Differentiated human podocytes endogenously express ... - CiteSeerX

2 downloads 0 Views 242KB Size Report
Nov 25, 2003 - David A. Gillatt,5 David O. Bates,1 and Steven J. Harper1,4 ...... Harper SJ, Bailey E, McKeen CM, Stewart AS, Pringle JH, Feehally.
Am J Physiol Renal Physiol 286: F767–F773, 2004. First published November 25, 2003; 10.1152/ajprenal.00337.2003.

Differentiated human podocytes endogenously express an inhibitory isoform of vascular endothelial growth factor (VEGF165b) mRNA and protein Tai-Gen Cui,1,2 Rebecca R. Foster,1 Moin Saleem,3 Peter W. Mathieson,4 David A. Gillatt,5 David O. Bates,1 and Steven J. Harper1,4 1

Microvascular Research Laboratories, Department of Physiology, Preclinical Veterinary School, University of Bristol, Bristol BS2 8EJ; 3Childrens’ and 4Academic Renal Unit, University of Bristol, and 5Bristol Urological Institute, Southmead Hospital, Bristol BS10 5NB, United Kingdom; and 2Institute of Nephrology, First Teaching Hospital, University of Beijing, Beijing 10034, Peoples’ Republic of China Submitted 22 September 2003; accepted in final form 21 November 2003

Cui, Tai-Gen, Rebecca R. Foster, Moin Saleem, Peter W. Mathieson, David A. Gillatt, David O. Bates, and Steven J. Harper. Differentiated human podocytes endogenously express an inhibitory isoform of vascular endothelial growth factor (VEGF165b) mRNA and protein. Am J Physiol Renal Physiol 286: F767–F773, 2004. First published November 25, 2003; 10.1152/ajprenal.00337. 2003.—Despite production by podocytes of the proangiogenic molecule vascular endothelial growth factor-A (VEGF), the glomeruli are not sites of angiogenesis. We recently described mRNA expression of an inhibitory splice variant of VEGF (VEGF165b) in normal kidney (Bates DO, Cui TG, Doughty JM, Winkler M, Sugiono M, Shields JD, Peat D, Gillatt D, and Harper SJ. Cancer Res 62: 4123–4131, 2002). Available anti-VEGF antibodies do not distinguish stimulatory from inhibitory VEGF families. To assess the production of VEGF165 (stimulatory) and VEGF165b (inhibitory) isoforms by human podocytes, we examined both primary cultured and conditionally immortalized human podocytes using family- and isoform-specific RT-PCR. In addition, VEGF protein production was analyzed in podocytes, using isoform-specific double-strand small-interference RNAs (siRNA). RT-PCR demonstrated the production of VEGF189 mRNA by podocytes of both phenotypes. In contrast, on differentiation there was a splicing change from VEGF165 to VEGF165b mRNA. In addition, VEGF protein in the supernatant of conditionally immortalized, differentiated podocytes was reduced by VEGF165b siRNA to 20 ⫾ 11% of the level of mock-transfected cells (P ⬍ 0.01). No reduction was seen with mismatch siRNA. Moreover, there was no reduction in VEGF protein concentration in the supernatant of primary cultured, dedifferentiated human podocytes (109 ⫾ 8% of mismatch siRNA, P ⬎ 0.1). In conclusion, differentiated but not dedifferentiated human podocytes secrete significant amounts of VEGF165b protein. It is possible that this may explain the paradox of high VEGF production in the glomerulus but no angiogenesis. Furthermore, the existence of this splicing switch in relation to podocyte phenotype suggests that alternative splicing of the VEGF pre-RNA is a regulated process that is open to manipulation and therefore could be a target for novel cancer therapies. angiogenesis; small-interference RNA; splicing VASCULAR ENDOTHELIAL GROWTH factor-A (VEGF) is the most potent and dominant proangiogenic factor in physiological and pathological angiogenesis and, as such, the overexpression of VEGF is believed to be a crucial pathophysiological step in many diseases, including cancer, atherosclerosis, arthritis, and psoriasis (8, 11). The multiple isoforms of VEGF stimulate

Address for reprint requests and other correspondence: D. O. Bates, Microvascular Research Laboratories, Dept. of Physiology, Preclinical Veterinary School, Univ. of Bristol, Southwell St., Bristol BS2 8EJ, UK (E-mail: [email protected]). http://www.ajprenal.org

