JOURNAL OF PATHOLOGY, VOL.

161: 93-97 (1 990)

RAPID COMMUNICATION

RENIN GENE EXPRESSION IN NEPHROBLASTOMA G. B. M. LINDOP*, K. DUNCAN*$, D. W. M. MILLAN*, A. A. M. GIBSON?, W. J. A. PATRICKT, B. J. LECKIES AND G. D. B R N I E ~

*University of Glasgow, Department of Pathology, Western InFrmary, Glasgow; ?Department of Pathology, Royal Hospitalfor Sick Children, Glasgow: SMRC Blood Pressure Unit, Western Infirmary, Glasgow: $Cancer Research Campaign Beatson Laboratories, Beatson Institute for Cancer Research, Glasgow Received 26 March 1990 Accepted 26 March 1990

SUMMARY Most cases of nephroblastoma have high plasma levels of prorenin which is biologically inactive. Plasma prorenin levels fall to normal following nephrectomy. In order to ascertain whether renin synthesis occurs in nephroblastomas we decided to search for renin-specificmRNA using a cDNA probe and Northern blot analyses on total RNA purified from snap-frozen human tumour tissue obtained at nephrectomy. We demonstrated renin-specific mRNA in 5/11 (45 per cent) nephroblastomas. It was 1.6 Kb in length, similar to the mRNA detected in normal kidney tissue and in kidneys with renal artery stenosis. In one of the cases of nephroblastoma, in which we could detect no normal renin mRNA at 1.6 Kb, the cDNA probe hybridized with a higher molecular weight mRNA 3 Kb in length. We conclude that some nephroblastomas synthesize renin. KEY WORDS-Renin,

nephroblastoma

INTRODUCTION We have previously shown that both renal cell carcinoma’**and nephrobla~toma~.~ can be associated with high levels of plasma renin. The renin is usually biologically inactive and may be a tumour marker.4In addition, most nephroblastomas possess a small population of cells which contain immunoreactive renin.5 However, renal tubular cells can reabsorb renin from the glomerular filtrate,6 and renal tumours possess some of the membrane characteristics of tubular cells.’ Therefore, in tumours the cells which contain immunoreactive renin could have taken it up from body fluids. Also, tumours growing in the kidney could increase intrarenal pressure, press on the renal vessels or even secrete a factor which stimulates renin release from the kidney. It is therefore uncertain whether the high plasma levels of inactive renin are secreted by the Addressee for correspondence: Dr. G. B M. Lindop, Department of Pathology, Western Infirmary, Glasgow GI 1 6NT.

0022-341 7/90/060093-05 $05.00 0 1990 by John Wiley & Sons, Ltd.

tumour or by the adjacent kidney. We decided to look for the presence of renin-specific mRNA in nephroblastoma tissue as direct evidence that renal tumours synthesize renin.

MATERIALS AND METHODS Tissues

Fourteen kidneys containing tumours were obtained fresh from the operating theatre. Pieces of fresh tumour tissue were snap-frozen in liquid nitrogen, and, in some cases, unaffected renal cortex from the same kidney was also collected. Tissue was obtained from other pathological kidneys including two cases of renal artery stenosis. All tissue was stored at - 70°C.Other pieces of the same tumours were fixed in neutral buffered formalin, embedded in paraffin wax and used for conventional histological stains and for immunocytochemistry.

94

G . B. M. LINDOP E T A L .

Fig. I-(a) A glomerulus in compressed renal cortex adjacent to a tumour. There are several cells which are staining positively at its vascular pole. Renin PAP counterstained with PAS: Nomarski illumination. (b) A glomerulus in a kidney with renal artery stenosrs. The vascular pole contains a large number ofcells which are staining positively. The individual cells are also much larger. Renin PAP counterstained with PAS: Nomarski illumination

Itnmunostaining

and the characteristics of this probe have been

Two blocks of parafin-embedded tumour were selected to include, where possible, some adjacent renal cortex: 5,um sections were cut and used for immunocytochemical demonstration of renin. As before’.’ we used an antiserum to human renin purified from cadaver kidneys and a peroxidaseantiperoxidase technique. The renin antiserum was incubated on the sections for 14 h at 4°C at a dilution of 1’1000 along with appropriate negative controls.

published.’

