Expression of the von Hippel-Lindau disease tumour suppressor gene during human embryogenesis
Expression of the von Hippel-Lindau disease tumour suppressor gene during human embryogenesisFrances M. Richards, Paul N. Schofield1, Stewart Fleming2 and Eamonn R. Maher*
Cambridge University Department of Pathology, Tennis Court Road, Cambridge CB2 1QP, UK, 1Cambridge University Department of Anatomy, Downing Street, Cambridge CB2 3DY, UK and 2Department of Pathology, University of Edinburgh, Edinburgh, UK
Received December 11, 1995;Revised and Accepted February 12, 1996
The von Hippel-Lindau (VHL) disease product is thought to down-regulate transcription by antagonizing elongin-enhanced transcriptional elongation. Germline VHL gene mutations predispose to the development of retinal, cerebellar and spinal haemangioblastomas, renal cell carcinoma and phaeochromocytoma. In addition, somatic inactivation of the VHL gene is frequent in sporadic renal cell carcinoma and haemangioblastoma. Regulation of transcript elongation is an important control mechanism for gene expression and the VHL gene might modify the expression of proto-oncogenes and growth suppressor genes during embryogenesis. We therefore investigated the expression of VHL mRNA during human embryogenesis by in situ hybridization studies at 4, 6 and 10 weeks post conception. Although VHL mRNA was expressed in all three germ layers, strong expression was noted in the central nervous system, kidneys, testis and lung. Within the kidney, VHL mRNA was differentially expressed within renal tubules suggesting that the VHL gene product may have a specific role in kidney development. Two alternatively spliced VHL mRNAs characterized by inclusion (isoform I) or exclusion (isoform II) of exon 2 are transcribed in adult tissues. To investigate if the two isoforms are differentially expressed during embryogenesis, VHL mRNA was reverse transcribed from 13 fetal tissues (8-10 weeks gestation). The quantitative distribution of VHL mRNA within fetal tissues reflected that seen by in situ hybridization and the ratio of the two VHL isoforms was similar between tissues. Although the genes regulated by the VHL gene product have not yet been identified, our findings are compatible with the hypothesis that VHL-mediated control of transcriptional elongation may have a role in normal human development.
Precise control of transcription is fundamental to normal development and abnormalities of transcription regulation are frequent in human cancers. The von Hippel-Lindau disease tumour suppressor (VHL) gene maps to chromosome 3p25 and was isolated by a positional cloning approach in 1993 (1 ). Germline mutations in the VHL gene cause: (i) von Hippel-Lindau (VHL) disease, a dominantly inherited familial cancer syndrome with an incidence of 1 in 36 000 characterized by the development of retinal, cerebellar and spinal haemangioblastomas, renal cell carcinoma and phaeochromocytoma (1 -6 ), and (ii) familial phaeochromocytoma without other features of VHL disease (7 ,8 ). Loss of the wild-type VHL allele is required to initiate tumours in VHL patients (9 ), and somatic inactivation of both VHL alleles is a critical step in the pathogenesis of sporadic clear cell RCC and central nervous system haemangioblastomas (10 -14 ). The VHL coding sequence is represented in three exons and two alternatively spliced VHL mRNAs have been detected reflecting the presence (isoform I) or absence (isoform II) of exon 2. The predicted VHL protein product comprises 213 amino acids (15 ). Although the VHL protein is novel and significant homology to other human genes was not detected, the VHL protein has been shown to associate with elongins B and C (16 ,17 ). These two proteins bind to elongin A to form the heterotrimeric transcription factor elongin which increases RNA transcription by suppressing polymerase II pausing (18 ). In the presence of the VHL protein, elongin's transcriptional elongation activity is suppressed and the transcription of target genes (so far unidentified) will be reduced. Recently, Kessler et al. (19 ), have reported ubiquitous VHL expression during mouse embryogenesis and VHL mediated regulation of transcription could have a role in normal human development. To provide further insight into the role of VHL in embryogenesis, we have: (i) performed in situ mRNA hybridization studies to investigate the principal sites of VHL expression during human embryogenesis; and (ii) analysed the levels of both VHL isoforms in a variety of fetal tissues.