endothelial cell proliferation, migration, and increased microvascular permeability by activation of VEGF receptor 1 (flt-1) and VEGFR-2 (KDR/flk1). VEGFR-1 also exists in a soluble form, sVEGFR-1 (sFlt), which is inhibitory when bound to free VEGF (12). Neuropilin-1 (NP-1) facilitates the binding of VEGF165 to VEGFR-2 (14). VEGF receptor signaling is incompletely understood but is rapidly becoming characterized both in vitro and in vivo (3, 4). Available data suggest that increased permeability is mediated by intracellular calcium (3) but compliance and mitogenesis via MAPK (4). VEGF-A is differentially spliced from eight exons, resulting in different proteins named by their amino acid number: VEGF121, VEGF165, (the dominant isoform) VEGF189, etc. (Fig. 1A). We recently identified an mRNA encoding a novel isoform VEGF165b (2) in normal kidney and other tissues. VEGF165b has a 3⬘-splicing structure that predicts a novel COOH-terminal peptide sequence. Usually, the COOH terminus of VEGF consists of six amino acids encoded by the first 18 nucleotides of exon 8. In VEGF165b, that is replaced by six amino acids coded for by an 18-nucleotide open-reading frame formed from a more distal splice site (DSS) selection in exon 8. We initially termed this new open-reading frame “exon 9” (2). However, there is no true “intron” between the two 18-base sequences and this new open-reading frame should more correctly be termed “exon 8b.” Thus DSS selection, “exon 8b,” in place of proximal splice site (PSS) selection, “exon 8a,” predicts the existence of a family of sister isoforms (Fig. 1B) with a novel COOH terminus. We termed the putative family VEGFxxxb where xxx is the amino acid number. The VEGFxxxb isoforms retain receptor binding and dimerization domains (exons 3&4) (2) and are the same size as their sister molecules, so are identified as the same product on most published blotting and RT-PCR analysis, which perhaps explains their previous elusiveness. Glomerular VEGF remains enigmatic. In situ hybridization (1) and immunohistochemical analyses (5) define the podocyte as the site of glomerular VEGF production in vivo. Glomerular endothelial cells (10) and podocytes (13, 16) express VEGF receptors. A paracrine and perhaps, more intriguingly, an autocrine role would therefore seem possible. However, despite high-level production of VEGF by podocytes and the expression of VEGFR-2 by glomerular endothelial cells, angiogenesis is not a feature of the normal glomerulus. This has The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

0363-6127/04 $5.00 Copyright © 2004 the American Physiological Society

F767

F768

PODOCYTES EXPRESS AN ANTIANGIOGENIC VEGF ISOFORM

Fig. 1. COOH-terminal exon structure of stimulatory and inhibitory vascular endothelial growth factor (VEGF) families. PSS and DSS, proximal and distal COOH-terminal splice site selection, respectively.

previously been explained by two proposals: first, the potential difficulty of podocyte-derived VEGF to target glomerular endothelial cell VEGF receptors, which microanatomically are upstream of the site of production. However, VEGF is clearly able to cross this filtration stream because immunogold electron microscopic studies showed a clear concentration gradient of labeled VEGF particles from glomerulus to endothelial cells, with VEGF being clearly apparent on the endothelial side of the glomerular barrier (13). Second, the expression of VEGF inhibitors within the normal glomerulus, for example, sFLT (27) or angiopoietin (22), prevents angiogenesis. While investigating the apparent paradox, we identified VEGF165b. In contrast to all exon 8 proximal splicing variants, VEGF165b inhibits VEGF165-mediated endothelial cell proliferation and migration in vitro and vasodilatation ex vivo (2). Furthermore, there is preliminary evidence that VEGF165b is not angiogenic in vivo (Woolard JW, unpublished observations). VEGF165b is potentially, therefore, an endogenous inhibitor of VEGF-mediated angiogenesis in the glomerulus. However, there is currently no evidence that VEGF165b is endogenously expressed by any human cells. There are no currently available antibodies that can distinguish proximal from distal COOH-terminal splice forms, and previously published work used recombinant VEGF165b, overexpressed in a mammalian expression system, so endogenous VEGF165b protein expression has not yet been shown in any tissue. Furthermore, we have not shown that the mRNA splice form of VEGF that would result in VEGF165b is actually produced in the glomerulus or in any endogenous glomerular cell type. Previous expression work used homogenized renal cortex, and, because isoform-specific in situ hybridization probes have not been successfully developed, the location of expression in the cortex is still uncertain. Although it is not clear whether the renal cortex cell type that expresses VEGF165b is the podocyte, the endothelial cells and mesangial cells rarely if ever express VEGF. Because no antibodies are available that distinguish between the isoforms, we used an interference RNA knockdown technique to determine whether cells endogenously produce VEGF protein. This technique relies on the recently described smallinterference RNAs (siRNA), double-strand 19-bp stretches of AJP-Renal Physiol • VOL

RNA that use an endogenous RNA degradation mechanism to regulate gene expression. The double-strand RNA is transfected into cells, where it binds to the RNA-induced silencing complex (RISC). RISC uses the siRNA to recognize endogenous mRNA sequences that contain the complementary sequence to the siRNA and degrade the endogenous mRNA (19). To test the hypothesis that VEGF165b protein and mRNA are secreted by human podocytes, we examined isoform-specific mRNA expression by RT-PCR and designed an RNA interference knockdown technique to determine whether VEGF165b protein is endogenously produced by human podocytes of both proliferating dedifferentiated and growth-arrested, differentiated phenotypes. METHODS

