Sequential analysis of kidney stone formation in the Aprt knockout mouse

Kidney International, Vol. 60 (2001), pp. 910–923

Sequential analysis of kidney stone formation in theAprt knockout mouse

ANDREW P. EVAN, SHARON B. BLEDSOE, BRET A. CONNORS, LI DENG, LI LIANG,CHANGSHUN SHAO, NAOMI S. FINEBERG, MARC D. GRYNPAS, PETER J. STAMBROOK,SHAO YOUZHI, AMRIK SAHOTA, and JAY A. TISCHFIELD

Departments of Anatomy and Cell Biology and Medicine, Indiana University School of Medicine, Indianapolis, Indiana andDepartment of Genetics, Rutgers University, Piscataway, New Jersey, USA; Department of Pathology and Laboratory Medicine,University of Toronto, Ontario, Canada; Department of Cell Biology, Neurobiology, and Anatomy, University of CincinnatiCollege of Medicine, Cincinnati, Ohio, USA; and Department of Histology, Jinzhou Medical College, Jinzhou, Liaoning,People’s Republic of China

crystals were spherical, with a diameter of 10 to 20 �m. TissueSequential analysis of kidney stone formation in the Aprt knock-staining and fixation procedures dramatically reduced theout mouse.amount of birefringent material in kidney sections. Aprt hetero-Background. We have previously shown that, as in humanzygotes of both sexes had low levels of crystalline material inadenine phosphoribosyltransferase (APRT) deficiency, Aprtthe kidneys and no pathology. Birefringent material or patho-knockout mice form 2,8-dihydroxyadenine (DHA) renal stones.logical changes were not seen in kidneys from wild-type mice.The disease develops earlier and is more severe in male than

Conclusions. Both male and female Aprt knockout micein female mice. To examine the biological bases for these differ-accumulate DHA. However, the area occupied by DHA crys-ences, the area occupied by DHA crystals was quantified intals was significantly greater in 120- to 240-day-old males com-kidney sections from male and female mice (strain 129) agedpared with the females of similar age. Also, substantial renalone day to eight months and this parameter was correlatedpathology was detected in kidneys of male mice that had verywith changes in renal histopathology. Aprt heterozygous and

wild-type mice were used as controls. high levels of stone material.Methods. Following anesthesia, the left kidney was removed

and immediately frozen in dry ice. Unstained cryosections wereexamined by polarized light to determine total area of birefrin-

Adenine phosphoribosyltransferase (APRT, EC 2.4.2.7)gent particles. The right kidney was perfused and embeddedis a purine salvage enzyme that catalyses the Mg2�-depen-in plastic, and stained sections were viewed by light microscopy

to examine the histopathology and to determine the location dent synthesis of 5�-adenosine monophosphate (AMP)of the birefringent particles. A pathological score was assigned from adenine and 5-phosphoribosyl-1-pyrophosphate. Into the histological findings. The scores from the right kidney APRT deficiency (MIM 102600), a rare autosomal-reces-were compared with crystal/particle area in the left kidney, and

sive disorder of purine metabolism, a fraction of thethe data were analyzed using two-way analysis of variance. Theadenine is oxidized by xanthine dehydrogenase (XDH,chemical composition of the particles was determined by x-ray

diffraction analysis. Several stone fragments from the bladder EC 1.2.1.37) to 2,8-dihydroxyadenine (DHA). This com-were also examined by scanning electron microscopy (SEM). pound is extremely insoluble at any urinary pH, and this

Results. Crystals were detected in kidney sections from one- insolubility can lead to its precipitation and crystalliza-to two-day-old Aprt knockout mice. The crystal burden re-tion in renal tubules as well as other components ofmained low in both sexes throughout the study except in malesthe urinary tract. DHA crystals can aggregate to format the 120- to 240-day period. Furthermore, there was a substan-

tial degree of renal pathology, primarily seen as interstitial fibro- macroscopically visible calculi, especially in the bladder.sis, in those males with a very high level of stone formation. The The major features of APRT deficiency are related tocrystalline material was identified as 6-amino-2,8(3,9)-purine DHA urolithiasis, but acute or chronic renal failure hasdione, a tautomeric form of DHA. SEM indicated that the

occurred in some patients. DHA formation and the re-sulting renal damage can be effectively controlled byallopurinol, an XDH inhibitor. Recent reviews of APRTKey words: adenine phosphoribosyltransferase deficiency, renal histo-

pathology, crystals, DHA renal stones, kidney stones, urolithiasis. deficiency are in [1, 2].Histological examination of tissue obtained at renalReceived for publication August 16, 2000

biopsy or nephrectomy has been reported in a smalland in revised form February 20, 2001Accepted for publication March 16, 2001 number of APRT-deficient patients [2–8]. These studies

showed intratubular crystal blockage, tubular degenera- 2001 by the International Society of Nephrology

910

Evan et al: Kidney stone formation in Aprt knockout mouse 911

tion, inflammation, and in some cases, glomerulosclerosis. METHODSThere was extensive crystalline precipitate in the tubular Micelumens and epithelia, especially in the cortex, with crys-