Expression of VHL mRNA was investigated in: (i) whole human embryo sections at 4, 6 and 10 weeks gestation, as assessed by crown/rump length and estimated menstrual age; and (ii) human fetal kidney sections at 23 weeks gestation. In situ hybridization was performed using a 630 bp VHL riboprobe (see Materials and Methods) with both antisense and competition controls. Extensive hybridization was seen over all of the embryos in most tissues. However, expression predominated in the specific locations described below. No marked differences were detected in the pattern of expression at different gestational ages.
At gastrulation the embryo becomes differentiated into three germ layers, the ectoderm, mesoderm and endoderm and it is convenient to use these lineage restricted progenitor tissues as a basis for describing the expression pattern of VHL. The central and peripheral nervous systems are derived from the neurectoderm, which becomes differentiated into the neural crest and the neural tube, giving rise to the brain, spinal cord, sensory, autonomic ganglia and a wide range of mesenchymal and neuroendocrine tissues. The endodermally derived tissues are mainly the derivatives of the primitive gut, which give rise to the gut lining, the bronchi of the lung, part of the lower urogenital tract and bladder, the pancreatic ducts and part of the liver. The mesoderm contributes to the bulk of the remaining organs, the musculature, kidneys, gonads, part of the adrenal glands, the haematopoietic system, the skeleton and cartilage elements not generated by the neural crest, and the vasculature. As specific tissues are discussed below their origins will be outlined.Mesodermal derivatives. VHL was expressed in all of the connective tissue and mesenchyme examined. Predominant areas were in the derivatives of the genital ridge and the nephrogenic cord. Skeletal muscle, the heart, liver and perichondral cartilage all expressed the gene above background levels.Genitourinary system. The genitourinary system develops from the intermediate mesoderm lying on either side of the axis of the embryo lateral to the somites. It is initially visible as a condensation of cells towards the cephalic end of the embryo which develops into the pronephros. In the human embryo this is a rudimentary structure and probably has no function. As the nephrogenic ridge extends caudally the mesonephros develops which is composed of a central tubule, the mesonephric duct, leading off from which are primitive but functional nephrons. The mesonephric duct extends caudally until it reaches and fuses with the ventral aspect of the cloaca. This structure persists until 10 weeks by which time it begins to regress. The definitive metanephric kidney begins to develop at day 28 when diverticulae (the ureteric buds) grow out of the extreme caudal end of the mesonephric ducts to penetrate the adjacent mesenchyme of the intermediate mesoderm (20 -23 ). This process is termed induction and results in the epithelialization of the mesenchyme to form nephrons, the branching of the ureteric bud and the proliferation of both tissues to form the definitive kidney. It is generally believed that the cells of the nephron are derived from the metanephric mesenchyme and the collecting ducts from the ureteric bud. The extending ureteric bud becomes the ureter, while the remaining mesonephric duct becomes the vas deferens in the male.
The gonads arise as lateral bulges on the anterior surface of the nephrogenic ridge during the fifth week and during the fifth and sixth weeks are colonized by primordial germ cells. They remain morphologically indifferent until week 7, when sexual differentiation begins. The germ cells remain a minor component of the gonad at the stages examined, with the supporting tissues dominating their appearance. The paramesonephric ducts which form during this process regress in the male and produce the oviducts and the uterus in the female.