Probes

We obtained from Professor W Brammar, School of Biological Sciences. University of Leicester, a cloned cDNA of 640 base pairs representing the 3’ end of human renin mRNA.8 This was cloned into plasmid pGEM- I from which we prepared labelled cDNA by random priming using a commercial kit (Amersham International, U.K.). To check the integrity of the RNA purified from the tissues we used a cDNA probe for j3,-microglobulin mRNA. /l,-microglobulin is a ubiquitous cellular protein

Purification of R N A

RNA was purified using a commercial kit (RNAzolTMfrom Biogenesis Ltd, Bournemouth). This method is based o n a phenol. guanidium thiocyanate and chloroform extraction procedure followed by an isopropanol precipitation. Briefly. the small blocks of frozen tissue were ground under liquid nitrogen then dispersed in the RNAzolTM solution in a ratio of about 1 cmj of tissue per 10 ml. A tenth volume of chloroform was then added and the solution stored on ice for 15 min. After centrifugation at IOOOOg at 4°Cfor 15 min the aqueous phase which contains the RNA was removed to a fresh tube. The RNA was precipitated by adding an equal volume of cold isopropanol and incubating at - 20°C for 45 min. The precipitated RNA was then pelleted by centrifugation as above. washed in 95 per cent ethanol and vacuum dried before resuspension in 1 ml of sterile water. The concentration of

RENIN GENE EXPRESSION IN NEPHROBLASTOMA

A

B

C

Fig. 2-An autoradiogram of a Northern blot hybridized with the renin cDNA probe. Lanes A, B and C each contain 25 pg of RNA; from a normal kidney (A), from a kidney with renal artery stenosis (B) and from a nephroblastoma (C). The blot shows the presence of renin mRNA (1.6Kb) in normal renal cortex, in kidney with renal artery stenosis and in nephroblastoma tissue

RNA was then determined by measurement of absorbance at 260 nm. Northern blotting and hybridization

Total RNA (20-25 pg) from each specimen was size fractionated by electrophoresis through formaldehyde-1 per cent agarose gels; molecular weight markers were included in each analysis. The RNA was then transferred to nitrocellulose membrane filters by a standard method." The filters were hybridized overnight in 50 per cent formamide, 5 x SSC (1 x SSC is 0.15 M NaCl, 0.015 M sodium citrate, pH 7), and 10 per cent (w/v) dextran sulphate at 42°C with the cDNA probes ( lo7 counts/ min/,ug) labelled with 32P by random priming (Amersham International Multiprime Kit). The filters were then washed to a stringency of 0.1 x SSC, 0.1 per cent SDS at 60°C for 1 h. Autoradiograms were prepared by exposing Kodak XRP film to the membranes with intensifying screens at - 70" for 4-7 days. The membranes were stripped for reprobing with the µglobulin probe by heating at 100°C for 30 min in 0.1 per cent SDS.

95

RESULTS Immunostainable renin was always present in the juxtaglomerular apparatus (JGA) and in the arterial tree of the renal cortical tissue examined for comparison (Fig. la). Where the renal cortex was compressed because it was adjacent to a tumour, the JGAs were often hyperplastic. The cases of renal artery stenosis contained even more renin-positive cells both in the JGAs and in the walls of the arteries. Individual renin-containing cells in the JGA were also larger than those in the normal kidneys (Fig. 1b). Northern blots of RNA extracted from the normal renal cortex of the kidneys consistently showed a specific band of renin mRNA with a molecular size of 1.6 Kb. This corresponds to the accepted size of renin mRNA." More intensely hybridizing bands of the same molecular weight were found in Northern blots of RNA purified from two kidneys with renal artery stenosis suggesting relatively more abundant renin mRNA in this tissue (Fig. 2). The nephroblastomas we examined were unselected. There was therefore a full range of differentiation in our series of tumours; some had well differentiated epithelial structure with welldifferentiated tubules and glomeruloid bodies, others had well differentiated mesenchyme and some consisted mainly of undifferentiated mesonephric blastema. We found immunoreactive renin in 2/11 nephroblastomas. The tumours which contained the renin-positive cells showed no other evidence of specialized differentiation. The cells containing immunoreactive renin were predominantly perivascular in distribution. They were mainly found in the differentiated mesenchyme, often related to the capillary vessels at the junction with the primitive blastema (Fig. 3a). Some of the cells were spindle-shaped but many were irregular in shape and some had long cell processes (Fig. 3b). Many cells had a high nucleus to cytoplasm ratio but there was no definite histological evidence that these cells were malignant. Special stains were unhelpful. Positive staining for renin was never seen in the cells of the primitive metanephric blastema or in the tubular elements in the nephroblastomas. We obtained intact mRNA from 11 of 14 nephroblastomas; in the other specimens smears on Northern blots hybridized with theµglobulin cDNA probe indicated that the RNA was degraded. We demonstrated 1-6Kb renin mRNA in 4 of 1 1 (36 per cent) of the nephroblastomas with intact RNA (Fig. 2). In one other case, the normal renin mRNA was undetectable; instead, the probe detected an