The Aprt knockout mice were originally generated intals tending to form aggregates within the tubules. Amor-a mixed background of 129/Sv embryonic stem cells andphous deposits in the interstitium were surrounded byC57BL/6J blastocyst donors [10]. The chimeras werevarying numbers of multinucleated giant cells, neutro-bred into strain 129/Sv to generate heterozygotes, whichphils, and lymphocytes. Associated with the DHA pre-were then intercrossed to produce homozygous-deficientcipitates were areas of interstitial nephritis, interstitialmice. Male and female homozygous-deficient and het-and peritubular fibrosis, patchy interstitial inflammation,erozygous mice within the age range of one day to eightand tubular atrophy. Polarizing microscopy revealedmonths were used in this study. A small number of three-many individual birefringent crystals scattered through-month-old wild-type mice were also examined. The Aprtout the renal parenchyma [7], and analysis of DHA stonesgenotypes were determined by PCR analysis of DNAwith the increased magnification of scanning electron mi-isolated from tail biopsies, as previously described [15].croscopy (SEM) showed a spherolite surface with matted

needle-shaped crystals [9]. Tissue harvestingAprt-deficient mice have been produced recently us-

Animals aged one month and over were anesthetizeding targeted homologous recombination in embryonicwith a single intraperitoneal injection of sodium pento-stem cells and subsequent breeding of heterozygous off-barbital (5 mg/100 g body weight), and the anterior ab-spring [10–12]. Unlike a number of other mouse modelsdominal wall was opened from the xiphoid process tofor purine metabolic disease, for example, the hypoxan-the pubic symphysis. Subsequently, the vessels at the renalthine-guanine phosphoribosyltransferase deficient mousehilus were clamped; the left kidney was quickly removed,[13], the Aprt-deficient mouse appears to mimic the cor-frozen in powdered dry ice, and stored at �80�C. Theresponding human disease. These animals deposit DHAanimal was then perfused through the left ventricle withcrystals and stones within the kidney, excrete DHA and30 mL of heparinized saline followed by 120 mL of 2.5%adenine in the urine, and form DHA bladder stones,glutaraldehye in 0.1 mol/L sodium cacodylate buffer (pHand the most severely affected mice die by approximately7.4) at room temperature. Next, the right kidney wassix months of age. Allopurinol treatment prevented theremoved and placed in a vial containing about 20 mLaccumulation of DHA and much of the resultant renalfresh fixative and stored at 4�C for future processing indamage in Aprt-deficient mice [12].JB-4 resin (Polysciences, Inc., Warrington, PA, USA) forWe have bred the mutant Aprt allele into several dif-light microscopic analysis. Animals under one week wereferent mouse strains, including C57BL6, C3H, 129, andnot perfused; the left kidney was frozen as described pre-Black Swiss. The severity of symptoms of Aprt deficiencyviously in this article, but the right kidney was placedin mice appears to be strain, age, and sex dependent,directly in fixative. Both the right and left kidneys werewith C57 males being the most severely affected. For aassigned a code number so that all future measurementsgiven strain, male mice show significantly more renaland scoring would be conducted as a blind analysis.damage than females of the same age. Recent differential

display and reverse transcription-polymerase chain reac-Cryosectioning, staining, and imagingtion (RT-PCR) studies from our laboratory also have

The frozen left kidney was embedded in OCT com-demonstrated a gender bias in the expression of severalpound (Miles Laboratories, Elkhart, IN, USA), and 7 �mgenes in the kidney, with these changes being most no-sections were cut in the midtransverse plane on a Reichert-ticeable in three-month-old male mice, an age when his-Jung cryostat in order to preserve the birefringent mate-tological damage becomes most obvious [14].rial located within the kidney [11]. Sections were pickedOur present study quantified the crystal burden (totalup on charged slides and kept frozen until fixed in ace-particle area) in Aprt knockout male and female micetone for 15 to 20 seconds. All odd-numbered slides wereof different ages and correlated these values with thedehydrated in one change of 100% ethanol, followed byextent of pathological changes in the kidneys. Animalstwo changes of Hemo-De clearing agent (Fisher Scien-of strain 129 were used because the disease severity intific, Pittsburgh, PA, USA) before coverslipping the un-these mice was less pronounced (compared to C57); thus,stained sections. The even-numbered slides were stainedsufficient Aprt-deficient animals could be generated bywith 1% toluidine blue, dehydrated in a series of gradedmating heterozygotes. We found that DHA accumula-ethanols, rinsed in Hemo-De, and then coverslipped [11].tion occurred in both male and female mice, but crystal

The protocol for crystal quantitation was performedburden was significantly higher in males 120 to 240 dayson four unstained cryosections from each kidney. Theof age. Also, substantial renal pathology was detectedsections were viewed by polarized light using a Leicain the kidneys of male mice that had very high levels of

stone material. DMR microscope and scanned with a Polaroid DMC

Evan et al: Kidney stone formation in Aprt knockout mouse912

digital camera (resolution set at 1600 � 1200 pixels) that with an exposure time of two hours at 40 kV and 25 mA.The resulting diffraction pattern was searched manuallywas linked to a Macintosh Power PC 8100 and operated

remotely by an Adobe Photoshop plug-in module. All in the Powder Diffraction file of organic compound of theInternational Center for Diffraction Data (Swarthmore,settings for the use of the camera and microscope were

standardized prior to starting the analysis. To measure PA, USA).In addition to x-ray diffraction analysis, standard SEMaccurately the crystal material (DHA) only within the

renal parenchyma, all extraparenchymal regions were protocols were used to examine several stones collectedfrom the bladder of a homozygous-deficient animal [17].eliminated after manually tracing the renal capsule and

renal sinus boundaries. The accuracy of the tracing was The amount of stone material from the heterozygotewas too small to be analyzed by SEM as well, but allconfirmed by viewing the alternate stained sections. The

images were then saved as TIFF files so that they could stone materials were examined by light microscopy.be analyzed with NIH Image software. The birefringent