Podocytes Proliferating dedifferentiated podocytes derive from two sources: first, from primary culture from unipolar renal tumor nephrectomy samples collected with local Ethical Committee approval; second, from a conditionally immortalized human podocyte cell line. Podocytes were isolated from the normal pole of specimens by standard sieving techniques (17). These cells have been previously characterized as cytokeratin and WT-1 positive (immunofluorescence); VEGF, WT-1 and synaptopodin mRNA positive (RT-PCR); and von Willebrand factor, CD45 and smooth muscle myosin negative (RT-PCR), excluding endothelial cell, leucocyte, or mesangial cell contamination, respectively, as we previously described (16, 20). This phenotype was confirmed by regular sampling. Differentiated, growth-arrested podocytes were derived from a cell line conditionally transformed from normal human podocytes with a temperature-sensitive mutant of immortalized SV-40 T-antigen as described elsewhere (21). At the “permissive” temperature of 33°C, the SV-40 T-antigen is active and allows the cells to proliferative rapidly. Thermoswitching the cells to the “nonpermissive” temperature of 37°C inactivates the T-antigen, and the cells become growth arrested and differentiate to express antigens appropriate to in vivo arborized podocytes. Cells were grown for a period of 16 days at 37°C to ensure growth arrest and differentiation. Molecular Biology Molecular biology tools were purchased from Invitrogen unless otherwise stated.

286 • APRIL 2004 •

www.ajprenal.org

F769

PODOCYTES EXPRESS AN ANTIANGIOGENIC VEGF ISOFORM

Table 1. Primer sequences 3⬘

Exon 7a 3⬘UTR Exon 4 Exon 8a (previously called 8) Exon 8b (previously called 9) T7 siRNA ⫺ VEGF165b sense T7 siRNA ⫺ VEGF165b anti-sense T7 siRNA ⫺ VEGF165b sense mismatch T7 siRNA ⫺ VEGF165b anti-sense mismatch T7 siRNA ⫺ common VEGF sense T7 siRNA ⫺ common VEGF anti-sense T7-siRNA ⫺ common VEGF sense mismatch T7-siRNA ⫺ common VEGF anti-sense mismatch

GTA ATG GAG TCA TCA AGA AAC AGA AAC CTG TAG CTG TAG

5⬘

AGC GAT ATG CCG GTC GAT GTA GAT GTA TAG GAG TAG GAG

TTG CCG AGC CCT TTT CTG CTT CTG CTA GAA AGA GAA AGA

TAC TAT TTC CGG CCT CAA GCA GTA CCA GCT TGA GGA TGT

AAG CAG CTA CTT GGT GTA GAT GTA GAT CAT GCT CAT CCT

ATC TCT CAG GTC GAG CGT CTC CGT CTC CTC TCC CTC TCC

CGC TTC CAC ACA AGA TCT TCT TCT TCT TCC TAC TCC TAC

AGA CG CT T TCT ATA ATA ATA ATA TAT TAT TAT TAT

GCA GTG GTG GTG GTG AGT AGT AGT AGT

AGT AGT AGT AGT GAG GAG GAG GAG

CGT CGT CGT CGT TCG TCG TCG TCG

ATT ATT ATT ATT TAT TAT TAT TAT

A A A A TA TA TA TA

Mismatch sequences are underlined in bold. UTR, untranslated region; siRNA, small-interference RNA; VEGF, vascular endothelial growth factor.

RT-PCR

siRNA Synthesis

RT-PCR was performed on dedifferentiated proliferating cultured podocytes (PCP), growth-arrested differentiated conditionally immortalized podocytes (DCIP), and sieved human glomeruli. RT-PCR to differentiate between exon 8a- and exon 8b-containing isoforms within the same sample in the same reaction was performed as previously described (2) using primers specific to exon 7a and 3⬘ untranslated region (UTR). Exon-specific RT-PCR for exon 8a (PSS)containing angiogenic mRNA isoforms was amplified using primers to exon 4 and exon 8a. Exon 8b (DSS)-containing inhibitory isoforms were detected using exon 4 and exon 8b sequences. Primer sequences are shown in Table 1. mRNA was extracted from ⬃10 sieved glomeruli, primary cultured, or conditionally immortalized podocytes from confluent 75-ml culture flasks using standard techniques (6). Six percent of the RNA was reverse transcribed using MMLV reverse transcriptase and poly-d(T) primer. One micromolar of each appropriate primer, 1.2 mM MgCl2, 200 ␮M dNTPs, and 1 unit of Taq (Abgene) were used. Reactions were cycled 35 times denaturing at 96°C for 30 s, annealing at 55°C for 30 s (exon 8a and 3⬘UTR primers) or 65°C (exon 8b primers), and extension at 72°C for 60 s. Products were run on 2% agarose gels containing 0.5 ␮g/ml ethidium bromide and visualized under a UV transluminator.

