View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Kidney International, Vol. 58 (2000), pp. 528–536

Altered expression in kidneys of mice with 2,8-dihydroxyadenine nephrolithiasis

LI WANG,NANDITA RAIKWAR,LI DENG,MIN YANG,LI LIANG,CHANGSHUN SHAO, ANDREW P. EVAN,PETER J. STAMBROOK,AMRIK SAHOTA, and JAY A. TISCHFIELD

Departments of Medical and Molecular Genetics and Anatomy, Indiana University School of Medicine, Indianapolis, Indiana; Department of Genetics, Rutgers University, Piscataway, New Jersey; and Department of Cell Biology, Neurobiology, and Anatomy, University of Cincinnati College of Medicine, Cincinnati, Ohio, USA

Altered gene expression in kidneys of mice with 2,8-dihydroxy- Conclusions. These findings suggest that (1) there are sex- adenine nephrolithiasis. related differences in gene expression in DHA lithiasis, possi- Background. We have developed a knockout mouse model bly caused by increased deposition of DHA crystals in male for adenine phosphoribosyltransferase (APRT) deficiency, a compared with female kidneys; and (2) the expression of cer- condition that often leads to 2,8-dihydroxyadenine (DHA) tain (for example, C10) may simply be an indication of nephrolithiasis in humans. Aprt knockout male mice develop nonspecific cellular stimulation and may not be related to renal severe renal damage by three months of age, but this is strain injury. specific. Renal damage in female mice is less pronounced than in males. The gene level changes that promote renal injury in APRT-deficient mice are not known. Methods. We used mRNA differential display polymerase Adenine phosphoribosyltransferase (APRT) is a ubiq- chain reaction (DD-PCR) to analyze renal gene expression uitously expressed enzyme that catalyzes the formation changes in APRT-deficient male and female mice (strain C3H) of AMP from adenine and 5-phosphoribosyl-1-pyro- compared with age- and sex-matched Aprt heterozygote con- phosphate. In human APRT deficiency, adenine is oxi- trols. The differentially amplified bands were reamplified, cloned, sequenced, and queried against the National Center dized by xanthine dehydrogenase to 2,8-dihydroxyade- for Biotechnology Information nonredundant databases using nine (DHA). The sparingly soluble nature of DHA at the Basic Alignment Search Tool. Relative quantitative reverse the normal urinary pH results in the excretion of DHA transcription-polymerase chain reaction was used to confirm crystals in the urine and, frequently, the deposition of the results of DD-PCR for a selected number of genes in one-, DHA stones in the kidneys. Clinical symptoms of APRT three-, and six-month-old male and female mice. Results. Sixty-three differentially amplified bands were identi- deficiency vary from benign to life threatening and may fied, including 21 for known genes, and 8 of these were examined be present from birth or may be onset late in life. Major further. In three-month-old APRT-deficient male mice, the signs and symptoms include crystalluria, hematuria, dys- expression of C10 was increased tenfold, and there was a four- uria, urinary tract infection, chronic interstitial nephritis, fold to sevenfold increase in the expression of a disintegrin and and, in some cases, chronic renal failure [1]. metalloproteinase with thrombospondin motifs (ADAMTS-1), MGP (matrix Gla ), and (LOX). The ex- We have generated APRT-deficient mice by gene pression of cholecystokinin-A receptor (CCKAR), imprinted knockout technology [2, 3]. These mice mimic the human multimembrane-spanning polyspecific transporter-like gene 1 disease, but the disease appears to be more severe in (IMPT-1), and kidney androgen-regulated protein (KAP) was male than in female mice [2, 3]. As a consequence of diminished twofold to fourfold, but there was little or no change in the expression of organic anion transporter (OATP). Except stone deposition in the kidneys, there is extensive struc- for a more than tenfold increase in C10 expression and up to tural damage (inflammation, tubular dilation, necrosis, tenfold decrease in KAP expression, APRT-deficient female and interstitial fibrosis) and loss of renal function in these mice did not show significant changes in gene expression com- mice. Laboratory measurements indicate that blood urea pared with controls. nitrogen is increased and creatinine clearance decreased in APRT-deficient male mice compared with wild-type Key words: adenine phosphoribosyltransferase deficiency, gene knock- controls [3]. Female APRT-deficient mice show signifi- out mice, renal gene, urinary tract infection, crystalluria, hematuria. cantly less renal damage than male mice of the same age. At age 12 weeks, for example, APRT-deficient females did Received for publication June 1, 1999 and in revised form February 25, 2000 not show a significant impairment in glomerular filtration Accepted for publication March 13, 2000 rate, as estimated by creatinine clearance [3].  2000 by the International Society of Nephrology The underlying molecular mechanisms involved in the