Data analysisDHA crystals were identified by thresholding, and thetotal particle area was calculated. The values were ex- The data were expressed as mean � SD. The values

for particle area (skewness � 7.0 � 0.2) and path scoreported in Microsoft Excel, and an average was deter-mined for each set of four values representing the four (skewness � 1.8 � 0.3) were skewed to the right. There-

fore, all analyses were done using a log transformationindividual sections for each kidney. The raw values forparticle area were converted from square pixels to square for particle area and a square root transformation for

path score. Two-way analysis of variance (ANOVA) wasmillimeters using this formula: 1 square pixel � 0.037mm2. The values for total particle area listed in the tables used to examine the effects of age group, gender, or Aprt

genotype on these parameters. Three statistical compari-are expressed as square millimeters per section.sons were carried out using this approach: (1) effect of

JB-4 embedment and pathological scoring age group pooled over gender, (2) effect of gender pooledover age group, and (3) an interaction effect. A significantA series of 2 �m sections was cut from each right

kidney at the midtransverse plane using a Sorval JB-4 interaction term implied that the gender effect differedamong the age groups. When only one main effect wasmicrotome, mounted on charged slides, and then stained

with Lee’s methylene blue before coverslipping. These significant and the other main effect and interaction termwere not statistically significant, data were pooled acrosssections were viewed under a light microscope to locate

the birefringent material seen in the cryosections in order the nonsignificant main effect and one-way ANOVA andthe least significant difference test were used to determineto quantify the birefringent material left after processing

the tissue under staining conditions and to determine a which subgroups were different. Linear regressions wereused to examine relationships between variables. Thesemiquantitative pathology score for each right kidney.

To assess a semiquantitative value for the type and criteria for statistical significance were set at P 0.05.degree of renal abnormalities present in the kidneys ofhomozygous-deficient and heterozygous animals, two

RESULTScoded sections (at least 100 �m apart) from each kidney

Morphology and x-ray analysis of crystalswere analyzed. The evaluation of renal pathology wasperformed in a blind fashion and was performed by two By routine light microscopy, DHA crystals were found

within the lumens of cortical and medullary collectingdifferent investigators as a means of reducing the subjec-tivity of this approach. The pathology grading scale as- ducts of Aprt-deficient male and female mice as early

as one to two days after birth (Fig. 1a). By two weekssigned value between 0 to 5, with 0 indicating no observ-able pathological abnormalities and 5 indicating severe of age, DHA crystals were seen in the lumens of proximal

and distal tubules as well as collecting ducts (Fig. 1b).abnormalities or disease [16]. The pathology scores ob-tained by the two investigators (listed as path scores 1 The crystals were easily identified by their brownish red

color and were generally in spherical shape (about 10 toand 2 in Table 1) were analyzed to determine the degreeof variability between the investigators. 20 �m in diameter), as seen in the kidney from a three-

month-old male (Fig. 1 c, d). At a higher magnificationX-ray diffraction protocols (Fig. 1d), DHA crystals clearly had a distinct substruc-

ture in the form of numerous slender needle- to lancet-The chemical composition of the stone material foundwithin the renal pelvis of a heterozygote and the bladder shaped crystals that radiated from the center of the larger

spherical crystal. With time, interstitial fibrosis was seenof homozygous-deficient mice was determined using es-tablished x-ray diffraction techniques. Briefly, the stones in both the male and female kidneys, with the males

showing much greater change (Table 1). In the regionswere finely powdered and placed in a glass capillaryfor examination in an x-ray diffraction Debye-Sherrer of interstitial fibrosis, DHA crystals were also found

surrounding atrophic cells primarily of the S3 segmentpowder camera using Cu K (alpha) radiation (1.5418 A)


Fig. 1. Morphology of 2,8-dihydroxyadenine (DHA) crystals by routine bright field (a–e) or polarizing optics ( f ) in kidneys from adenine phosphoribo-syltransferase (Aprt)-deficient mice. (a) The location of crystals only in the collecting ducts (arrow) of one- to two-day-old mice. By one to twoweeks of age, both male and female mice showed crystals in proximal and distal tubules and collecting ducts (b, arrow). (c and d) From a three-month-old male with a pathology score of 3. (c) Two proximal tubules (PT, arrows) and a distal tubule (DT, arrow), each with large brownish-red crystals in the tubular lumens. Evidence of interstitial fibrosis is seen around glomeruli and various tubular segments. (d) A higher magnificationof a large crystal in the lumen of a proximal tubule to show the substructure of these crystals is shown. Note the numerous slender needle- tolancet-shaped structures that radiate from the center of the crystal. (e and f) The same field from a three-month-old male mouse. The bright-fieldphotomicrograph (e) shows a number of atrophic cells in the S3 segment of a proximal tubule encased by slender, brownish red crystals (arrows),whereas the birefringent pattern of the same crystals is readily seen under polarizing optics (f, arrows). (a and b) �350; (c) �250; and (d–f) �1700.


of the proximal tubule (Fig. 1 e, f), as well as atrophiccells of the distal tubule and collecting ducts. The crystalswere birefringent under polarizing optics regardless oftheir location in the tubular lumen or the interstitialspace (Figs. 1f and 3a). The DHA crystals surroundingthe atrophic tubular segments were not spherical, butinstead were composed of numerous slender rod- to nee-dle-shaped structures.