Double-strand siRNA was synthesized using in vitro transcription (Fig. 2). Briefly, primers designed to degrade the target sequence were used to make double-strand DNA encoding the T7 RNA polymerase promoter. Single-strand RNA was made from each primer pair, denatured at 96°C, and annealed by cooling to form dsRNA. This was transfected into human embryonic kidney (HEK) cells or podocytes using lipofectamine. Double-strand siRNAs specific for exon 8bcontaining mRNA (i.e., across exon 7–8b boundary) were made in this way, but unfortunately, it was not possible to construct a similar siRNA for exon 8a-containing VEGF mRNA species (e.g., VEGF165 mRNA) because the optimal C-(N)19-G sequence across the splice site of exons 7–8a was not present. siRNA was therefore synthesized for an area of the mRNA common to all isoforms, the exon3/exon4 boundary. Primers were synthesized for VEGF165b sense (forward and reverse), VEGF165b anti-sense (forward and reverse), VEGF165b sense with a 2-bp mismatch (reverse only), and VEGF165b anti-sense with a 2-bp mismatch (reverse only). Mismatch siRNA acted as a control for nonspecific degradation of secreted protein. For each pair of primers, 1) double-strand DNA templates were synthesized using 1 ␮M primer, 1.2 mM MgCl2, 200 ␮M dNTPs, and 1 unit of Taq

Fig. 2. Interference RNA. Primers designed to degrade the target sequence (a) are used to make double-strand DNA (b) encoding the t7RNA polymerase promoter. Single-strand RNA is made from each primer pair and then hybridized to form dsRNA (c). This is transfected (d) into cells where it binds with an RNA-induced silencing complex (RISC), which recognizes the target sequence and enables degradation of the endogenous mRNA.

AJP-Renal Physiol • VOL

286 • APRIL 2004 •

www.ajprenal.org

F770

PODOCYTES EXPRESS AN ANTIANGIOGENIC VEGF ISOFORM

(Abgene) in 1⫻ Taq buffer. Reactions were cycled five times denaturing at 94°C for 30 s, annealing at 37°C for 1 min, and extension at 72°C for 3 min. Products were phenol-chloroform extracted and pellets were resuspended in 20 ␮l diethylpyrocarbonate (DEPC) water. 2) Single-strand RNA was synthesized using 1 ␮M DNA templates, 20 U RNA guard, 200 ␮M rNTPs, and 80 units of T7 polymerase (Pharmacia) in transcription buffer. The above were incubated for 1 h at 37°C and then digested with 1 unit RNase-free DNAase (Promega) at 37°C to release single-strand RNA. Products were phenol-chloroform extracted and pellets were resuspended in 20 ␮l DEPC water, and a further round of phenol-chloroform extraction was performed and pellets were resuspended in annealing buffer (10 mM Tris䡠HCl, 100 mM NaCl). 3) Double-strand RNA was produced by mixing equal amounts of the appropriate two products, heating to 95°C for 5 min and cooling to room temperature. Aliquots of each double-strand RNA were run on a 1.5% agarose gel to confirm synthesis. Transfection Protocol and Assessment of Protein Production Cells were grown in six-well plates, each well seeded with 3 ⫻ 105 cells. Initial experiments were conducted on HEK cells transfected with pcDNA3-VEGF165b or pcDNA3-VEGF165 (2). Stable cell lines were produced using Geneticin selection. Stable or transiently transfected HEK cells and podocytes were transfected with siRNA using lipofectamine. The concentration of VEGF secreted into the media was assayed using a pan-VEGF ELISA (R&D Systems, Duo-set DY293). RESULTS

RT-PCR Consistent products were produced on repeated analysis. DCIP produced products consistent with the inhibitory isoform VEGF165b and the stimulatory isoform VEGF189 (Fig. 3A). We were unable to identify exon 8b (inhibitory)-containing VEGF mRNA isoforms in PCP, despite efficient mRNA extraction and reverse transcription as assessed by the presence of 8a products and efficient PCR, as seen in Fig. 3A. In contrast, the stimulatory VEGF isoform VEGF165 and VEGF189 were readily detected (Fig. 3B). RT-PCR for exon 8a- and exon 8b-containing isoforms using exon 7 and 3⬘UTR primers in DCIP and freshly isolated human glomeruli confirmed the presence of exon 8b-containing isoforms (Fig. 3C). Previous work from this laboratory showed that both exon 8b- and exon 8a-containing isoforms are found in freshly isolated human glomeruli (27). VEGF Secretion by Cells in Culture To determine whether cells in culture could secrete VEGF into the media, conditioned media was tested from a variety of cell lines using the commercially available ELISA. VEGF was not detected in conditioned media from HEK 293Q, Chinese hamster ovary, MCF7 breast cancer cells, fibroblasts, human umbilical vein endothelial cells, or human dermal microvascular endothelial cells. Very low concentrations (50–100 pg/ml) were detected in the conditioned media of A375 cells as previously described (23). Conditionally immortalized podocytes, on the other hand, expressed significant concentrations of VEGF during proliferation (3.1 ⫾ 0.14 ng/ml), after transfection (⬃800 pg/ml), and after differentiation (⬃150 pg/ml). We therefore used these cells to determine the effects of siRNA specific for VEGF165b on VEGF protein production. AJP-Renal Physiol • VOL