528 Wang et al: Alterations in renal gene expression 529 pathogenesis of kidney stone disease, particularly alter- month-old male C3H Aprt heterozygote and a homozy- ations in the expression of disease-related genes, are un- gous-deficient mouse was reverse transcribed with one known. The availability of a mouse model for this disease of 3 HT11M (H ϭ HindIII site, T11 ϭ 11Ts, M ϭ A, C, enables us, to our knowledge for the first time, to explore or G) downstream primers using Superscript II RNase HϪ the molecular pathological basis of DHA-induced neph- reverse transcriptase (GIBCO BRL). The same primer rolithiasis. In the present study, mRNA differential dis- was used subsequently as the 3Ј primer to amplify the play polymerase chain reaction (DD-PCR) was used to single-stranded cDNAs together with 1 of 16 different identify 63 cDNAs, including 21 known genes, in which random 13-mer 5Ј primers, in the presence of [␣-33P]- the expression was altered in APRT-deficient mice. The dATP (Amersham, Arlington Heights, IL, USA). The known genes encode such as extracellular ma- PCR products were separated on a 6% denaturing poly- trix (ECM) proteins, membrane transporters, metabolic acrylamide gel, dried, and exposed to Kodak Max x-ray enzymes, and hormone-regulated proteins involved in film for three to seven days. Gel slices containing the various physiological and pathological processes. The differentially amplified products were soaked in 20 mL expression of eight of these genes was further examined water for 15 minutes to remove the urea and were then by relative quantitative reverse transcription-PCR (RT- reamplified using the same primers [5]. PCR) in APRT-deficient male and female mice of vari- ous ages. Our findings suggest that there are sex-related DNA cloning and sequencing differences in gene expression in DHA lithiasis and that The reamplified PCR products were cloned into vector the expression of certain genes may be related to nonspe- CR2.1 using the TA cloning kit (Invitrogen, Carlsbad, cific cellular stimulation rather than to renal injury. CA, USA). The recombinant plasmids were examined for inserts by colony PCR using M13 reverse and forward primers [5] and were then sequenced using the Thermo METHODS Sequenase radiolabeled terminator cycle sequencing kit Animals (Amersham, Cleveland, OH, USA). The sequences were The Aprt gene knockout mice were originally gener- queried against the National Center for Biotechnology ated in a mixed background of strains 129/Sv (embryonic Information (NCBI) GenBank and dbEST databases using stem cells) and C57BL/6 J (blastocyst donors) [2]. The Basic Local Alignment Search Tool algorithm (BLAST). mixed-strain heterozygotes were backcrossed to strain C3H/HeJ mice for 10 generations, after which the prog- Relative quantitative RT-PCR eny was expected to be identical at 99.9% of their loci. Relative quantitative RT-PCR for a select number C3H/HeJ Aprt heterozygotes were then mated to gener- of genes was carried out using the QuantumRNA kit ate APRT-deficient homozygotes. Using the same strat- following the manufacturer’s instructions (Ambion, Inc., egy, we have also introduced the Aprt mutation into Austin, TX, USA). Briefly, DNA-free total RNA (1.2 ␮g) strains 129, Black Swiss, and C57. The Aprt genotypes was reverse transcribed with Superscript II RNase HϪ were determined by PCR analysis of DNA isolated from reverse transcriptase (GIBCO BRL) using random de- tail biopsies, as described previously [4]. The gross renal camers as downstream primers. The second-round PCR histologic changes in the different strains were similar, reaction was performed with single-stranded cDNA orig- but the extent and timing of any strain-specific differences inating from the equivalent of 1 to 10 ng total RNA in remain to be established. We used C3H Aprt heterozy- a10␮L volume containing 1 ϫ PCR buffer, 200 ␮mol/L gotes as controls in the following experiments, as pheno- each dNTP, 500 nmol/L gene-specific primers, and 0.25 typic appearance and gross renal histology did not show units Taq DNA polymerase (GIBCO BRL). The prim- any differences compared with wild-type animals [2]. ers were designed based on cDNA sequence analysis and were obtained from GIBCO BRL. The PCR parameters Total RNA isolation were a 15-second denaturation at 93ЊC, a 30-second an- Total RNA was isolated from whole mouse kidneys nealing at 52ЊCto58ЊC (depending on the gene), and a using the Qiagen RNAeasy mini kits (Santa Clarita, CA, 30-second extension at 72ЊC for 22 to 40 cycles. The PCR USA). Contaminating DNA was removed by treating products were resolved on a 1% agarose gel, stained the samples with RNase-free DNase I (GIBCO BRL, with ethidium bromide, and analyzed by densitometery. Gaithersburg, MD, USA). A mixed ratio of 18S ribosomal RNA primers and com- petimers was used to amplify rRNA as an internal con- Differential display PCR trol under the same conditions as the genes of interest. Differential display-PCR was performed using RNA- A competimer is a primer that has been modified at its image kits according to the manufacturer’s instructions 3Ј end to block extension by DNA polymerase. (GenHunter Corporation, Nashville, TN, USA). Briefly, For a given sample, a standard curve was obtained by 0.2 ␮g of total RNA from kidney tissue from a three- plotting the densitometric signal intensity of the rRNA 530 Wang et al: Alterations in renal gene expression

PCR product versus the primers:competimers ratio using C10, a macrophage inflammatory protein; cholecystoki- GraphPad Prism software (GraphPad Software Inc., San nin-A receptor (CCKAR); imprinted multimembrane- Diego, CA, USA). With different mixed ratios of 18S spanning polyspecific transporter-like gene 1 (IMPT-1); rRNA primers and competimers (usually primers:com- kidney androgen-regulated protein (KAP); lysyl oxidase petimers ratios from 10:0 to 3:7), signal intensity of the (LOX); matrix Gla protein (MGP); and an organic anion 488 bp 18S rRNA PCR band ranged from strong to weak. transporter (OATP). CCKAR and IMPT-1 are tubule The sample signal intensities were normalized as de- cell surface proteins. MGP is an ECM protein that inhib- scribed in the Ambion manual, and the change in gene its tissue calcification. LOX catalyzes the cross-linking expression was expressed as the ratio of the signal inten- of in tissue fibrosis, and KAP and OATP are two sity of the RT-PCR product from an APRT-deficient hormone-regulated proteins. The PCR primer sequences mouse divided by the intensity of the product from a and the expected product sizes for these gene fragments heterozygote control. Because of inherent biological and are shown in Table 