Scanning electron microscopy visualization of the DHAstone material in the tubular lumens of nephrons and/orcollecting ducts of homozygous-deficient animals clearlyshowed the spherical shape (�10 to 20 �m in diameter)of each individual crystal (Fig. 2 a, b). Occasionally, severalcrystals were aggregated within the tubular lumen, form-ing a much larger stone (data not shown). At a highermagnification (Fig. 2b), each individual crystal was foundto have a substructure composed of numerous needle-to lancet-shaped crystals. Once the individual sphericalcrystals passed into the renal pelvis (Fig. 2c) and then theurinary bladder (Fig. 2d), extensive aggregation occurredmaking such stones visible to the naked eye. The individ-ual DHA crystals when found in the urinary bladderretained their spherical shape (Fig. 2e) and their compo-sition of numerous needle-shaped substructures (Fig. 2f).

The chemical composition of the stone material foundwithin the renal pelvis of a heterozygous animal and theurinary bladder of the homozygous-deficient animals wasdetermined by x-ray diffraction analysis as C5H6N5O2 [6-amino-2,8(3,9)-purine dione], a tautomeric form of DHA.

Effects of tissue staining and fixation oncrystal quantification

Figure 3 illustrates the dramatic effect of tissue fixationand staining on the amount of DHA birefringent mate-rial detected within the mouse kidney. Alternate un-stained (Fig. 3 a, b) and stained (Fig. 3 c, d) sectionswere taken from the snap-frozen left kidney, while plasticsections (Fig. 3 e, f) were taken from the perfused rightkidney of the same mouse. Tissue that was frozen andunstained provided the highest yield of DHA crystals,whereas frozen and stained sections exhibited a substan-tial loss of material during the staining and subsequentdehydration steps. Because of the processing steps in-volving dehydration and the high heat produced from thepolymerization of the JB-4 plastic, most of the crystallinematerial appears to have been lost from the perfusedkidney.

National Institutes of Health (NIH) imaging softwarewas used to quantify the amount of DHA birefringentmaterial seen in Figure 3 b, d, and f. Frozen unstainedsections exhibited a total particle area of 169.9 mm2 (Fig.3b). Frozen, toluidine blue-stained sections had a loss ofapproximately 44% of the birefringent material and a totalparticle area of 94.9 mm2 (Fig. 3d). The JB-4-embedded

Tab

le1.

Pat

holo

gysc

ores

and

DH

Acr

ysta

lqu

anti

fica

tion

inA

prt

hom

ozyg

ous-

defi

cien

tm

ale

and

fem

ale

mic

eof

diff

eren

tag

egr

oups

Age

days

1to

27

to14

30to

4560

to90

120

to24

0

Fem

ale

Mal

eF

emal

eM

ale

Fem

ale

Mal

eF

emal

eM

ale

Fem

ale

Mal

eP

valu

e

Pat

holo

gysc

ore

10.

12�

0.16

0.10

�0.

070.

88�

0.61

0.85

�0.

860.

88�

0.62

0.86

�0.

480.

50�

00.

67�

0.41

0.74

�0.

652.

51�

1.47

cP

atho

logy

scor

e2

0.09

�0.

090.

10�

0.0

0.62

�0.

550.

55�

0.66

0.53

�0.

410.

47�

0.34

0.15

�0.

070.

33�

0.21

0.44

�0.

502.

39�

1.68

cT

otal

part

icle

area

mm

219

3.5

�10

4.7

289.

8�

166.

863

3.5

�61

5.9

960.

1�

213.

561

0.6

�23

3.9

660.

6�

283.

231

3.5

�82

.583

4.3

�44

2.9

1126

.9�

2191

.236

35.6

�51

86.7

abN

95

64

67

26

77

Dat

aar

eex

pres

sed

asm

ean

�SD

.Abb

revi

atio

nsar

ein

the

App

endi

x.aO

vera

llva

lues

for

mal

esar

esi

gnif

ican

tly

grea

ter

than

fem

ales

bytw

o-w

ayA

NO

VA

bV

alue

sin

crea

sesi

gnif

ican

tly

wit

hag

eby

two-

way

AN

OV

AcO

ne-

to2-

day-

old

anim

als

are

sign

ific

antl

ydi

ffer

ent

from

all

othe

rag

egr

oups

byon

e-w

ayA

NO

VA

and

the

leas

tsi

gnif

ican

tdi

ffer

ence

test

plastic sections from the right kidney exhibited a loss of


Fig. 2. Scanning electron microscopy (SEM) of DHA crystals and stones in the kidneys from Aprt-deficient mice. (a and b) Each show anindividual crystal in the lumen of a proximal tubule (a) and a cortical collecting duct (b), respectively, from a three-month-old male mouse. Thecrystal is spherical with a substructure of numerous needle-shaped projections. (c) A DHA stone (arrow) lodged between the renal papilla (asterisk)and the wall of the renal pelvis. The stone was composed of multiple individual spherical DHA crystals. The individual spherical crystals are easilyseen in an aggregated DHA stone obtained from the urinary bladder of a three-month-old male (d). At higher magnifications (e and f ), thesubstructure of the spherical crystal is seen to be composed of numerous needle- to lancet-shaped crystals. (a) �2000; (b) �4000; (c) �100; (d)�500; (e) �1200; and (f) �8000.