Fig. 3. A: exon (Ex)-specific RT-PCR on differentiated, growth-arrested conditionally immortalized podocytes (DCIP). Lanes 1-3: PCR for exon 8bcontaining isoforms. Lane 1: VEGF165b cDNA (positive control). Lane 2: exon 8b product, size compatible with VEGF165b. Lane 3: water control. Lanes 4 and 5: PCR for exon 8a-containing isoforms. Lane 4: VEGF165 cDNA (positive control). Lane 5: exon 8a product, size compatible with VEGF189. Lane 6: molecular weight ladder. B: exon-specific RT-PCR on proliferating primary cultured podocytes (PCP). Lanes 7 and 9: PCR for exon 8b-containing products. No products were detected. Lanes 8 and 10: PCR for exon 8acontaining products showed PCR products consistent with VEGF165 and VEGF189. C: RT-PCR for composite exon 8a and 8b products in DCIP (lane 11) and freshly harvested human glomeruli (lane 12). No 8a products were identified; however, exon 8b isoforms were identified in both.

siRNA Studies HEK cell line. To determine whether siRNA could specifically target VEGF165b protein expression, studies were carried out in HEK cells (which do not constitutively express VEGF) stably transfected with VEGF165b cDNA. VEGF165b siRNA significantly inhibited VEGF165b protein expression. VEGF165b protein in the supernatant of HEK cells transfected with VEGF165b cDNA was significantly lower when cotransfected with VEGF165b siRNA (39 ⫾ 0.9 ng/ml) than with mismatch siRNA (123 ⫾ 4 ng/ml, P ⬍ 0.007). In contrast, VEGF165 production was not affected in cells transfected with VEGF165 cDNA and VEGF165b siRNA (76.7 ⫾ 4 ng/ml compared with 80 ⫾ 4 ng/ml with mismatch siRNA; Fig. 4). Transient cotransfection of HEK cells with VEGF165b and siRNA showed that the inhibition was dose dependent (Fig. 5). The transfected cells produce large amounts of VEGF protein as the production is under the control of the cytomegalovirus promoter. Despite this, the siRNA reduced the production by over 200 ng/ml.

286 • APRIL 2004 •

www.ajprenal.org

PODOCYTES EXPRESS AN ANTIANGIOGENIC VEGF ISOFORM

Fig. 4. Small-interference (si)RNA inhibits VEGF165b but not VEGF165 protein expression in transfected human embryonic kidney (HEK) cells, VEGF expression of media of cells cotransfected with siRNA for VEGF165b, or a 2-base mismatch RNA and pcDNA3-VEGF165b (left) or pcDNA3-VEGF165 (right).

DCIP. DCIP expressed VEGF protein that was blocked by VEGF165b-specific siRNA (Fig. 6). Transfection of DCIP with VEGF165b siRNA reduced VEGF production to 26 ⫾ 10% compared with before transfection from 151 ⫾ 11 to 38 ⫾ 6.8 pg/ml (P ⬍ 0.001, paired t-test). In contrast, no reduction was seen with mismatch siRNA (from 157 ⫾ 6 to 190 ⫾ 31 pg/ml, P ⬎ 0.1, paired t-test), demonstrating that this effect was specific for VEGF165b and not a nonspecific effect on total protein production. Primary cultured podocytes. The VEGF concentration of media from PCP was 3.1 ⫾ 0.14 ng/ml (n ⫽ 15). Transfection of PCP with siRNA reduced the total VEGF production, irrespective of the sequence of the siRNA. However, transfection with pan-VEGF siRNA reduced the VEGF production to a significantly lower level (to 586 ⫾ 17 pg/ml) compared with that with the 2-bp mismatch siRNA (867 ⫾ 41 pg/ml; Fig. 7). A pan-VEGF siRNA could therefore significantly reduce the VEGF in primary PCP to 67% of control (P ⬍ 0.01, t-test). This was not seen with VEGF165b-specific siRNA. Transfection with VEGF165b siRNA reduced VEGF protein production to 928 ⫾ 65 pg/ml, which was not significantly different from that induced by transfection with 2-bp mismatch siRNA 855 ⫾

Fig. 5. VEGF165b siRNA dose dependently inhibits VEGF165b protein expression in transfected cells. VEGF concentration of media taken from HEK293 cells transfected with 2 ␮g VEGF165b cDNA (in expression vector pcDNA3) alone or with increasing amounts of double-strand siRNA directed across the exon 7 exon 8b splice site is shown. AJP-Renal Physiol • VOL

F771

Fig. 6. DCIP express VEGF protein that is blocked by VEGF165b-specific siRNA. Effect of transfecting siRNA for VEGF165b or a 2-base mismatch RNA on VEGF expression in the media of DCIP is shown.

69 pg/ml (109% of control). Primary cultured podocytes therefore express VEGF isoforms that do not contain DSS selection, i.e., they express angiogenic, stimulatory forms. The use of VEGF165-transfected cells, mismatch sequences, and a panVEGF siRNA confirms the specificity of the VEGF165b siRNA for sequence and target mRNA. DISCUSSION

Unipolar renal tumors demonstrate a paradox: one pole contains a highly angiogenic lesion expressing high levels of VEGF mRNA and protein (17, 26); the opposite, histologically and functionally normal pole also expresses high levels of VEGF mRNA and protein in the absence of new vessel formation. We showed that the VEGF165b mRNA (“exon 8b” inhibitory), normally expressed in whole kidney, is downregulated in renal cell carcinoma, in contrast to all other isoforms in all other tumors (8, 12). Investigation of exon 8b-containing isoforms in vivo at the protein level presents some difficulties because “exon 8b” distal COOH-terminal splice site-specific antibodies are unavailable. In situ hybridization will not differentiate between exon 8a (PSS) and exon 8b (DSS) mRNA species because the

Fig. 7. Freshly isolated, proliferating, dedifferentiated podocytes express VEGF protein that is not blocked by VEGF165b-specific siRNA. Effect of transfecting siRNA for VEGF165b, pan-VEGF, or 2-base mismatch RNA (pan or VEGF165b, respectively) on VEGF expression in media of proliferating dedifferentiated podocytes is shown.