2. experimental variation, expression ratios in the range For relative quantitative RT-PCR, 18S rRNA was used 0.5 to 2.0 were not considered significant. both as an internal control and as an indicator that PCR amplification was proportional to the template concen- tration. Figure 1A shows PCR products from 18S rRNA RESULTS and ADAMTS-1 mRNA. Using calibration curves simi- Identification of differentially expressed mRNAs lar to the one shown in Figure 1B, the expression of Kidneys from a three-month-old male APRT-deficient ADAMTS-1 in APRT-deficient mice was normalized C3H mouse were pale yellow, atrophic, and of irregular against the internal control and then compared with ex- shape, compared with healthy kidneys from an Aprt het- pression in heterozygous mice. The RT-PCR results were erozygote of the same age. Total RNA extracted from similar to those from DD-PCR, indicating a significant kidneys of APRT-deficient and heterozygous mice was elevation in gene expression for ADAMTS-1, C10, LOX, reverse transcribed and amplified by PCR. The RT-PCR and MGP, and decreased expression for KAP, CCKAR, products from the two mice were electrophoresed side and IMPT-1 (Table 3). A modest but insignificant de- by side, and 63 differentially amplified bands were identi- crease in OATP expression was detected by RT-PCR. fied. Of these, 26 were strongly expressed in the Aprt heterozygous mouse, and 37 were strongly expressed in Analysis of gene expression at different ages the APRT-deficient mouse. The bands were cut from the To evaluate the onset and significance of altered gene gels, reamplified, cloned, sequenced, and queried against expression, kidneys from male and female mice aged the NCBI nonredundant databases. cDNA fragments in one, three, and six months were examined for the above the size range 100 to 500 bp provided sufficient sequence eight genes. The gross renal morphology of one-month- information for DNA database searches. old APRT-deficient male mice was similar to heterozy- Fifty-three of the 63 cDNA clones (about 83%) gote controls, but the kidneys from three- and six-month- matched a total of 44 nonredundant GenBank and dbEST old male mice were highly abnormal, showing atrophy, database entries. The 63 cDNA fragments were divided pale color, and irregular shape. At a given age, kidneys into the following categories: (1) previously identified from APRT-deficient female mice were much less se- mouse genes (25 clones), (2) mouse homologues of iden- verely affected than from male mice. tified rat or human genes (4 clones), (3) mouse expressed The renal expression of C10 was increased tenfold sequence tag (EST) clones (20 clones), (4) mouse homo- in a three-month-old, APRT-deficient male mouse, and logues of rat or human ESTs (4 clones), and (5) no match there was a fourfold to sevenfold increase in the expres- found by BLAST (10 clones). Twenty-one known genes sion of ADAMTS-1, MGP, and LOX. There was little (Table 1), in which the renal expression was altered in or no change in the expression of these four genes at APRT-deficient mice, encoded proteins such as ECM one and six months (Fig. 2A). The expression of CCKAR proteins, membrane transporters, metabolic enzymes, and IMPT-1 was diminished twofold to fourfold in male and hormone-regulated proteins involved in various mice at three months, with a modest decrease at one physiological and pathological processes. and six months (Fig. 2B). There was also a fourfold decrease in the expression of KAP, and this decrease Quantitative analysis of gene expression persisted at six months (Fig. 2C). There was little or no Because of their functional importance and to confirm change in the expression of OATP in the three age the results of DD-PCR, the expression of eight known groups (Fig. 2C). Except for a more than tenfold increase genes in three-month-old homozygous-deficient mice and in C10 expression and up to tenfold decrease in KAP in heterozygous mice was further examined by relative expression at three months, APRT-deficient female mice quantitative RT-PCR. These were a disintegrin and metal- did not show significant changes in gene expression com- loproteinase with thrombospondin motifs (ADAMTS-1); pared with heterozygote controls (Fig. 2 D–F). Wang et al: Alterations in renal gene expression 531

Table 1. Differentially expressed genes in kidneys of a 3-month-old male Aprt Ϫ/Ϫ C3H mouse identified by DD-PCR Change in Product size Clone expression bp Sequence identity GenBank # A2-1 ϩ 340 Pentylenetetrazol-related mRNA D45203 A3-2, G3-1 ϩ 400 Arginase II AF032466 A4ϩ1 Ϫ 360 Imprinted multimembrane-spanning polyspecific transporter-like gene 1 (IMPT-1) AF028739 A9-2 ϩ 420 Proteolipid protein (PLP) M15442 A11ϩ3 Ϫ 300 Cholecystokinin type-A receptor (CCKAR) D85605 C11ϩ1, 360 G11ϩ1, 360 G11ϩ2, 240 G11-2 360 A11-3 ϩ 120 Macrophage inflammatory protein (C10) M58004 C4-1 ϩ 300 Homeodomain protein AJ000507 C4-4 ϩ 120 A disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS-1) AB001735 C5ϩ3 Ϫ 180 Fragile X mental retardation L23971 G4-1, G15-3 ϩ 180 Mouse lysyl oxidase (LOX) M65142 G5ϩ1 Ϫ 200 Hydrophobic protein L07095 G1ϩ1, Ϫ 200 Kidney androgen-regulated protein (KAP) M22810, G7ϩ1, M63707 G7ϩ2 G9-1 ϩ 100 Matrix Gla protein (MGP) D00613, S77350 G9-3 ϩ 120 Arylsulfatase A X73230 G9-4 ϩ 120 EAT/MCL-1 mRNA U35623 G15-1 ϩ 600 T cell receptor AE000664 G16-1 ϩ 240 T cell receptor ␣/␦ M64239 A16ϩ2 Ϫ 490 Similar to rat cytochrome b5 D13205 C2-1 ϩ 200 Similar to rat liver proteasomal ATPase, rat Tat-binding protein, human immuno- D83522, deficiency virus tat transactivator binding protein-1 U77918, M34079 G9ϩ1 Ϫ 400 Similar to rat organic anion transporter (OATP) L19031 G15-2 ϩ 100 Similar to rat diacylglycerol kinase (DGK) S49760 Symbols are: ϩ, increased renal expression in APRTϪ/Ϫ mouse; Ϫ, decreased renal expression in APRTϪ/Ϫ mouse.