96

G. B. M. LINDOP E r AL.

Fig. 3-(a) and (b) Perivascular cells which contain imrnunoreactive renin in a case of nephroblastoma. The cells are around capillary blood vessels and are situated at the junction between the well differentiated mesenchyme and the primitive blastema

A

B C

mRNA with a molecular weight of about 3 Kb, approximately twice the size of normal renin mRNA (Fig. 4). Intact mRNA was extracted from the two tumours in which we found cells containing immunoreactive renin, but Northern blot analyses failed to demonstrate renin mRNA in either of them. DISCUSSION

Fig. 4-An autoradiogram of a Northern blot preparation hybridized with the renin cDNA probe. Lanes A and B contain samples of RNA from a kidney with renal artery stenosis. These show a band corresponding to renin mRNA ( b1.6 Kb). The RNA from the case of nephroblastoma in lane C has no detectable normal renin mRNA: instead. there is a higher molecular weight mRNA ( .33 Kb) which hybridizes with the renin probe

Most of the renin in the blood is synthesized and secreted by the myoepithelioid cells in the juxtaglomerular apparatus and in the renal arterial tree. A stimulus such as narrowing of the renal arteries causes increased levels of renin in the blood and an increase in the number of renin-secreting cells in the kidney. In tissue sections, this is reflected in the increased numbers of cells containing immunoreactive renin.”.13 Our immunostaining results confirm this, and the increased abundance of renin mRNA revealed by Northern blot analysis is also in accord with this idea. Most patients with nephr~blastoma~ and a smaller number of patients with other renal

RENIN GENE EXPRESSION IN NEPHROBLASTOMA

tumours3 have high (up to 20 times normal) levels of inactive (pro)renin in their blood. Plasma prorenin levels fall to normal after removal of the kidney containing the t ~ m o u r . Since ~ ‘ ~ the renin in the blood could have been secreted either from the tumour or from the adjacent kidney, the primary aim of this work was to seek evidence of renin gene expression in tumour tissue. We suggest that our finding of renin mRNA in RNA isolated from nephroblastomas constitutes the most direct evidence provided hitherto that these tumours can synthesize renin. The cases in which we found renin mRNA by Northern blot analysis were not those in which we demonstrated cells containing immunoreactive renin. However, nephroblastomas are large and heterogeneous tumours, and they vary in differentiation from area to area; therefore, selection of small pieces of tissue for analysis makes sampling error the likeliest explanation for this apparent discrepancy. In nephroblastomas, the cells which contain immunoreactive renin are typically perivascular in distribution.’ If they are renin-secreting cells, they could have grown into the tumours along with the blood vessels from the neighbouring kidney. Similar perivascular renin-containing cells have been found around the blood vessels in congenital mesoblastic nephroma, and in situ hybridization studies with a probe for renin mRNA suggested that these cells synthesize renin.14 Small amounts of renin have been found in some lung carcinomas and immunostaining using a renin antiserum showed positive cells related to the blood vessels in and around the t u m o ~ r s ;these ’ ~ workers regarded these renin-containing cells as stromal

cells.15 In a previous study we have suggested on histological grounds that some of the cells which contain immunoreactive renin in nephroblastomas are indistinguishable from tumour cells.5 Also, other renal tumours and tumours from a variety of organs have been associated with renin secretion, and in some cases there is good evidence that the renin was secreted by tumour cells; for review see ref. 16. It is therefore possible that renin can be synthesized both by stromal elements within and around the tumour and by tumour cells themselves. The RNA from one of our cases of nephroblastoma contained no detectable normal sized renin mRNA; instead, there was an RNA of about 3 Kb in length which hybridized with our cDNA probe. We have found no other similar RNA species which hybridizes with our renin cDNA probe in Northern blot analysis of other specimens of normal and pathological renal tissue. The interpretation of