286 • APRIL 2004 •

www.ajprenal.org

F772

PODOCYTES EXPRESS AN ANTIANGIOGENIC VEGF ISOFORM

difference between full-length mRNAs is only 18 bases. A single oligonucleotide (for instance, across exon 7–8b splice site) will not produce enough visible signal isotopically or after optimization of hapten-labeling and detection as we previously described (15). Furthermore, an exon 8b-containing sequence is present in both stimulatory and inhibitory VEGF families, as part of the 3⬘UTR and as coding sequence, respectively. We therefore approached this study using conventional RT-PCR and siRNA technologies to investigate VEGF expression of human podocytes in culture. The use of podocytes in vitro attracts two potential criticisms, the issues of purity and differentiation status. We did our utmost to ensure purity of our primary cultures and covered this issue elsewhere (16, 20). We also studied a conditionally transformed podocyte cell line that can be investigated in both dedifferentiated and differentiated phenotypes and is a pure population (18, 21). The data we present are the first evidence that human cells can endogenously express exon 8b-containing isoforms at the protein level and that they can regulate this expression during differentiation, presumably by regulation of splicing. In addition, this study provides new information concerning changes in podocyte VEGF splicing that relate to the level of differentiation. VEGF189 mRNA is expressed by both podocyte phenotypes. VEGF189 is avidly heparin bound and in vivo would tend to remain cell and glomerular basement membrane associated. However, to our surprise, a splicing alteration seems to occur in VEGF165/VEGF165b isoforms in association with podocyte differentiation status. There is a switch from exon 8b-containing inhibitory VEGF165b in differentiated podocytes to exon 8a-containing stimulatory VEGF165 in dedifferentiated podocytes. The VEGF glomerular literature is conflicting. Floege and colleagues (18) attempted to inhibit VEGF production in vivo in healthy animals using aptamers, failing to define any glomerular change after 3 wk of administration. However, targeted pan-VEGF (both stimulatory and inhibitory) inhibition and overexpression in podocytes have now shown that a balance of VEGF is required for normal glomerular well-being. A Cre-recombinase knockout of even a single gene copy leads to nephrotic syndrome, uremia, and death 9 wk postpartum, and complete knockout results in death a few hours postpartum (9). In a transgenic model, VEGF overexpression resulted in death a few days postpartum with renal hemorrhages (9). In addition, a recent study demonstrated that in vivo inhibition of VEGF with antibody or soluble receptor caused nephrotic syndrome in mature mice (24). In glomerular disease, aptamer VEGF inhibition in a model of glomerulonephritis characterized by endothelial cell damage impaired glomerular repair (18). In contrast, inhibition of VEGF in streptozotocin-induced diabetic animals produced a beneficial reduction in proteinuria (7). Some of this data may suggest that VEGF balance is important to glomerular well being and repair. However, overall these contradictory findings cannot be reconciled if VEGF is only considered as a proangiogenic, propermeability vasodilator. However, should the splicing changes we demonstrated in vitro be mirrored in vivo, some of the contradictions are explained. We hypothesize, therefore, that in healthy glomeruli in vivo there is a balance between COOH-terminal distal and proximal AJP-Renal Physiol • VOL