Fig. 3. Dramatic effect of tissue fixation and staining on the amount of DHA birefringent material detected within the kidney. Three differentlyprepared sections from the kidney of the same animal (five-month-old Aprt-deficient female) were viewed by both bright-field (a, c, and e) andpolarizing optics (b, d, and f ). (a and b) Unstained sections cut from the frozen kidney. There was a total particle area of 169.9 mm2 (b). (c andd) A toluidine blue-stained section cut from the same frozen kidney. There was a total particle area of 94.9 mm2 (d), a 44% decrease comparedwith (b). (e and f) A JB-4-embedded plastic section (2 �m thick) stained with Lee’s methylene blue. There was a total particle area of 5.3 mm2

(f), a 95% loss of crystal material compared with (b). (a–f) �23.


Fig. 4. Differences in DHA particle amount as a function of age and sex in kidneys from Aprt-deficient mice. (a, male, and b, female) Thedifference in particle area at age one to two days. (c, male, and d, female) The variation in particle area at age 120 days. (e, male, and f, female)The difference at age 120 days. There was a visible difference in the amount of crystalline material between the male (e) and the female (f), aswell as between the male and female in (e) and (f) as compared with the male and female in (c and d). (a–f) �23.


over 95% of crystal material, with a total particle areaof 5.3 mm2 (Fig. 3f). These data clearly show that therewas a substantial loss of crystalline material during thedehydration and processing steps.

Crystal quantification in homozygous-deficient mice

The pathological changes in Aprt-deficient mice havebeen described previously [10–12, 14], and these wereconfirmed and extended in the present study. The pri-mary pathological change was interstitial fibrosis, whichextended from the renal capsule to the papilla. Otherchanges included (1) crystal deposition within lumens ofproximal and distal tubules, collecting ducts, as well asaround atrophic tubular segments; (2) dilation and atro-phy of primarily the proximal tubules; (3) leukocyteinfiltration around injured tubular segments; (4) perivas-cular cuffing with interstitial infiltrates; (5) loss of post-glomerular capillaries; and (6) dilation of Bowman’sspace and presence of flocculent material [14]. Fig. 5. Square root of the pathology score as a function of the log of

particle area. The square root of the pathology scores was plottedTable 1 lists the values for total particle area for theagainst the log of all particle areas and separate linear regressions werebirefringent DHA crystals and pathology scores deter- calculated for Aprt-deficient male (�) and female (�) mice. Females:

mined in the kidneys of Aprt-deficient male and female f(x) � 0.335X � 0.223, R2 � 0.108; Males: f(x) � 1.738X � 3.864, R2 �0.528.mice of five different age groups. Figure 4 clearly demon-

strates this for male and female mice aged 120 days (Fig.4 e, f) compared with male and female mice aged oneto two days (Fig. 4 a, b). In addition, while there was compares the degree of pathological change (pathologysome variation in the values for the birefringent particles

scores of 1, 2, and 4) with crystal area (246.7, 1261.2,between animals (Fig. 4 c and d), overall, the values for

and 14,957.4 mm2) for three 120-day-old male mice.total particle area were significantly greater (P 0.05)in males compared with females when analyzed over the Crystal quantification in heterozygousentire time course of this study (Table 1). Particle size and wild-type micewas estimated to be approximately 500 square microns.

Initially, heterozygotes were used as controls in theseTable 1 also lists both of the pathology scores (1 andexperiments, as phenotypic appearance and gross renal2) obtained for both male and female mice at each ofhistology did not show any differences compared withthe five time points studied. While the pathology scoreswild-type animals [10]. However, some of the heterozy-determined by the second reader (pathology scores 2)gotes exhibited crystalline material in the renal pelviswere consistently lower than the first reader, the resultsat autopsy (Fig. 7). By x-ray diffraction analysis, thesefor the two readers were highly correlated (correlationcrystals were found to be identical to the DHA crystalsbetween readings: r � 0.95, P 0.001; paired t test be-found in homozygous-deficient mice. Table 2 lists thetween readings: t � 6.0, P 0.010). The average differencevalues for total particle area occupied by DHA crystalsbetween the two readings was 27.8% � 31.1%, whichin kidney sections from heterozygous males and femalesrepresents an actual difference in the two scores of 0.23 �at 7 to 14 and 30 to 45 days of age. A small amount of0.29. With regard to the histological changes, the one- toDHA crystal material was present in the kidneys of bothtwo-day-old animals (both male and female) were foundthe male and female heterozygous mice; however, thereto be significantly different from all other age groupswas no significant difference between the males and fe-(P 0.05; Table 1).males at either age group. Since all heterozygous animalsThe individual values for the square root of the pathol-had a pathology score of 0, analysis of the pathologyogy scores were plotted against the log of the total parti-score as a function of crystal burden was not carried out.cle areas for both homozygous-deficient males and fe-

The values for DHA crystal burden in heterozygousmales at each of the five time points (Fig. 5). Statisticalanimals (male and female combined) were comparedanalysis of these data showed a highly significant correla-with the corresponding values in homozygous-deficienttion in males (r � 0.73, P 0.001; Fig. 5), suggestingmice (male and female combined) of the same age groups.that substantial pathology is seen in those kidneys withThe values in homozygous-deficient animals were sig-a very high stone burden. Further support of this idea

is seen in Figure 6 where a series of light micrographs nificantly higher (Table 3 and Fig. 7). After finding crys-


Fig. 6. Relationship between the pathology score and DHA particle area in Aprt-deficient male mice. The figure shows three sets of polarizing(a, c, and e) and bright-field (b, d, and f ) images obtained from three different 120-day-old male mice. The bright-field images show the degreeof pathology in each kidney, while the polarizing images show the amount of birefringent crystals in the same kidney. (a and b) A mouse withlittle pathology (pathology score of 1) and a small particle area (246.7 mm2). (c and d) A mouse with mild pathology (pathology score of 2) anda moderate particle area (1261.2 mm2). (e and f) A mouse with advanced pathology (pathology score of 4) and a large particle area (14,957.4mm2). Panels (a–f) �150.