DISCUSSION served in lung tissue cultures from mice treated with NG- APRT-deficient mice generated by gene knockout de- nitro-l-arginine-methyl ester, an inducer of granuloma- velop DHA nephrolithiasis, as in humans. We have re- tous inflammatory lesions in the lung [7]. Our studies cently described the renal pathological characteristics show that the level of C10 renal mRNA expression in- and some of the functional consequences of kidney dam- creased about threefold in one-month-old male mice and age in these mice [2, 3]. Because of the complexity of reached a tenfold increase in both male and female mice kidney damage associated with APRT deficiency, it is at three months of age, compared with heterozygote controls. These findings implicate C10 as a mediator of expected that there will be alterations in the expression inflammation and infiltration of lymphocytes in kidneys of many genes in this disease. In the present study, we of APRT-deficient mice. These data are consistent with have shown that there are at least 63 differentially ex- previous observations that kidneys of one-month-old, pressed mRNAs (including 21 known genes) in kidneys APRT-deficient mice exhibit mild-to-moderate inflam- of APRT-deficient mice. The corresponding proteins for mation and some crystal deposition associated with the known genes are involved in many physiological and lymphocytic infiltration [2, 3]. Since the extent of renal pathological processes, suggesting that they may also be damage in APRT-deficient females was significantly important in the pathogenesis of DHA nephrolithiasis. lower than in APRT-deficient males, it is also possible Based on information about their function, eight of these that the expression of C10 is not a marker for renal genes were selected for further investigation, as de- injury but may simply be a consequence of nonspecific scribed later in this article. cellular stimulation. Our finding that the renal expression of ADAMTS-1 Inflammation-associated genes (C10 and ADAMTS-1) is increased in APRT-deficient male mice but not in fe- C10, a mouse macrophage inflammatory protein, be- male mice is very interesting, since this gene has been longs to the C-C chemokine family, and it induces direc- reported to be associated with inflammation. ADAMTS-1 tional migration of T cells, monocytes, and eosinophils. encodes an ECM-binding protein that belongs to the C10 is induced in macrophages by an interleukin-4 stimu- cellular disintegrin and metalloproteinases (ADAM) lus [6]. Up-regulation of C10 expression has been ob- family [8]. In control mice, a very weak signal for 532 Wang et al: Alterations in renal gene expression

Table 2. Sequences of primers for relative quantitative RT-PCR and the sizes of PCR products Gene Primers Size bp ADAMTS-1 5ЈGGTGCAAGCTCACCTGTGAAGC 3Ј;5ЈCATCTTCTTGCATGTGGAACCG 3Ј 212 C10 5ЈAGGATGAGAAACTCCAAGACTG 3Ј;5ЈTCAAGCAATGACCTTGTTCCCA 3Ј 354 CCKAR 5ЈTCTGGAGCTCTACCAAGGAATC 3Ј;5ЈGACCACAATGACAATGAGCATG 3Ј 286 IMPT-1 5ЈTCTTCGGGATCCTCCAGATGA 3Ј;5ЈGTTGCACAGATGCAGAGAGGC 3Ј 246 KAP 5ЈTCACTGTCTTCTGTGGTCTGAC 3Ј;5ЈACAATATCCTGAATGGCAGTCG 3Ј 390 LOX 5ЈCTGCCTGGCCAGTTCAGCATAT 3Ј;5ЈTCCACTGGCAGTCTATGTCTGC 3Ј 302 MGP 5ЈTAATATTTGGCTCCTCGGCGCT 3Ј;5ЈCGAGACACCATGAAGAGCCTGC 3Ј 323 OAPT 5ЈACAGCCATACCTGGGTACATG 3Ј;5ЈGATAGCTTGATCCTCTTAGTGC 3Ј 262 For each gene, the sequence on the left is for the forward primer.

Fig. 1. Relative quantitation of ADAMTS-1 mRNA from a three-month-old adenine phos- phoribosyltransferase (APRT)-deficient mouse by reverse transcription-polymerase chain reac- tion (RT-PCR). (A) Gel showing the RT-PCR products from 18S rRNA and ADAMTS-1 mRNA stained with ethidium bromide. The 18S primers/competimers ratios ranged from 6:4 to 2:8. (B) Standard curves for 18S rRNA RT-PCR. The intensities of the 18S rRNA PCR products for Aprt heterozygote (ϩ/Ϫ; ᭛; female) and homozygous-deficient (Ϫ/Ϫ; ᭜; female) samples were plotted against the 18S rRNA primers/compatimers ratio by linear re- gression analysis. The intensity of the signal for ADAMTS-1 and other genes was normal- ized, as described in the QuantumRNA man- ual (Ambion). Each data point is the mean of five replicates from one animal.