splice site selection. When podocytes dedifferentiate or are injured in glomerular disease, there is a switch from inhibitory exon 8b-containing VEGF isoform expression to stimulatory exon 8a-containing VEGF isoform expression. This would be an appropriate physiological response because exon 8a-containing VEGF isoforms are well-characterized survival factors for endothelial cells. In addition, we recently reported the ability of VEGF165 to act as an autocrine survival factor for podocytes themselves via a PI3 kinase-dependent pathway (13). Such a splicing event, therefore, in the context of glomerular injury would be beneficial for both glomerular endothelial cells and podocytes. However, we also hypothesize that this survival response for podocytes and endothelial cells occurs at the expense of proteinuria. If we accept this hypothesis, some conflicting findings are resolved; for example, the aptamer studies that failed to demonstrate any change in health by aptamer administration but produced a detrimental effect of glomerular repair when administered in a model of glomerulonephritis in which endothelial damage predominated (18). Aptamers are sequences of nucleic acids that bind to specific areas of proteins because of their three-dimensionsal shape not because of their sequence. They are very specific; indeed the aptamer used in these studies was specific for exon 7 in VEGF165. Substitution of exon 8a by exon 8b (as in VEGF165b) would result in significant conformational change to the terminal part of exon 7 because at least one disulfide bond is lost (2). This aptamer may well therefore have no effect in healthy glomeruli in which exon 8b isoforms predominate but would have a potent effect on glomerular endothelial cell survival in glomerular disease in which exon 8a isoforms predominate. Furthermore, the detrimental effect of VEGF inhibition in the above animal model contrasts with the beneficial effect of VEGF inhibition in streptozotocininduced diabetic nephropathy (7), in which we presume endothelial cell damage is not a major feature but that the proteinuria induced by the splicing change is amenable to manipulation. The most recent study to shed light on glomerular VEGF biology (24) contrasted with both the above studies by showing that inhibition of VEGF (a propermeability agent) caused an increase in microvascular permeability and precipitated nephrotic syndrome. This is indeed an unexpected finding. These authors claim that the inhibitory agents affect primarily the circulating VEGF pool. This may be the case but will require further study. What is clear is that many tissues seem to produce both stimulatory and inhibitory VEGF isoforms, at least at the mRNA level, and that within normal tissues a balance may exist (2). This may also be the case with the circulating VEGF pool, the balance of which may bear no relationship to the balance within a particular microanatomic site. Interpretation of studies based on systemic administration of VEGF inhibitors may always, therefore, be fraught with difficulty because it is impossible to know which pool or pools (circulating and/or tissue, e.g., glomerular) are inhibited and the site at which this inhibition occurs. The VEGF splicing change we demonstrated is identical to that which tissues undergo during malignant change. We initially demonstrated this in renal cell carcinoma and more recently showed the same change in prostate cancer; indeed in this lesion the splicing change occurs at the prostatic intraepithelial neoplasia (carcinoma in situ) stage (25). The detailed

286 • APRIL 2004 •

www.ajprenal.org

PODOCYTES EXPRESS AN ANTIANGIOGENIC VEGF ISOFORM

control of VEGF splicing is one of the few areas of VEGF that have been poorly investigated. A fuller understanding of the VEGF splicing mechanism that podocytes use may therefore bring benefits to glomerular pathology but also to angiogenesis-based disease. In summary, VEGF165b protein, a VEGF isoform that is downregulated in renal and prostate cancer, inhibits VEGF165mediated endothelial cell proliferation and migration, and vasodilatation, is not angiogenic in vivo, and is the dominant form of VEGF expressed by differentiated human podocytes. It is not expressed, however, by PCP.

11.

12. 13.

14.

ACKNOWLEDGMENTS The authors thank Dr. K. Zavitz for Martini-inspired experimental design.

15.

GRANTS This work was supported by the Association for International Cancer Research Project Grant 02–053, The Showering Fund SF61, The Luff Cancer Fund, and The Richard Bright VEGF Research Trust. S. J. Harper is supported by Wellcome Trust Grant 057936/Z/99. D. O. Bates is supported by British Heart Foundation Grant BB-2000003.

16.

17.

REFERENCES 1. Bailey E, Bottomley MJ, Westwell S, Pringle JH, Furness PN, Feehally J, Brenchley PE, and Harper SJ. Vascular endothelial growth factor mRNA expression in minimal change, membranous, and diabetic nephropathy demonstrated by non-isotopic in situ hybridisation. J Clin Pathol 52: 735–738, 1999. 2. Bates DO, Cui TG, Doughty JM, Winkler M, Sugiono M, Shields JD, Peat D, Gillatt D, and Harper SJ. VEGF165b, an inhibitory splice variant of vascular endothelial growth factor, is down-regulated in renal cell carcinoma. Cancer Res 62: 4123–4131, 2002. 3. Bates DO and Curry FE. Vascular endothelial growth factor increases microvascular permeability via a Ca2⫹-dependent pathway. Am J Physiol Heart Circ Physiol 273: H687–H694, 1997. 4. Bates DO, Heald RI, Curry FE, and Williams B. Vascular endothelial growth factor increases Rana vascular permeability and compliance by different signalling pathways. J Physiol 533: 263–272, 2001. 5. Brown LF, Berse B, Tognazzi K, Manseau EJ, Van de Water L, Senger DR, Dvorak HF, and Rosen S. Vascular permeability factor mRNA and protein expression in human kidney. Kidney Int 42: 1457– 1461, 1992. 6. Chomczynscki P and Sacchi N. Single-step method of RNA isolation by acid quanidinium thiocyanate phenol chloroform extraction. Anal Biochem 162: 156–159, 1987. 7. De Vriese AS, Tilton RG, Elger M, Stephan CC, Kriz W, and Lameire NH. Antibodies against vascular endothelial growth factor improve early renal dysfunction in experimental diabetes. J Am Soc Nephrol 12: 993– 1000, 2001. 8. Dvorak HF. Vascular permeability factor/vascular endothelial growth factor: a critical cytokine in tumor angiogenesis and a potential target for diagnosis and therapy. J Clin Oncol 20: 4368–4380, 2002. 9. Eremina V, Sood M, Haigh J, Nagy A, Lajoie G, Ferrara N, Gerber HP, Kikkawa Y, Miner JH, and Quaggin SE. Glomerular-specific alterations of VEGF-A expression lead to distinct congenital and acquired renal diseases. J Clin Invest 111: 707–716, 2003. 10. Feng D, Nagy JA, Brekken RA, Pettersson A, Manseau EJ, Pyne K, Mulligan R, Thorpe PE, Dvorak HF, and Dvorak AM. Ultrastructural localization of the vascular permeability factor/vascular endothelial growth factor (VPF/VEGF) receptor-2 (FLK-1, KDR) in normal mouse

AJP-Renal Physiol • VOL

18.