The most abundant source of VHL messenger RNA was found in the indifferent gonad (Fig. 1 D). Both the cortical and medullary cells of the gonad expressed VHL at high levels. In the mesonephros low levels of expression were seen in the mesenchyme of the nephrogenic cord, but the predominant pattern was that of expression in the mesonephric duct epithelia and mesonephric tubules. Both the mesonephric and paramesonephric ducts expressed VHL at similar levels within the developmental window examined. No expression was noted in the early part of metanephric development, but by 23 weeks expression was clearly visible throughout the metanephric kidney in both the ureteric bud and metanephric mesenchyme derived components (Fig. 1 A-C). Expression was detected in the visceral and parietal layers of Bowman's capsule and proximal and distal convoluted tubule, but the most abundant source of RNA was in the ascending/descending limbs of the loops of Henle. The medulla of the kidney showed heterogeneous labelling over the thin tubular epithelium in the medulla and clear labelling over the collecting ducts. Labelling of parallel sections with an antibody to WT1 (data not shown and P. Ramani, personal communication) indicated a similar distribution of WT1 labelling with VHL expression in this region, which together with the morphology suggests that the tubules expressing the VHL gene are the long limbs of the loops of Henle from maturing nephrons developing within the medulla of the metanephros.
To investigate the representation of VHL isoforms I (containing all three exons) and isoform II (exons 1 and 3) in fetal tissues, RT-PCR analysis was performed to amplify both transcripts. In addition, RT-PCR for glyceraldehyde-3 phosphate dehydrogenase (GAPDH), WT1 and c-fos mRNA was performed as control experiments as the patterns of expression of these genes are known. Thirteen tissues from embryos of approximately 9 weeks gestation (range 8.0-10.0 weeks) were analysed (see Materials and Methods for details). For comparison, cultured primary adult renal tubular cells were also analysed. The levels of VHL, WT1 and c-fos mRNAs were compared with that of GAPDH (Fig. 2 ). Both isoforms of VHL mRNA were detected in all tissues and the ratio of the two isoforms was similar between tissues. In general the level of expression of VHL mRNA in the different tissues reflected that seen by in situ hybridization. In contrast, WT1 showed a more tissue-specific pattern of expression, consistent with previous studies (24 ); high levels in the kidney and testis; low levels in the spinal cord, intestine, pancreas, heart and lung; and not detectable in the liver, brain, eye, adrenal, limbs or placenta. c-fos also showed a tissue-specific pattern of expression, with very high levels of mRNA detected in the eye, spinal cord, intestine, testis and lung.
The pattern of expression of VHL mRNA in human embryogenesis appears to be similar to that described in the mouse (19 ). Thus, although VHL expression is ubiquitous, we detected particularly high levels of expression in the urogenital system, brain, spinal cord, sensory ganglia, eyes and bronchial epithelium (see Table 1 ). This pattern correlates to some extent with the pattern of organ involvement in VHL associated disease, but there are significant differences. The most frequent complications of VHL are retinal and CNS haemangioblastomas, renal carcinoma and cysts (2 ) and somatic VHL mutations are found in most sporadic clear cell renal carcinomas (10 -13 ). However, germ cell tumours are rare in VHL disease and somatic VHL mutations are not a feature of gonadal tumours (25 ). Although chromosome 3p allele loss is found in the majority of lung tumours, somatic VHL mutations are rare (26 ). Despite phaeochromocytoma and pancreatic apudomas being recognized complications of VHL disease, we did not find VHL expression in the fetal adrenals or pancreas at 4 and 6 weeks gestation. However, it is clear that other neural crest derivatives are a major source of VHL expression. Much of the cranial mesenchyme giving rise to the viscerocranium and that contributing to the endochondral bone of the neurocranium, together with the facial structures derived from the first two branchial arches, are derived from the neural crest. The three cranial ganglia, V, VII and VIII all have major contributions from the neural crest, and the dorsal root ganglia are similarly derived. The marked expression of VHL mRNA in the developing vestibular apparatus was of interest as endolymphatic sac tumours are a recently described complication of VHL disease (27 ).