ADAMTS-1 transcripts was detected in the heart and diseases [10, 11], and increased LOX expression has been kidney but not in other organs (lung, liver, brain, and observed in fibrotic renal lesions during chronic adriamy- muscle). Upon lipopolysaccharide administration, there cin nephropathy [12]. In adriamycin-treated rats, the ex- was systemic inflammation as well as stimulation in the pression of LOX in total kidney, glomeruli, and medulla expression of ADAMTS-1 in kidney and heart. Interleu- increased up to threefold between weeks 8 and 12. These kin-1 stimulates ADAMTS-1 mRNA expression in vitro studies suggested that collagen cross-link formation by in colon 26 cells [9]. LOX might be implicated in the pathogenesis of irrevers- ible, fibrotic renal lesions. Although mesangial cells, kid- LOX and kidney fibrosis ney fibroblasts, and epithelial cells all express LOX, only Lysyl oxidase is an enzyme that catalyzes collagen expression in epithelial cells is induced by transforming cross-linking through oxidative deamination of lysyl growth factor-␤ (TGF-␤). Our previous studies showed ε-amino groups to form stable and insoluble collagen that kidney lesions typical of chronic interstitial nephritis products. The imbalance between collagen deposition were apparent in homozygous-deficient male mice [2]. and removal results in an abnormal accumulation of Our current observation that the renal expression of ECM proteins. Several lines of evidence suggest that LOX is elevated in these mice implicates a role for this LOX expression is significantly elevated in fibrotic liver gene in renal fibrosis in APRT deficiency. Wang et al: Alterations in renal gene expression 533

Table 3. Changes in renal gene expression in a 3-month-old Hormone-regulated proteins (KAP and OATP) male Aprt Ϫ/Ϫ mouse identified by DD-PCR or relative quantitative RT-PCR Kidney androgen-regulated protein (KAP) represents Change in expression the most abundant mRNA species in mouse kidney and Change in expression by RT-PCR is expressed in epithelial cells of proximal convoluted Gene by DD-PCR (fold of control)a tubules [20]. The expression of KAP is multihormone C10 ϩ 10.62 regulated. Androgens stimulate KAP gene expression in LOX ϩ 7.69 epithelial cells of proximal convoluted tubules of cortical MGP ϩ 4.72 ADAMTS-1 ϩ 4.33 nephrons, while estrogens and pituitary hormones con- CCKAR Ϫ 0.33 trol its expression in the juxtamedullary S3 segment of IMPT-1 Ϫ 0.41 the tubules [21]. Furthermore, the androgenic response KAP Ϫ 0.25 OATP Ϫ 0.90 of KAP gene expression is independent of pituitary func- Symbols are: ϩ, increased renal expression in APRTϪ/Ϫ mouse; Ϫ, decreased tion but dependent on thyroid hormone for maximal renal expression in APRTϪ/Ϫ mouse induction [22]. OATP is another hormone-regulated gene a Data are the mean of triplicate measurements from the same mouse that encodes an organic anion transporter and is expressed only in the apical membrane of S3 proximal tubules in the kidney [23]. Regulation of renal OATP mRNA MGP and tissue calcification expression, stimulated by testosterone and inhibited by estrogen, is important in modulating the renal tubular Matrix Gla protein is a low molecular weight (14 kD), secretion of estradiol conjugated to form estradiol-17-␤- ␥-carboxyglutamic acid (Gla)-containing, - D-glucoronide inside the tubule cells [24]. The impair- dependent protein, and is a member of the family of ment of OATP transporter function may induce the accu- ECM mineral-binding Gla proteins. It is highly expressed mulation of estradiol in the kidney, and this may inhibit in developing and , acting as a regulator OATP expression. of calcification of the ECM [13, 14]. It is also expressed at a lower level in soft tissues, including kidney, heart, APRT deficiency in male and female mice lung, and aorta, but its role is not clearly understood As reported previously, APRT-deficient female mice [15]. Kidney epithelial cells express MGP mRNA that are much less severely affected than males [3]. We have can be developmentally regulated in vivo and is regu- carried out studies on urinary excretion of DHA (Stam- lated by growth factors and cell density in vitro [16, 17]. brook et al, unpublished data) and DHA crystal deposi- The physiological role of MGP in kidney is not yet clear. tion in kidneys (Evan et al, manuscript in preparation) However, the increased expression of MGP in the kid- in male and female APRT-deficient mice of different neys of APRT-deficient mice may implicate its impor- age groups. Males excrete significantly more DHA than tance in DHA nephrolithiasis. females, and they also have more DHA crystals in the kidneys, suggesting that the observed sex difference in Transmembrane proteins (CCKAR and IMPT-1) disease severity may be related to increased crystal depo- Cholecystokinin-A receptor (CCKAR), which is widely sition in male kidneys. The increased expression of LOX, expressed in the gastrointestinal system as well as in brain ADAMTS-1, MGP, CCKAR, and IMPT-1 in kidneys and kidney, functions as a receptor for gastrin/cholecys- from APRT-deficient male mice supports the gender bias