Fig. 7. Differences in DHA particle amount in kidneys from Aprt heterozygous and homozygous-deficient male mice. A very small amount ofcrystalline material is present in this 7- to 14-day-old heterozygous animal (a, arrows). However, (b) a much large number of birefringent particlesare seen in a homozygous-deficient animal of the same age (a and b; �45).

Table 3. Comparison of the total area occupied by DHA crystalsTable 2. Total area occupied by DHA crystals in kidney sectionsfrom Aprt heterozygous male and female mice of two in kidney sections from Aprt heterozygous and homozygous-

deficient male and female mice of two different age groupsdifferent age groups

Age Female Male Age Heterozygous Homozygous Significance

7 to 14 days 50.1�40.0 764.1�504.4 a7 to 14 days 44.4�11.7 52.5�49.5N 4 9 N 13 10

30 to 45 days 33.1�15.1 637.5�252.1 a30 to 45 days 34.2�18.4 30.5�0.7N 5 2 N 7 13

Data are expressed as mean � SD, mm2. Data are expressed as mean � SD, mm2.a Values for homozygous animals are significantly greater than those for het-

erozygous animals by two-way ANOVA

tals in heterozygous mice, we examined the kidneys ofwild-type mice and found no crystalline material or pa- oped in the late 1990s, and the renal pathological changesthology in any of these animals (Fig. 8). associated with DHA urolithiasis were confirmed [10–12].

The availability of these mice has enabled investigatorsto address several aspects of APRT deficiency that areDISCUSSIONnot possible in humans, including sequential analysis ofInterest in the nephrotoxicity associated with the ad-DHA crystallization and stone formation and the associ-ministration of adenine to experimental animals datesated pathological changes in animals of different sex andback to over a century, and it was suggested that theage; detailed SEM and x-ray diffraction analyses of DHAtoxicity was due to DHA formation in vivo [18]. Thiscrystals and stones; loss of DHA during organ perfusion,was confirmed in the rat in the 1950s [19, 20], and thefixation, and staining; and the effects, if any, of DHAenzymatic pathway was identified [21]. Because of inter-formation, in Aprt heterozygotes.est in the use of adenine to improve the preservation of

As observed in earlier studies in mice of mixed back-red blood cells, numerous studies on the toxicology andgrounds (129/Sv � C57BL/6J; 129/Sv � Black Swiss; andpharmacology of this purine were initiated in animals129/Ola � BALB/c) [10–12], kidneys from strain 129/Svand humans in the mid-1960s [22]. HistopathologicalAprt-deficient mice in the present study were dispropor-studies in kidneys from a patient who died soon aftertionately small, with a yellowish appearance, particularlyopen-heart surgery and who received large amounts ofin the severely affected males. The main pathologicalACD-adenine–preserved blood during and after surgerychanges were confined to the tubules and interstitiumrevealed the presence of rosette-shaped birefringentand included destruction of the medulla and part of thecrystals identified as DHA in renal tubules [23]. Sooncortical region; tubular obstruction, dilation, collapse,thereafter, DHA was identified as the principal constit-and atrophy; areas of tubular damage and regenerationuent of recurrent stones in young children with APRT

deficiency [2, 18, 24]. Aprt knockout mice were devel- with sloughing of tubular epithelial cells; and interstitial


Fig. 8. Bright-field and polarizing image analysis of kidneys from wild-type and Aprt-deficient male mice. (a and b) The histology and polarizingimage, respectively, from a three-month-old wild-type animal. There was no pathology or crystalline material in this or other wild-type animals.(c and d) The histology and polarizing image from a homozygous-deficient animal of the same age. Despite minimal cellular changes, this malemouse still displayed a noticeable amount of crystalline material (arrows) (a–d �200).

fibrosis that extended from the renal capsule to the pa- caused renal injury [26]. However, definitive distinctionbetween isoguanine and DHA could not be made bypilla. The histological changes were more diffuse and

interstitial fibrosis more extensive in male kidneys, the high-performance liquid chromatography (HPLC)system used. DHA stones may contain trace amountswhereas in females, much less of the kidney was affected

and fibrosis was more focal. Similar pathological changes of other materials, including calcium oxalate, calciumphosphate, or uric acid [26–29]. These authors also sug-have been observed in the urate oxidase-deficient mice

[25] and in wild-type mice receiving adenine or various gested that allopurinol did not alleviate the adenine-induced renal injury in ddY mice. This again is at vari-adenine analogues [26, 27].