. Summary of VHL mRNA expression during human embryogenesis
Germ layer
Tissue
VHL expression
Endoderm
Gut
-
Pancreas
-
Lung
++
Pharyngeal pouches
-
Ectoderm
Dorsal root ganglia
+
Periderm
(+)
Brain
++
Cranial ganglia (V, VII, VIII)
+++
Vestibular apparatus
+++
Eye
+
Cephalic mesenchyme
+
Mesoderm
Heart
(+)
Indifferent gonad
++++
Mesonephric duct
++
Paramesonephric duct
++
Bowman's capsule
+
Loop of Henle
++++
Nephrogenic cord
+
Metanephric collecting ducts
++
Proximal convoluted tubule
++
Perichondrium
(+)
Liver (haematopoietic tissue)
+
Adrenal
(+)
Striated muscle
+
At the stages of development examined, the adrenal gland has few neural crest derived medullary cells. At week 6 neural crest cells have not entered the gland and at weeks 7-8 they are present but not clustered into medullary structures; therefore, lack of signal from the adrenal may indicate that we have insufficient cells or that during their immediate migratory phase they do not express the gene. Kessler et al. (19 ), however, failed to find significant expression in the mouse at much later equivalent gestational ages, but we did demonstrate VHL mRNA by RT-PCR at 8-10 weeks gestation. We did not detect expression of VHL in vascular endothelial cells, and this would be compatible with the hypothesis that the VHL gene product has a role in angiogenic regulatory process normally exercised by other cell types. We suggest that VHL is necessary for the control of genes in neural crest derivatives, possibly cartilage or glia, whose normal function is to control endothelial cell growth. In the light of this, it is interesting to note that Wizigmann-Voos et al. (28 ) have reported upregulation of vascular endothelial growth factor (VEGF) gene expression in VHL and sporadic central nervous system haemangioblastomas, the source of the VEGF being the tumour stromal cells. Also, Brown et al. (29 ) demonstrated increased VEGF expression in renal cell carcinoma but not in endothelial cells. As the VHL gene product might be involved in controlling transcriptional elongation, and so expression, of several genes, the consequences of VHL gene inactivation may differ according to the spectrum of gene expression in different tissues. In addition, loss of control of the expression of one gene product may not be rate limiting for tumorigenesis in all tissues and under all circumstances. It is clear, however, that the pattern of normal expression does correlate with tumorigenic potential in the kidney.
We also considered that a possible explanation for the apparent contradiction of ubiquitous VHL mRNA expression and the specificity of tissue involvement in VHL disease, might be that the ratio of the two VHL mRNA isoforms varied between tissues.VHL patients with germline deletions of exon 2 have a reversal of the normal ratio of isoform I to II and have classical VHL disease, so that it seems that the full length transcript is required for full VHL function. However, the ratio of the two VHL mRNA transcripts was broadly similar between tissues. We cannot exclude the possibility that intratissue differences in the expression of VHL mRNA isoforms by different cell types might exist or that the pattern of expression of alternative VHL proteins might vary (30 ).
The distribution of VHL expression within the kidney was not uniform, but was concentrated within the renal tubules, particularly in the loops of Henle. Kessler et al. (19 ) have also analysed VHL mRNA expression in human fetal kidneys, but at a different stage (14 weeks). Although we both detected VHL mRNA in proximal tubular cells, we were able to detect a more specific pattern of differential expression within the renal tubules. These findings would be compatible with a role for VHLmediated control of transcription in normal renal development. Although it is not yet known which genes are dependent on VHL for transcriptional control, it is known that regulation of transcript elongation is important in determining the expression of many genes including c-myc, N-myc, L-myc and c-fos which are expressed in the developing kidney (31 ). Similarly, little is known about the regulation of VHL expression, although the VHL promoter region does contain putative binding sites for the Pax family of transcription factors, members of which are also expressed during nephrogenesis (30 ,31 ). Although further work is necessary to confirm a role for the VHL gene in renal development, our results are compatible with this hypothesis and the expectation is that the VHL gene product will downregulate renal tubular cell growth and division. Hyperplasia of renal tubular cells is thought to be important in the pathogenesis of renal cysts in addition to predisposing to renal cell carcinoma (32 ).
