression and lowered RANKL expression (Figures 1D

and 5), even when rats (Figure 1) and mice

(Figure 2D–E) were dosed after severe arthritis was

established. These results have significant impli-

cations for the therapeutic potential of IL17A

antagonism in RA. IL-17A is not required for arthritis

initiation (IL-17A 2 /2 are susceptible to CIA,

although less so than wild-type controls) [24], and

therapeutic IL-17A neutralization does not absolutely

halt disease in the highly joint-destructive B10.RIII

CIA model.

Similar conclusions are also drawn for TNF, since

TNF 2 /2 mice are susceptible to arthritis [25] and

therapeutic TNF antagonism does not stop, but

diminishes, joint-swelling responses [26,27]. IL-17A

blockade has two mechanisms to inhibit joint

histology. The first is via the conventional concept of

inhibiting joints from becoming inflamed. The

analogous human RA scenario would be using anti-

IL17A as front-line therapy in “early RA” patients

who have risk factors that predict that destructive joint

disease will quickly ensue in the initial inflamed joint

and will spread to other joints. A second, independent

mechanism by which anti-IL17A inhibits joint

pathology occurs in joints that are already inflamed.

The analogous human RA scenario would be in

patients who have inadequate or no response to the

current gold-standard TNF antagonists.

It should be noted that anti-IL17A may have joint

protective benefits, even in patients where the outward

signs of arthritis are not modulated by IL-17A

antagonism. This concept has some similarities to

Denosumab/Anti-RANKL, which has profound bone

protective effects, but no effects on the outward

systemic signs of arthritis. Anti-IL17A therapy differs

from RANKL blockade in animal models, because

neutralizing RANKL in treated arthritic animals has

no effect on joint swelling and minimal cartilage

protection. Therefore, anti-IL-17A is both anti-

inflammatory and joint-preserving, whereas, anti-

RANKL is only the latter.

Chronic arthritis in RA can cause peri-articular

bone erosion due to an imbalance in bone erosion and

bone formation. In these inflammatory arthritic

models, we have also found that new bone formation

along with bone erosion was displayed in the cortical

surface of different areas. The higher activity of bone

turnover frequently occurs in the areas of the bones

having more severe erosion. Since anti-IL-17A

treatment inhibits bone loss, less activity of bone

remodeling and new bone formation were observed in

the anti-IL-17A-treated paws. Bone erosion is not a

purely resorptive process, but a complex interaction of

bone resorption and formation. An interplay between

cellular (osteoclasts and osteoblasts) and molecular

(RNAKL, BMPs/Wnt) levels, and a net outcome of

bone remodeling will be further investigated in our

models [28].

In summary, treatment with anti-IL-17A comple-

tely alleviates arthritis, decreases inflammation-driven

RANKL levels, and provides protection in the bone-

erosive rAIA model. Furthermore, therapeutic

IL-17A neutralization after CII challenge diminishes

the DSS by inhibiting the number of paws that

become inflamed in the mCIA model. Unexpectedly,

even paws that became outwardly severely inflamed in

the presence of an IL-17A antagonist had diminished

joint histopathology scores compared to severely

inflamed control-treated mice. No evidence was

obtained that anti-IL-17A therapy altered T cell

cytokine response or diminished B-cell anti-collagen

responses (date not shown). IL-17A blockade rep-

resents a novel approach to attack the underlying

pathomechanism that initiates and perpetuates

inflammatory joint disease.

Declaration of interest: All authors were employees

of Schering-Plough (now Merck) during the time of

these studies, there are no other conflicts of interest.

References

[1] FossiezF,DjossouO,ChomaratP,Flores-RomoL,Ait-YahiaS,

Maat C, Pin JJ, Garrone P, Garcia E, Saeland S, Blanchard D,

Gaillard C, Das Mahapatra B, Rouvier E, Golstein P,

Banchereau J, Lebecque S. T cell interleukin-17 induces

stromal cells to produce proinflammatory and hematopoietic

cytokines. J Exp Med 1996;183:2593–2603.