in disease severity. tokinin for trophic control of the pancreas and gastroin- In kidneys from three-month-old male mice, the pri- testinal tract [18]. The IMPT-1 gene encodes a predicted mary pathological change was interstitial fibrosis, which multimembrane-spanning protein similar to bacterial extended from the renal capsule to the papilla. This and eukaryotic polyspecific metabolite transporter and change appeared diffuse in that the entire kidney usually multidrug resistance pump [19]. IMPT-1 mRNAs are possessed extensive regions of fibrosis. Interstitial fibro- highly expressed in tissues with metabolite transport sis appeared more focal in kidneys from three-month-old functions, including liver, kidney, intestine, extra-embry- female mice, and much less of the kidney was affected. A onic membranes, and placenta. In our studies, the finding midcoronal section of a female kidney typically showed that the renal expression of these two genes is decreased three to four small regions of fibrosis. Mild hydronephro- in APRT-deficient mice suggests an impairment of re- sis seen as dilation of the renal pelvis and some cortical ceptor and transporter functions, which may ultimately thinning were found in both male and female mice; how- affect renal function. These data are consistent with our ever, again, this was less common in females. previous observations that kidney tubule damage is a Within the regions of renal interstitial fibrosis in male striking feature in APRT deficiency and results in the and female mice, the following alterations were com- physiological impairment of renal function [3]. monly seen: (1) tubular atrophy of primarily the proximal 534 Wang et al: Alterations in renal gene expression

Fig. 2. RT-PCR for renal gene expression in male (A–C) and female (D–F) APRT-deficient mice at various ages. The graphs show the ratio of renal gene expression in APRT-deficient mice compared with heterozygote controls of the same age. Each data point is the mean of triplicate measurements from one animal. Symbols in A and D are: (᭿) C10; (ᮀ) LOX; (ϫ) MGP; (᭢) ADAMTS-1. Symbols in B and E are: (᭹) CCKAR; (᭺) IMPT. Symbols in C and F are: (᭜) OATP; (᭛) KAP. Wang et al: Alterations in renal gene expression 535 tubules, (2) tubular dilation of the proximal tubules and interactions [30–34]. The binding of COM crystals to collecting ducts, (3) lymphocytic and macrophage infil- apical microvilli facilitates their internalization into vesi- tration around injured tubular segments, (4) perivascular cles and leads to changes in the cytoskeletal structure cuffing with interstitial infiltrates, (5) loss of postglomer- [35]. The endocytosed COM crystals were shown to acti- ular capillaries, (6) crystal deposition around atrophic vate a number of immediate early genes (for example, tubular segments, (7) crystalline formations within tubu- c-myc, EGR-1, Nur-77), as well as plasminogen activator lar lumens of the proximal and distal tubules as well inhibitor, platelet-derived growth factor, and connective as the collecting ducts, (8) dilation of Bowman’s space tissue growth factor [36]. The particular combination of resulting in glomerular cysts, (9) flocculent material in activated genes was believed to favor the accumulation dilated Bowman’s space suggestive of proteinuria, and of ECM proteins and the stimulation of fibroblast prolif- (10) increased matrix material in the interstitial space eration, suggesting a role for COM crystals in the devel- (Evan et al, manuscript in preparation). opment of interstitial fibrosis. Bikunin is a component The progression of various forms of renal disease is of a family of serine proteases that appears to inhibit more rapid in men than in women [25, 26]. Men also have the crystallization of oxalate. The expression a higher incidence of calcium oxalate nephrolithiasis, of bikunin is reported to be significantly increased in but the reasons for this are unclear. In a rat model of cultured epithelial cells exposed to oxalate, suggesting urolithiasis, it has been shown that androgens increase that this may be a protective response to the nephrotoxic and estrogens decrease urinary oxalate excretion, plasma effects of oxalate [37, 38]. oxalate concentration, and kidney calcium oxalate crys- We used a cDNA array (Clontech, Palo Alto, CA, tal deposition. These findings may partly explain why USA) to monitor changes in gene expression in cultured nephrolithiasis is a predominantly male disease [27]. human renal epithelial (NHK-C) and African green Sex hormones per se may be important determinants monkey kidney epithelial (BSC-1) cells exposed to DHA of the greater susceptibility of male kidneys to progressive crystals (Wang et al, manuscript in preparation). The renal injury. Potential mechanisms for this include effects array contains 588 cDNAs associated with various physi- of sex hormones on matrix accumulation as well as inter- ological and pathological processes. Our findings suggest action of sex hormones with growth factors such as TGF-␤. that the binding of DHA crystals to cultured cells and In glomerulosclerosis, synthesized by mesan- subsequent