A NotI-XbaI fragment was excised from the VHL cDNA clone g7 (1 ). This fragment, encoding nucleotides 386-917 (GenBank accession no. L15409), was blunt cloned into the EcoRV site of pBluescript II SK vector. Plasmid DNA was then linearized with HindIII and purified by treatment with pronase (0.1 mg/ml) for 3 h at 37oC followed by two phenol/chloroform extractions and ethanol precipitation. 35S-labelled antisense riboprobe (630 bp) was synthesized from the T3 promoter of this template using an RNA transcription kit (Stratagene), according to the manufacturer's instructions. One [mu]g of plasmid DNA was used in each 25 [mu]l reaction, which contained 125 [mu]Ci of uridine 5'-(alpha-thio 35S)-triphosphate (NEN Dupont; 1000-1500 Ci/mmol). A parallel reaction was included without isotope, for preparation of the control probe. After transcription, the probes were treated with 20 U DNaseI (RNase-free, Pharmacia), then purified by phenol/chloroform extraction and ethanol precipitation (twice) in the presence of >5 mM dithiothreitol (DTT). The 35S antisense probe was then diluted to 1*105 c.p.m./[mu]l in hybridization buffer (50% formamide, 10% dextran sulphate, 1*Denhardt's, 20 mM Tris-HCl pH 8.0, 0.3 M NaCl, 10 mM sodium pyrophosphate, 0.5 mg/ml yeast tRNA and 50 mM DTT). Part of this probe was prepared as a control at 1*105 c.p.m./[mu]l, by the addition of 10-fold excess of unlabelled antisense transcript as competitor.
In situ hybridization studies were performed on three human fetuses of 4, 6 and 10 weeks gestation, and a fetal kidney of 23 weeks gestation. Tissues for RNA extraction were dissected from approximately 8 to 10 week old human fetuses obtained from the MRC Tissue Bank. Thirteen different tissues were available; liver (8.1 weeks), kidney (8.1 weeks), brain (8.0 weeks), eye (8.8 weeks), spinal cord (8.3 weeks), adrenal gland (8.9 weeks), intestine (9.3 weeks), pancreas (10.0 weeks), heart (8.0 weeks), testis (9.1 weeks), lung (8.3 weeks), limbs (8.8 weeks) and placenta (8.5 weeks). A culture of primary human adult renal tubule cells was provided by Krystyna Czapla (Department of Nephrology, Hope Hospital, Salford) as a control. Ethical approval was given by the Cambridge Health Authority Ethics Committee.
Tissue was fixed in 4% paraformaldehyde and embedded in paraffin wax. Sections were cut, mounted on glass slides and dewaxed prior to hybridization essentially according to the method of Brice et al. (32 ). The slides were hybridized at 50oC for 19 h then washed to a stringency of 2* SSC, 0.1 M DTT, 50% formamide at 50oC, followed by RNase treatment and further washing at 50oC with 2* SSC, 0.1 M DTT, 50% formamide, then at 42oC in 2* SSC and at room temperature in 0.1* SSC. After autoradiography with K5 emulsion (Ilford) for 11 days, the slides were developed and counterstained with toluidine blue.
Total cellular RNA was extracted from each sample using TRI REAGENT (Molecular Research Center). The concentration of each RNA sample was measured, each was diluted to 1 [mu]g/[mu]l, and the concentrations were again measured.