[2] McGeachy MJ, Cua DJ. The link between IL-23 and Th17

cell-mediated immune pathologies. Semin Immunol 2007;19:

372–376.

[3] Kawaguchi M, Adachi M, Oda N, Kokubu F, Huang SK.

IL-17 cytokine family. J Allergy Clin Immunol 2004;114:

1265–1273; quiz 1274.

[4] Hwang SY, Kim HY. Expression of IL-17 homologs and their

receptors in the synovial cells of rheumatoid arthritis patients.

Mol Cells 2005;19:180–184.

[5] Honorati MC, Meliconi R, Pulsatelli L, Cane S, Frizziero L,

Facchini A. High in vivo expression of interleukin-17 receptor

in synovial endothelial cells and chondrocytes from arthritis

patients. Rheumatology (Oxford) 2001;40:522–527.


Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Lib

rary

Utr

echt

on

08/0

1/13

For

pers

onal

use

onl

y.

[6] Joosten LA, Radstake TR, Lubberts E, van den Bersselaar LA,

van Riel PL, van Lent PL, Barrera P, van den Berg WB.

Association of interleukin-18 expression with enhanced levels

of both interleukin-1beta and tumor necrosis factor alpha in

knee synovial tissue of patients with rheumatoid arthritis.

Arthritis Rheum 2003;48:339–347.

[7] Page G, Sattler A, Kersten S, Thiel A, Radbruch A, Miossec P.

Plasma cell-like morphology of Th1-cytokine-producing cells

associated with the loss of CD3 expression. Am J Pathol 2004;

164:409–417.

[8] Aarvak T, Chabaud M, Miossec P, Natvig JB. IL-17 is

produced by some proinflammatory Th1/Th0 cells but not by

Th2 cells. J Immunol 1999;162:1246–1251.

[9] Ziolkowska M, Koc A, Luszczykiewicz G, Ksiezopolska-

Pietrzak K, Klimczak E, Chwalinska-Sadowska H,

Maslinski W. High levels of IL-17 in rheumatoid arthritis

patients: IL-15 triggers in vitro IL-17 production via

cyclosporin A-sensitive mechanism. J Immunol 2000;164:

2832–2838.

[10] Chabaud M, Durand JM, Buchs N, Fossiez F, Page G,

Frappart L, Miossec P. Human interleukin-17: A T cell-

derived proinflammatory cytokine produced by the rheuma-

toid synovium. Arthritis Rheum 1999;42:963–970.

[11] Kotake S, Udagawa N, Takahashi N, Matsuzaki K, Itoh K,

Ishiyama S, Saito S, Inoue K, Kamatani N, Gillespie MT,

Martin TJ, Suda T. IL-17 in synovial fluids from patients with

rheumatoid arthritis is a potent stimulator of osteoclastogen-

esis. J Clin Invest 1999;103:1345–1352.

[12] Raza K, Falciani F, Curnow SJ, Ross EJ, Lee CY, Akbar AN,

Lord JM, Gordon C, Buckley CD, Salmon M. Early

rheumatoid arthritis is characterized by a distinct and transient

synovial fluid cytokine profile of T cell and stromal cell origin.

Arthritis Res Ther 2005;7:R784–R795.

[13] Kirkham BW, LassereMN, Edmonds JP, Juhasz KM, Bird PA,

Lee CS, Shnier R, Portek IJ. Synovial membrane cytokine

expression is predictive of joint damage progression in

rheumatoid arthritis: A two-year prospective study (the

DAMAGE study cohort). Arthritis Rheum 2006;54:

1122–1131.

[14] Mori H, Kitazawa R,Mizuki S, NoseM,Maeda S, Kitazawa S.

RANK ligand, RANK, and OPG expression in type

II collagen-induced arthritis mouse. Histochem Cell Biol

2002;117:283–292.

[15] Gravallese EM, Manning C, Tsay A, Naito A, Pan C,

Amento E, Goldring SR. Synovial tissue in rheumatoid

arthritis is a source of osteoclast differentiation factor. Arthritis

Rheum 2000;43:250–258.