19. 20. 21.

22.

23.

24.

25.

26.

27.

F773

kidney and in the hyperpermeable vessels induced by VPF/VEGF-expressing tumors and adenoviral vectors. J Histochem Cytochem 48: 545–556, 2000. Ferrara N. Role of vascular endothelial growth factor in physiologic and pathologic angiogenesis: therapeutic implications. Semin Oncol 29: 10– 14, 2002. Ferrara N and Davis-Smyth T. The biology of vascular endothelial growth factor. Endocr Rev 18: 4–25, 1997. Foster RR, Hole R, Anderson K, Satchell SC, Coward RJ, Mathieson PW, Gillatt DA, Saleem MA, Bates DO, and Harper SJ. Functional evidence that vascular endothelial growth factor may act as an autocrine factor on human podocytes. Am J Physiol Renal Physiol 284: F1263– F1273, 2003. Fuh G, Garcia KC, and de Vos AM. The interaction of neuropilin-1 with vascular endothelial growth factor and its receptor flt-1. J Biol Chem 275: 26690–26695, 2000. Harper SJ, Bailey E, McKeen CM, Stewart AS, Pringle JH, Feehally J, and Brown T. A comparative study of digoxigenin, 2,4-dinitrophenyl, and alkaline phosphatase as deoxyoligonucleotide labels in non-radioisotopic in situ hybridisation. J Clin Pathol 50: 686–690, 1997. Harper SJ, Xing CY, Whittle C, Parry R, Gillatt D, Peat D, and Mathieson PW. Expression of neuropilin-1 by human glomerular epithelial cells in vitro and in vivo. Clin Sci (Lond) 101: 439–446, 2001. Nicol D, Hii SI, Walsh M, Teh B, Thompson L, Kennett C, and Gotley D. Vascular endothelial growth factor expression is increased in renal cell carcinoma. J Urol 157: 1482–1486, 1997. Ostendorf T, Kunter U, Eitner F, Loos A, Regele H, Kerjaschki D, Henninger DD, Janjic N, and Floege J. VEGF165 mediates glomerular endothelial repair. J Clin Invest 104: 913–923, 1999. Paddison PJ and Hannon GJ. RNA interference: the new somatic cell genetics? Cancer Cell 2: 17–23, 2002. Parry RG, Gillespie KM, and Mathieson PW. Effects of type 2 cytokines on glomerular epithelial cells. Exp Nephrol 9: 275–283, 2001. Saleem MA, O’Hare MJ, Reiser J, Coward RJ, Inward CD, Farren T, Xing CY, Ni L, Mathieson PW, and Mundel P. A conditionally immortalized human podocyte cell line demonstrating nephrin and podocin expression. J Am Soc Nephrol 13: 630–638, 2002. Satchell SC, Harper SJ, Tooke JE, Kerjaschki D, Saleem MA, and Mathieson PW. Human podocytes express angiopoietin 1, a potential regulator of glomerular vascular endothelial growth factor. J Am Soc Nephrol 13: 544–550, 2002. Siddiqui FA, Desai H, Siddiqui TF, and Francis JL. Hemoglobin induces the expression and secretion of vascular endothelial growth factor from human malignant cells. Hematol J 3: 264–270, 2002. Sugimoto H, Hamano Y, Charytan D, Cosgrove D, Kieran M, Sudhakar A, and Kalluri R. Neutralization of circulating vascular endothelial growth factor (VEGF) by anti-VEGF antibodies and soluble VEGF receptor 1 (sFlt-1) induces proteinuria. J Biol Chem 278: 12605– 12608, 2003. Sugiono M, Perrin RM, Harper SJ, Oxley JD, Gillatt DA, and Bates DO. Vascular endothelial growth factor (VEGF) 165b, a novel and inhibitory VEGF isoform is down-regulated in prostate carcinoma (Abstract). Kuala Lumpur, Malaysia: Asian Congress of Urological Cancer, Kuala Lumpur, 2002. Tricarico C, Salvadori B, Villari D, Nicita G, Della Melina A, Pinzani P, Ziche M, and Pazzagli M. Quantitative RT-PCR assay for VEGF mRNA in human tumors of the kidney. Int J Biol Markers 14: 247–250, 1999. Whittle C, Gillespie K, Harrison R, Mathieson PW, and Harper SJ. Heterogeneous vascular endothelial growth factor (VEGF) isoform mRNA and receptor mRNA expression in human glomeruli, and the identification of VEGF148 mRNA, a novel truncated splice variant. Clin Sci (Lond) 97: 303–312, 1999.

286 • APRIL 2004 •

www.ajprenal.org