pattern of gene expression is similar to that gial cells contribute to the pathological changes. It has seen with COM crystals. These observations and the been reported that estradiol in micromolar concentra- availability of the APRT-deficient mouse model (in tions suppressed total collagen, , and type which the severity of the disease can be experimentally IV collagen synthesis, and decreased steady-state mRNA modulated) suggest that investigation of DHA lithiasis levels for type I collagen in mesangial cells [28]. The may provide insights into the molecular pathophysiology accumulation of glomerular ECM may contribute to the of the more common stone diseases. development of glomerular sclerosis. In contrast, testos- terone had no effect on collagen synthesis by mesangial ACKNOWLEDGMENTS cells. More recently, it has been shown that estradiol This work was supported by National Institutes of Health grants inhibits gene expression by antagonizing the actions of DK38185, ES05652, and ES 06096. A portion of this work was pre- TGF-␤ [29]. sented in abstract form (Wang et al, Am J Hum Genet 63:183, 1998). Li Wang was supported by a National Insitutes of Health Training Grant (PHS T32 HD07373). We gratefully acknowledge the assistance Cell culture models for kidney stone disease of Dr. Sudhanshu Raikwar in the preparation of this manuscript. The present study highlights the complexity of investi- Reprint requests to Amrik Sahota, Ph.D., Department of Genetics, gating gene expression changes in whole kidneys. The Nelson Biological Laboratories, Rutgers University, 604 Allison Road, observed changes may occur in tubular epithelial cells, Piscataway, New Jersey 08854-8082, USA. interstitial fibroblasts, or infiltrating cells, and the expres- E-mail: [email protected] sion of several genes may be altered in a given cell or tissue. To dissect these change at the cellular level and to APPENDIX assess the physiological and pathological importance of Abbreviations used in this article are: ADAMTS-1, a disintegrin the various genes in the development and progression and metalloproteinase with thrombospondin motifs; APRT, adenine of DHA nephrolithiasis, we are investigating the effects phosphoribosyltransferase; BLAST, Basic Local Alignment Search of DHA crystals on gene expression in different types of Tool; C10, a macrophage inflammatory protein; CCKAR, cholecystoki- nin-A receptor; COM, calcium oxalate monohydrate; DD-PCR, differ- cultured kidney cells. ential display PCR; DHA, 2,8-dihydroxyadenine; ECM, extracellular Calcium oxalate monohydrate (COM) is the most matrix; EST, expressed sequence tag; Gla, ␥-carboxyglutamic acid; common cause of renal stone disease. COM is injurious IMPT-1, imprinted multimembrane-spanning polyspecific transporter- like gene 1; KAP, kidney androgen-regulated protein; LOX, lysyl oxi- to cultured renal epithelial cells, and this system has been dase; MGP, matrix Gla protein; NCBI, National Center for Biotechnol- widely used to model the early events in crystal–cell ogy Information; and OATP, organic anion transporter. 536 Wang et al: Alterations in renal gene expression

REFERENCES Molecular structure of the mouse CCK-A receptor gene. Biochem Biophys Res Commun 236:630–635, 1997 1. Simmonds HA, Sahota A, Van Acker KJ: Adenine phosphoribo- 19. Dao D, Frank D, Qian N, O’Keefe D, Vosatka RJ, Walsh syltransferase deficiency and 2,8-dihydroxyadenine lithiasis, in The CP, Tycko B: IMPT1, an imprinted gene similar to polyspecific Metabolic and Molecular Bases of Inherited Disease (7th ed), edited transporter and multi-drug resistance genes. Hum Mol Genet by Scriver CR, Beaudet AL, Sly WS, Valle D, New York, 7:597–608, 1998 McGraw-Hill, 1995, pp 1707–1723 20. Meseguer A, Catterall JF: Mouse kidney androgen-regulated 2. Engle SJ, Stockelman MG, Chen J, Boivin G, Yum M, Davies protein messenger ribonucleic acid is expressed in the proximal PM, Ying MY, Sahota A, Simmonds HA, Stambrook PJ, Tisch- convoluted tubules. Mol Endocrinol 1:535–541, 1987 field JA: Adenine phosphoribosyltransferase-deficient mice de- 21. Meseguer A, Catterall JF: Effects of pituitary hormones on the velop 2,8-dihydroxyadenine nephrolithiasis. Proc Natl Acad Sci cell-specific expression of the KAP gene. Mol Cell Endocrinol USA 93:5307–5313, 1996 89:153–162, 1992 3. Stockelman MG, Lorenz JN, Smith FN, Boivin G, Sahota A, 22. Sole E, Calvo R, Obregon MJ, Meseguer A: Effects of thyroid Simmonds HA, Stambrook PJ, Tischfield JA: Chronic renal fail- hormone on the androgenic expression of KAP gene in mouse ure in a mouse model of human adenine phosphoribosyltransferase kidney. Mol Cell Endocrinol 119:147–159, 1996 deficiency. Am J Physiol 275:F154–F163, 1998 23. Bergwerk AJ, Shi X, Ford AC, Kanai N, Jacquemin E, Burk 4. Stambrook PJ, Shao C, Stockelman M, Boivin G, Engle SJ, RD, Bai S, Novikoff PM, Stieger B, Meier PJ, Schuster VL, Tischfield JA: APRT: A versatile in vivo resident reporter of Wolkoff AW: Immunologic distribution of an organic anion trans- local mutation and loss of heterozygosity. Environ Mol Mutagen port protein in rat liver and kidney. Am J Physiol 271:G231–G238, 