One [mu]g of each RNA (accurately measured) was reverse transcribed with an oligo(dT) primer in a 20 [mu]l reaction volume, using a Reverse Transcription System (Promega) according to the manufacturer's instructions. Two [mu]l of undiluted reverse transcription (RT) product was used for quantitative polymerase chain reaction (PCR) amplification of each VHL, c-fos and WT-1 mRNAs, and 2 [mu]l of a 1/10 dilution of each RT product was used for quantitative PCR of GAPDH. The `forward' primer for each PCR was labelled with ([gamma]-32P)-adenosine triphosphate (Amersham) using T4 polynucleotide kinase (BRL). The PCR reactions were all designed to work under the same conditions; a 15 [mu]l reaction containing 1.5 mM MgCl2, 10 mM Tris-Cl pH 8.8, 50 mM KCl, 0.01% gelatin, 0.2 mM each dNTP, 0.53 [mu]M each primer and 0.2 U Taq polymerase (NBS). PCR amplification was performed (94oC for 1 min, 62oC for 1 min, 72oC for 1 min) for 28 cycles. Preliminary experiments had established that amplification was exponential up to at least 30 cycles for all four genes amplified under these conditions. Equal volumes of each PCR reaction were then separated on 6% polyacrylamide gels (45 cm long * 1 mm thick) in 0.5* TBE (45 mM Tris-borate, 1 mM EDTA), at 3 W per gel for 16 h. The gels were dried and autoradiography was performed using Kodak X-Omat LS film. Band densities were quantitated using a Lynx Densitometer (Applied Imaging), and the level of expression of each gene was then standardized with respect to the level of GAPDH expression in each sample.
The primers used for PCR of VHL cDNA were VHLF1 (5'-GCGTCGTGC-TGCCCGTATG-3') and VHLR1 (5'-TTCTGCACATTTGGGTGGTCTTC-3') These amplified VHL nucleotides 458-800 (from GenBank L15409), giving a 343 bp fragment of VHL mRNA isoform 1 (exons 1+2+3) and a 219 bp fragment from VHL mRNA alternatively spliced isoform 2 (exons 1+3). Primers for amplification of a 180 bp fragment of WT1 cDNA were WT1F (5'-GATGCAYAGCMGGAAGCACACT-3') in exon 7 and WT1R (5'-TGGGTCTTCAGRTGGTCGGA-3') in exon 9; designed using the sequence from Haber et al. (34 ). Primers for amplification of a 452 bp fragment of c-fos mRNA were FOSF1 (5'-ATGATGTTCTCGGGYTTCAACG-3') at position 289 in exon 1 and FOSR1 (5'-GCAGCCATCTTATTCCKTTCCCT-3') at position 1924 in exon 3 using sequence from F. Van Straaten et al. (35 ). Primers for amplification of a 369 bp cDNA fragment of the control gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were GAPDHF (5'-GACCCCTTCATTGACCTCAACTACA-3') at position 3189 and GAPDHR (5'-CTAAGCAGTTGGTGGTGCAGGA-3') at position 3959 (sequence from L. Ercolani et al. GenBank J04038).
We thank Dr John Kingdom and Dr L. Wong (MRC Tissue Bank) for their assistance in providing embryo material for this study and Dr Krystina Czapla for providing human renal tubular cell cultures. This work was supported by Action Research.