97

this finding is uncertain. The size of this transcript excludes it being an unprocessed primary renin gene transcript. However, it could represent a partly processed gene transcript or an aberrant transcript containing part or all of the normal renin mRNA sequence. Such an abnormality of renin gene expression may constitute evidence that the abnormal mRNA is derived from tumour cells rather than from vascular or stromal cells which have grown into the tumour from the adjacent kidney. ACKNOWLEDGEMENT

This work was supported by the Scottish Home and Health Department Grant K/MRS/SO/C 801. REFERENCES I. Lindop GBM, Stewart J, Downie TT.An immunocytochemical and ultrastructural study of a renal carcinoma secreting inactive renin. J Clin Parhol1983; 36:639-645. 2. Lindop GBM, Leckie B, Winearls CG. Malignant hypertension due to a rcnin-secreting renal cell carcinoma-an ultrastructural and immunocytochemical study. Hisroparhology 1986; 1 0 1325-1332. Renin 3. LeckieB,McIntyrcGD,MillanDW,LindopGBM,CarachiR. and inactive renin (prorenin) in the plasma of patients with malignant renal tumours. Clin €xper Hyperfens 1987; A9(8&9): 1325-1332. 4. Carachi R, Lindop GBM, k k i e B. Inactive renin-a tumor marker in nephroblastoma. JPedriar Surg 1987; 2 2 278-280. 5. Lindop GBM, Fleming S,Gibson AAM. Renin in nephroblastomaan immunocytochemical study. J Clin Parhof I984 Ji: 738-742. 6. Taugner R, Hackenthal E,Inagami T,Nobiling R, Poulsen K. Glomerular and tubular r a i n in the kidneys of mice. Hisrochemisfry 1982; 75: 473484. 7. Fleming S, LindopGBM, Gibson AAM. Thedistribution ofepithelial membrane antigen in the normal and developing human kidney and its tumours. Hisroparhology 1985; PI 729-739. 8. Samani NJ, Ruprai AK,Brarnrnar WJ, Swalcs JD. The renin gene in patients with malignant hypertension and raised plasma renin activity. CIinSci16 151-155. 9. Suggs SV, Wallace RB, Hirose T,Kawashima EH, Itakura K. Use of synthetic oligonucleotides as hybridisation probes: isolation of cloned cDNA sequences for human &-microglobulin. Proc Nor Acad Sci 1981;18 181O-l819. 10. Maniatis T, Fritxh EF, Sambrook 1. Molecular Cloning-a Loboralory Manual, Cold Spring Harbour Laboratory, USA; 199-206. 1. Soubrier F. Panthier J-J, Corvol P, Rougeon F. Molecular cloning and nucleotide sequence of a human renin cDNA fragment. Nucleic Acids Res 1983; 11: 7181-7190. 2. Graham PC, Stewart HVA, Downie I, Gardiner DS, More IAR, Lindop GBM. The distribution of rcnin-containing cells in kidneys with renal artery stenosis-an immunocytochemical study. Hisropathology 1990; 1 6 347-355. 3. Graham PC, Lindop GBM. The renin-secreting cell in polyarteritisan immunocytochemical study. Hisroporhology 1990; 1 6 339-345. 14. Taylor M, Cook T,Parson C, Risdon RA, Peart S.Renin messenger RNA localization in congenital mesoblastic nephroma using in situ hybridisation. J Hyperrension 1989; 7 733740. 15. Taylor GM, Cook HT, Sheffield EA. Hanson C, Peart WS. Renin in blood vessels in human pulmonary tumors-an immunohistochemical and biochemical study. Amer J Parhol1988; 130.534-551. 16. Lindop GBM, Lever AF. The anatomy of the rcnin-angiotensin system in the normal and pathological human kidney. Histopurhofogy 1986; 10 335-362.

Renin gene expression in nephroblastoma.

Most cases of nephroblastoma have high plasma levels of prorenin which is biologically inactive. Plasma prorenin levels fall to normal following nephr...
681KB Sizes 0 Downloads 0 Views