28:471–482, 1996 1996 5. Corton JC, Gustafsson J: Increased efficiency in screening large 24. Lu R, Kanai N, Bao Y, Wolkoff AW, Schuster VL: Regulation numbers of cDNA fragments generated by differential display. of renal OATP mRNA expression by testosterone. Am J Physiol Biotechniques 22:802–810, 1997 270:F332–F337, 1996 6. Orlofsky A, Lin EY, Prystowsky MB: Selective induction of 25. Neugarten J, Silbiger SR: Effects of sex hormones on mesangial the ␤-chemokine C10 by IL-4 in mouse macrophages. J Immunol cells. Am J Kidney Dis 26:147–151, 1995 152:5084–5091, 1994 26. Silbiger SR, Neugarten J: The impact of gender on the progres- 7. Hogaboam CM, Chensue SW, Steinhauser ML, Huffnagle GB, sion of chronic renal disease. Am J Kidney Dis 25:515–533, 1995 Lukacs NW, Strieter RM, Kunkel SL: Alteration of the cytokine 27. Fan J, Chandhoke PS, Grampsas SA: Role of sex hormones in phenotype in an experimental lung granuloma model by inhibiting experimental calcium oxalate nephrolithiasis. J Am Soc Nephrol nitric oxide. J Immunol 159:5585–5593, 1997 10(Suppl 14):S376–S380, 1999 8. Kuno K, Matsushima K: ADAMTS-1 protein anchors at the extra- 28. Kwan G, Neugarten J, Sherman M, Ding Q, Fotadar U, Lei J, cellular matrix through the thrombospondin type I motifs and its Silbiger S: Effects of sex hormones on mesangial cell proliferation spacing region. J Biol Chem 273:13912–13917, 1998 and collagen synthesis. Kidney Int 50:1173–1179, 1996 9. Kuno K, Kanada N, Nakashima E, Fujiki F, Ichimura F, Matsu- 29. Lei J, Silbiger S, Ziyadeh FN, Neugarten J: Serum-stimulated shima K: Molecular cloning of a gene encoding a new type of ␣1 type IV collagen gene transcription is mediated by TGF-␤ and metalloproteinase-disintegrin family protein with thrombospondin inhibited by estradiol. Am J Physiol 274:F252–F258, 1997 motifs as an inflammation associated gene. J Biol Chem 272:556– 30. Lieske JC, Norris R, Swift H, Toback FG: Adhesion, internaliza- 562, 1997 tion and metabolism of calcium oxalate monohydrate crystals by 10. Desmouliere A, Darby I, Costa AM, Raccurt M, Tuchweber renal epithelial cells. Kidney Int 52:1291–1301, 1997 B, Sommer P, Gabbiani G: deposition, lysyl 31. Verkoelen CF, Van der Boom BG, Houtsmuller AB, Schroder oxidase expression, and myofibroblastic differentiation during the FH, Romlin JC: Increased calcium oxalate monohydrate crystal initial stages of cholestatic fibrosis in the rat. Lab Invest 76:765–778, binding to injured renal tubular epithelial cells in culture. Am J Physiol 274:F958–F965, 1998 1997 32. Thamilselvan S, Khan SR: Oxalate and calcium oxalate crystals 11. Wakasaki H, Ooshima A: Synthesis of lysyl oxidase in experimen- are injurious to renal epithelial cells: Results of in vivo and in vitro tal hepatic fibrosis. Biochem Biophys Res Commun 166:1201–1204, studies. J Nephrol 11(Suppl 1):66–69, 1998 1990 33. Thamilselvan S, Hackett RL, Khan SR: Cells of proximal and 12. Di Donato A, Ghiggeri GM, Di Duca M, Jivotenko E, Acinni distal tubular origin respond differently to challenges of oxalate R, Campolo J, Ginevri F, Gusmano R: Lysyl oxidase expression and calcium oxalate crystals. J Am Soc Nephrol 10(Suppl 14):S452– and collagen cross-linking during chronic adriamycin nephropathy. S456, 1999 Nephron 76:192–200, 1997 34. Khan SR, Byer KJ, Thamilselvan S, Hackett RL, McCormack 13. Luo G, D’souza R, Jogue D, Karsenty G: The matrix Gla protein WT, Benson NA, Vaughn KL, Erdos GW: Crystal-cell interaction gene is a marker of the chondrogenesis cell lineage during mouse and apoptosis in oxalate-associated injury of renal epithelial cells. development. J Bone Miner Res 10:325–334, 1995 J Am Soc Nephrol 10(Suppl 14):S457–S463, 1999 14. Luo G, Ducy P, McKee MD, Pinero GJ, Loyer E, Behringer RR, 35. Lieske JC, Swift H, Martin T, Patterson B, Toback FG: Renal Karsenty G: Spontaneous calcification of and cartilage in epithelial cells rapidly bind and internalize calcium oxalate mono- mice lacking matrix GLA protein. Nature 386:78–81, 1997 hydrate crystals. Proc Natl Acad Sci USA 91:6987–6991, 1994 15. Fraser JD, Price PA: Lung, heart, and kidney express high levels 36. Hammes MS, Lieske JC, Pawar S, Spargo BS, Toback FG: Calcium of mRNA for the vitamin K-dependent matrix Gla protein. J Biol oxalate monohydrate crystals stimulate gene expression in renal Chem 263:11033–11036, 1988 epithelial cells. Kidney Int 42:501–509, 1995 16. Cancela ML, Hu B, Price PA: Effect of cell density and growth 37. Iida S, Peck AB, Johnson-Tardieu J, Moriyama M, Glenton factors on matrix Gla protein expression by normal rat kidney PA, Byer KJ, Khan SR: Temporal changes in mRNA expression cells. J Cell Physiol 171:125–134, 1997 for bikunin in the kidneys of rats during calcium oxalate nephroli- 17. Zhao J, Warburton D: Matrix Gla protein gene expression is thiasis. J Am Soc Nephrol 10:986–996, 1999 induced by transforming growth factor-␤ in embryonic lung culture. 38. Iida S, Peck AB, Byer KJ, Khan SR: Expression of bikunin mRNA Am J Physiol 273:L282–L287, 1997 in renal epithelial cells after oxalate exposure. J Urol 162:1480– 18. Lacourse KA, Lay JM, Swanberg LJ, Jenkins C, Samuelson LC: 1486, 1999