1 Latif, F., Tory, K., Gnarra, J., Yao, M., Duh, F-M., Orcutt, M.L., Stackhouse, T., Kuzmin, I., Modi, W., Geil, L., Schmidt, L., Zhou, F., Li, H., Wei, M.H., Chen, F., Glenn, G., Choyke, P., Walther, M.M., Weng, Y., Duan, D.R., Dean, M., Glavac, D., Richards, F.M., Crossey, P.A., Ferguson-Smith, M.A., Le Paslier, D., Chumakov, I., Cohen, D., Chinault, C.A., Maher, E.R., Linehan, W.M., Zbar, B. and Lerman, M.I. (1993) Science,260, 1317-1320.MEDLINE Abstract
2 Maher, E.R., Yates, J.R.W., Harries, R., Benjamin, C., Harris, R. and Ferguson-Smith, M.A. (1990) Q. J. Med., 77, 1151-1163.MEDLINE Abstract
3 Maher, E.R., Iselius, L., Yates, J.R.W., Littler, M., Benjamin, C., Harris, R., Sampson , J., Williams, A., Ferguson-Smith, M.A. and Morton, N. (1991) J. Med. Genet., 28, 443-444.MEDLINE Abstract
12 Shuin, T., Kondo, K., Torigoe, S., Kishida, T., Kubota, Y., Hosaka, M., Nagashima, Y., Kitamura, H., Latif, F., Zbar, B., Lerman, M.I. and Yao, M. (1994) Cancer Res., 54, 2852-2855.MEDLINE Abstract
13 Whaley, J.M., Naglich, J., Gelbert, L., Hsia, Y.E., Lamiell, J.M., Green, J.S., Collins, D., Neumann, H.P.H., Laidlaw, J., Li, F.P., Klein-Szanto, J.P., Seizinger, B.R. and Kley, N. (1994) Am. J. Hum. Genet.,55, 1092-1102.MEDLINE Abstract
14 Kanno, H., Kondo, K., Ito, S., Yamamoto, I., Fujii, S., Torigoe, S., Sakai, N., Hosaka, M., Shuin, T. and Yao, M. (1994) Cancer Res., 54, 4845-4847.MEDLINE Abstract
15 Iliopoulos, O., Kibel, A., Gray, S. and Kaelin, W.G. (1995) Nature Med., 1, 822-826.
22 Schofield, P.N. and Boulter, C.A. (1996) Exp. Nephrol. (in press).
23 Larsen, W.J. (1993) Human Embryology. Churchill Livingstone.
24 Pritchard-Jones, K., Fleming, S., Davidson, D., Bickmore, W., Porteous, D., Gosden, C., Bard, J., Buckler, A., Pelletier, J., Housman, D., van Heyningen, V. and Hastie, N. (1990) Nature, 346, 194-197.MEDLINE Abstract
25 Foster, K., Osborne, R.J., Huddart, R.A., Affara, N.A., Ferguson-Smith, M.A., Maher, E.R. (1995) Eur. J. Cancer31A, 2392-2395.
26 Sekido, Y., Bader, S., Latif, F. et al. (1994) Oncogene,9, 1599.MEDLINE Abstract
27 Delisle, M.B., Uro, E., Rouquette, I., Yardeni, E. and Rumeau, J.L. (1994) J. Clin. Pathol.,47, 959-961.MEDLINE Abstract
28 Wizigmann-Voos, S., Brier, G., Risau, W. and Plate, K.H. (1995) Cancer Res., 55, 1358-1364.MEDLINE Abstract
29 Brown, L.F., Berse, B., Jackman, R.W., Tognazzi, K., Manseau, E.J., Dvorak, H.F. and Senger, D.R. (1993) Am. J. Pathol., 143, 1255-1262.MEDLINE Abstract
30 Kuzmin, I., Duh, F-M., Latif, F., Geil, L., Zbar, B. and Lerman, M.I. (1995) Oncogene, 10, 2185-2194.MEDLINE Abstract
31 Bard, J.B.L., McConnell, J.E. and Davies, J.A. (1994) Mechanisms Dev.,48, 3-11.MEDLINE Abstract
32 Boulter, C.A. and Sandford, R. (1996) Exp. Nephrol. (in press).
33 Brice, A.L., Cheetham, J.C., Botton, V.J., Hill, N. and Schofield, P.N. (1989) Development, 106, 543-547.MEDLINE Abstract
35 van Straaten, F., Muller, R., Curran, T., van Beveren, C. and Verma, I.M. (1983) Proc. Natl Acad. Sci. USA,80,3183-3187.MEDLINE Abstract
*To whom correspondence should be addressed at: University of Cambridge, Box 134, Addenbrooke's Hospital, Hills Road, Cambridge CB2 2QQ, UK
This page is maintained by OUP admin. Last updated Thu Oct 31 15:23:49 GMT 1996. Part of the OUP Journals World Wide Web service.Copyright Oxford University Press, 1996