The identity of the renal stone from the Aprt heterozy- ance with other published data, but the possibility of anextreme strain-specific effect cannot be ruled out.gous mouse, as well as bladder stones from homozygous-

deficient mice, was confirmed to be DHA by x-ray dif- The disease phenotype is more severe in Aprt-defi-cient male mice [10, 12, 14]. To determine whether thisfraction analysis. With one exception [27], all studies on

the structural analysis of the stone material from APRT- variation was related to variations in DHA crystal depo-sition, we compared the total area occupied by DHAdeficient patients or from adenine-fed experimental ani-

mals have identified the material as DHA. Minami et al particles with renal histopathology in mice of differentage groups. At age one to two days, comparable numbersstudied six-week-old male ddY strain mice and suggested

that adenine was converted to isoguanine, which then of DHA crystals were present in both male and female


mice, and there was no observable pathology. A strong In summary, we have demonstrated that both male andcorrelation between the pathology score and crystal bur- female Aprt-deficient mice accumulate DHA. However,den was detected in those males that had a very high the area occupied by DHA crystals was significantlystone burden. Stone burden remained low in the female greater in the 120- to 240-day-old males compared with129 strain mice, as did the level of renal pathology. Thus, the females of similar age. Substantial renal pathologywe were not able to determine the effects of a high stone was detected in the kidneys of male mice that had veryburden on renal pathology in these mice. Data on renal high levels of stone material. Both male and female Aprthistopathology and DHA crystal/stone burden are not heterozygotes have a modest crystal burden but no pa-available for the vast majority of APRT-deficient pathology. We have recently identified several genes thattients. The largest number of such patients has been promote renal injury in Aprt-deficient mice (Wang etidentified in Japan, and examination of the clinical data al, submitted for publication; Tzortzaki et al, submittedsuggests that urolithiasis and renal failure also may be for publication) [14]. Functional studies of these genesmore common in males [1]. are in progress.

A number of factors may account for the increaseddisease severity in male mice. Crystallization occurs when ACKNOWLEDGMENTSthe concentration of DHA exceeds the supersaturation

This work was supported by NIH grants DK38185, ES05652, andpoint. Differences in the ability to supersaturate the urine ES06096.with DHA may contribute to the observed phenotypic

Reprint requests to Andrew P. Evan, Ph.D., Department of Anatomydifferences in mice, as has been noted in a number ofand Cell Biology, Indiana University School of Medicine, 635 Barnhill

APRT-deficient patients [2]. Male mice also may pro- Drive (MS 5055), Indianapolis, Indiana 46202-5120, USA.E-mail: [email protected] more polyamines, the endogenous source of ade-

nine, and this could predispose Aprt-deficient males tomore severe renal disease [2, 12]. Since DHA is formed

APPENDIXfrom adenine via XDH, variations in the activity of thisAbbreviations used in this article are: AMP, 5�-adenosine mono-enzyme in males and females may account for the differ-

phosphate; APRT, adenine phosphoribosyltransferase; DHA, 2,8-dihy-ences in phenotype. The increased expression of genes droxyadenine; RT-PCR, reverse transcription-polymerase chain reaction;related to fibrosis, tissue calcification, and transmembrane SEM, scanning electron microscopy; SDH, xanthine dehydrogenase.proteins in kidneys from Aprt-deficient male mice sup-ports the gender bias in disease severity [14]. It is also REFERENCESlikely that sex-related differences in disease severity may

1. Sahota AS, Tischfield JA, Kamatani N, Simmonds HA: Adeninesimply reflect intrinsic differences in kidney structure phosphoribosyltransferase deficiency and 2,8-dihydroxyadenine li-

thiasis, in The Metabolic and Molecular Bases of Inherited Diseaseand function, blood pressure, or hormonal changes [30].(8th ed), edited by Scriver CR, Beaudet AL, Sly WS, et al, NewHuman APRT obligate heterozygotes do not have anyYork, McGraw-Hill, 2001, pp 2571–2584clinical, biochemical, or immunological abnormalities, 2. Simmonds HA, Sahota AS, Van Acker KJ: Adenine phosphoribo-

and the urinary and plasma levels of adenine and DHA syltransferase deficiency and 2,8-dihydroxyadenine lithiasis, in TheMetabolic and Molecular Bases of Inherited Disease (7th ed), editedin normal individuals and heterozygotes are at the limitby Scriver CR, Beaudet AL, Sly WS, Valle D, New York,of detection of the commonly used analytical techniques McGraw-Hill, 1995, pp 1707–1724

[2]. The urinary levels of adenine and DHA in wild- 3. Arnadottir M, Laxdal T, Hardarson S, Asmundsson P: Acuterenal failure in a middle-aged woman with 2,8-dihdroxyadeninuria.type and Aprt heterozygous mice are also at the limit ofNephrol Dial Transplant 12:1985–1987, 1997detection [10, 11]. The finding of a stone in the renal

4. Brown HA: Recurrence of 2,8-dihydroxyadenine tubulointerstitialpelvis of an Aprt heterozygous mouse was therefore un- lesions in a kidney transplant recipient with a primary presentation

of chronic renal failure. Nephrol Dial Transplant 13:998–1000, 1998expected. In animal studies, little DHA was formed from5. de Jong DJ, Assmann KJ, De Abreu RA, et al: 2,8-Dihydroxyade-adenine at doses below 10 mg/kg, but with increasing

nine stone formation in a renal transplant recipient due to adenineadenine dosage, yellowish spheres of DHA appeared in phosphoribosyltransferase deficiency. J Urol 156:1754–1755, 1996the urine and tubular lumens [2]. The finding of low 6. Gagne ER, Deland E, Daudon M, et al: Chronic renal failure

secondary to 2,8-dihydroxyadenine deposition: The first report oflevels of DHA birefringent material in kidneys fromrecurrence in a kidney transplant. Am J Kidney Dis 24:104–107,Aprt heterozygous male and female mice, without any 1994

increase with age or any associated pathology, clearly 7. Gelb AB, Fye KH, Tischfield JA, et al: Renal insufficiency sec-ondary to 2,8-dihydroxyadenine urolithiasis. Hum Pathol 23:1081–demonstrated that the endogenous production of ade-1085, 1992nine exceeded its conversion to AMP by Aprt. The Km 8. Glicklich D, Gruber HE, Matas AJ, et al: 2,8-Dihydroxyadenine

for adenine for the human enzyme is in the �mol/L range urolithiasis: Report of a case first diagnosed after renal transplant.Q J Med 69:785–795, 1988[2]. These observations suggest that the phenotype in