[16] Lubberts E, Joosten LA, Oppers B, van den Bersselaar L,

Coenen-de Roo CJ, Kolls JK, Schwarzenberger P, van de Loo

FA, van den Berg WB. IL-1-independent role of IL-17 in

synovial inflammation and joint destruction during collagen-

induced arthritis. J Immunol 2001;167:1004–1013.

[17] Lubberts E, Koenders MI, Oppers-Walgreen B, van den

Bersselaar L, Coenen-de Roo CJ, Joosten LA, van den

Berg WB. Treatment with a neutralizing anti-murine inter-

leukin-17 antibody after the onset of collagen-induced arthritis

reduces joint inflammation, cartilage destruction, and bone

erosion. Arthritis Rheum 2004;50:650–659.

[18] Koenders MI, Lubberts E, Oppers-Walgreen B, van den

Bersselaar L, HelsenMM,Di Padova FE, Boots AM,GramH,

Joosten LA, van den Berg WB. Blocking of interleukin-17

during reactivation of experimental arthritis prevents joint

inflammation and bone erosion by decreasing RANKL and

interleukin-1. Am J Pathol 2005;167:141–149.

[19] Burchill MA, Nardelli DT, England DM, DeCoster DJ,

Christopherson JA, Callister SM, Schell RF. Inhibition of

interleukin-17 prevents the development of arthritis in

vaccinated mice challenged with Borrelia burgdorferi. Infect

Immun 2003;71:3437–3442.

[20] Bush KA, Farmer KM, Walker JS, Kirkham BW. Reduction of

joint inflammation and bone erosion in rat adjuvant arthritis by

treatment with interleukin-17 receptor IgG1 Fc fusion protein.

Arthritis Rheum 2002;46:802–805.

[21] Lam J, Takeshita S, Barker JE, Kanagawa O, Ross FP,

Teitelbaum SL. TNF-alpha induces osteoclastogenesis by

direct stimulation of macrophages exposed to permissive levels

of RANK ligand. J Clin Invest 2000;106:1481–1488.

[22] Nakamura I, Kadono Y, Takayanagi H, Jimi E, Miyazaki T,

Oda H, Nakamura K, Tanaka S, Rodan GA, Duong LT. IL-1

regulates cytoskeletal organization in osteoclasts via TNF

receptor-associated factor 6/c-Src complex. J Immunol 2002;

168:5103–5109.

[23] Adamopoulos IE, Sabokbar A, Wordsworth BP, Carr A,

Ferguson DJ, Athanasou NA. Synovial fluid macrophages are

capable of osteoclast formation and resorption. J Pathol 2006;

208:35–43.

[24] Nakae S, Nambu A, Sudo K, Iwakura Y. Suppression of

immune induction of collagen-induced arthritis in IL-17-

deficient mice. J Immunol 2003;171:6173–6177.

[25] Campbell IK, O’Donnell K, Lawlor KE, Wicks IP. Severe

inflammatory arthritis and lymphadenopathy in the absence of

TNF. J Clin Invest 2001;107:1519–1527.

[26] Joosten LA, Helsen MM, van de Loo FA, van den Berg WB.

Anticytokine treatment of established type II collagen-induced

arthritis in DBA/1 mice. A comparative study using anti-TNF

alpha, anti-IL-1 alpha/beta, and IL-1Ra. Arthritis Rheum

1996;39:797–809.

[27] Butler DM, Malfait AM, Maini RN, Brennan FM, Feldmann

M. Anti-IL-12 and anti-TNF antibodies synergistically

suppress the progression of murine collagen-induced arthritis.

Eur J Immunol 1999;29:2205–2212.

[28] Zwerina J, Tuerk B, Redlich K, Smolen JS, Schett G.

Imbalance of local bone metabolism in inflammatory arthritis

and its reversal upon tumor necrosis factor blockade: Direct

analysis of bone turnover in murine arthritis. Arthritis Res

Ther 2006;8:R22.


Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Lib

rary

Utr

echt

on

08/0

1/13

For

pers

onal

use

onl

y.

Documents

Anti-IL-17A therapy protects against bone erosion in experimental models of rheumatoid arthritis