9. Winter P, Hesse A, Klocke K, Schaefer RM: Scanning electronhuman APRT heterozygotes depends on the level atmicroscopy of 2,8-dihydroxyadenine crystals and stones. Scanning

which it is examined, a finding that maybe equally appli- Microsc 7:1075–1080, 199310. Engle SJ, Stockelman MG, Chen J, et al: Adenine phosphoribo-cable to other autosomal-recessive diseases.


syltransferase-deficient mice develop 2,8-dihydroxyadenine neph- 21. Wyngaarden JB, Dunn DT: 8-Hydroxyadenine as the metabolicintermediate in the oxidation of adenine to 2,8-dihydroxyadeninerolithiasis. Proc Natl Acad Sci USA 93:5307–5312, 1996

11. Redhead NJ, Selfridge J, Wu CL, Melton DW: Mice with ade- by xanthine oxidase. Arch Biochem Biophys 70:150–156, 195722. Warner WL: Toxicology and pharmacology of adenine in animalsnine phosphoribosyltransferase deficiency develop fatal 2,8-dihy-

droxyadenine lithiasis. Hum Gene Ther 7:1491–1502, 1996 and man. Transfusion 17:326–332, 197723. Falk JS, Lindblad GTO, Westman BJM: Histopathological stud-12. Stockelman MG, Lorenz JN, Smith FN, et al: Chronic renal failure

in a mouse model of human adenine phosphoribosyltransferase ies on kidneys from patients treated with large amounts of bloodpreserved with ACD-adenine. Transfusion 12:376–381, 1972deficiency. Am J Physiol 275: F154–F163, 1998

13. Finger S, Heavens RP, Sirinathsinghji DJ, et al: Behavioral and 24. Cartier P, Hamet M, Hamburger J: Une nouvelle maladie meta-bolique: Le deficit complet en adenine phosphoribosyltransferaseneurochemical evaluation of a transgenic mouse model of Lesch-

Nyhan syndrome. J Neurol Sci 86:203–213, 1988 avec lithiasis de 2,8-DHA. Comptes Rendus Hebdomadaires Se-ances Acad Sci 279:883–886, 197414. Wang L, Raikwar N, Deng L, et al: Altered gene expression in

kidneys of mice with 2,8-dihydroxyadenine nephrolithiasis. Kidney 25. Wu X, Wakamiya M, Vaishnav S, et al: Hyperuricemia and uratenephropathy in urate oxidase-deficient mice. Proc Natl Acad SciInt 58:528–536, 2000

15. Stambrook PJ, Shao C, Stockelman M, et al: APRT: A versatile in USA 91:742–746, 199426. Minami T, Nakagawa H, Ichii M, et al: Nephrotoxicity inducedvivo resident reporter of local mutation and loss of heterozygosity.

Environ Mol Mutagen 28:471–482, 1996 by a single dose of adenine: Effects of 4-aminopyrazolo[3,4-d]pyri-midine and allopurinol. Biol Pharm Bull 17:201–206, 199416. Gardner KD, Evan AP, Reed WP: Accelerated renal cyst devel-

opment in deconditioned germ-free rats. Kidney Int 29:1116–1123, 27. Minami T, Nakagawa H, Nabeshima M, et al: Nephrotoxicityinduced by adenine and its analogs: Relationship between structure1986

17. Khan SR, Hackett RL: Retention of calcium oxalate crystals in and renal injury. Biol Pharm Bull 17:1032–1037, 199428. Estepa-Maurice L, Hennequin C, Marfisi C, et al: Fourier trans-renal tubules. Scanning Microsc 5:707–712, 1991

18. Simmonds HA, Van Acker KJ, Cameron JS, Sneddon W: The form infrared microscopy identification of crystal deposits in tissue:Clinical importance in various pathologies. Clin Chem 105:576–582,identification of 2,8-dihydroxyadenine, a new component of uri-

nary stones. Biochem J 157:485–487, 1976 199629. Szonyi P, Bereni M, Toth B: A rare enzyme deficiency causing19. Bendich A, Bosworth-Brown G, Philips FS, Thiersch JB: The

direct oxidation of adenine in vivo. J Biol Chem 183:267–277, 1950 formation of 2,8-dihydroxyadenine (purine body) calculi. Int UrolNephrol 17:231–233, 198520. Philips FS, Thiersch JB, Bendich A: Adenine intoxication in

relation to in vivo formation and deposition of 2,8-dihydroxyade- 30. Silbiger SR, Neugarten J: The impact of gender on the progres-sion of chronic renal disease. Am J Kidney Dis 25:515–533, 1995nine in renal tubules. J Pharmacol Exp Ther 104:20–30, 1952

Documents

Sequential analysis of kidney stone formation in the Aprt knockout mouse