JOURNAL OF PATHOLOGY, VOL.

167: 273-277 (1992)

RAPID COMMUNICATION

MITOCHONDRIAL AND CHROMOSOMAL DNA ALTERATIONS IN HUMAN CHROMOPHOBE RENAL CELL CARCINOMAS ANGELA KOVACS, STEPHAN STORKEL, WOLFGANG THOENES AND GYULA KOVACS

National Cancer Center Research Institute, Genetics Division, Tokyo, Japan, and Institute of Pathology, Johannes Gutenberg University, Mainz, Germany Received 21 February 1992 Accepted2 April 1992

SUMMARY Renal cell tumours are characterized by the loss of chromosome 3p and trisomy of 5q segments (common, nonpapillary renal cell carcinoma), or by trisomy ofchromosomes 7 and 17 and loss of the Y chromosome (papillary renal cell carcinoma), or by random karyotype changes and mitochondrial DNA alterations (renal oncocytoma). We have studied by means of RFLP analysis the genomic and mitochondrial DNA in 11 chromophobe renal cell carcinomas, which have a unique morphology among kidney cancers. We found a loss of the constitutional heterozygosity at chromosomal regions 3p, 5q, 17p, and 17q, a combination of allelic losses that has not been found in other types of renal cell tumours. Three of the tumours showed a gross alteration in the restriction pattern of the mitochondrial DNA. A combined morphological and genetic analysis suggests that chromophobe renal cell carcinoma is a distinct entity. KEY

worm-Chromophobe

renal cell carcinoma, RFLP analysis, mitochondrial DNA alterations

INTRODUCTION Through the use of molecular cytogenetic techniques, it has become possible to divide renal cell tumours (RCTs) into three genetically well-defined groups:' ( I ) Non-papillary renal cell carcinomas (RCCs), which comprise about 80 per cent of kidney tumours in adults, show a specific loss of the 3p 13pter and trisomy of the 5q22-qter chromosomal regions as well as non-random loss of 6q, 8p, 9, and 14q segment^.^-^ (2) Papillary RCTs comprise about 10 per cent of kidney cancers. Papillary renal cell adenomas (RCAs) are characterized by a highly specific combination of trisomy of chromosomes 7 and 17 and loss of the Y chromosome. During progression, papillary RCCs, which probably evolve from papillary RCAs, acquire additional chromosomal alterations such as trisomy of chromosomes Addressee for correspondence: Dr G . Kovacs, National Cancer Center Research Institute, Genetics Division, 1-1 Tsukiji 5-chome, Chuo-ku. Tokyo 104, Japan.

0022-341 7/92/07027345 $07.50 0 1992 by John Wiley & Sons, Ltd.

16, 12, and 20."8 Renal oncocytomas (5 per cent of RCTs) are not affected by recurrent alteration of the chromosomal D N A , but they display a n alteration in the restriction pattern of mitochondrial Recently, a new type of renal cell carcinoma was described-chromophobe RCC-which comprises about 5 per cent of R C C S . " , ' ~Its unique histology was first recognized among kidney tumours of the rat.13 Chromophobe R C C is characterized by cells having a pale, fine reticular cytoplasm. The cells are positive with Hale's acid iron colloid stain. Ultrastructurally, chromophobe cells display pathognomonic cytoplasmic vesicles and a variable number of morphologically altered mitochondria. The present study was designed to gain a n insight into the molecular genetics of chromophobe RCC. O u r question was whether chromophobe RCC, with its unique phenotype, conforms t o one of the three known genotypic patterns of renal cell tumours o r whether it is marked by distinct genetic changes.

274

A. KOVACS E T A L .

METHODS Non-papillary RCCs exhibit a loss of chromosome 3p sequences in 96 per cent ofcases and trisomy of the chromosome 5q sequences in 50 per cent of cases. These changes never occur in papillary RCTs or in renal oncocytomas. Trisomy 17 has been found in 100 per cent of papillary RCAs and 85 per cent of papillary RCCs, but never in non-papillary RCCs, or in renal oncocytomas. Neither monosomy 5 nor monosomy 17 occurs in any of these tumours. Therefore, we examined the allelic status in 11 chromophobe RCCs by using polymorphic DNA markers assigned to chromosomes 3p, 5q, 17p, and 17q. High molecular weight DNA was isolated from normal and tumour tissues by proteinase K digestion with sodium dodecyl sulphate (1 per cent) for 5 h at 50°C. After phenol-chloroforrn extractions and ethanol precipitation, DNA was dissolved in TE buffer. Seven pg of DNA was digested to completion with restriction enzymes appropriate for the probes used. DNA fragments were separated on 1 per cent agarose gel and transferred to nylon membranes. Filters were hybridized to probes radiolabelled by the random-primer method. DNA probes used in this study were homologous to the loci on chromosomes 3,5, and 17: pBH302 (THRB) at 3p24.1-p22: pH3H2 (D3F15S2E) and 3p21; J0209E-B (D5S12) at 5~15.2-15.1; J0205H-C (D5S22) at 5q32-qter; pYNZ22.1 (D17S30) at 17~13.3;and pTHH59 (D17S4) at 17q23-25.2. For a description of the RFLPs see Human Gene Mapping Chromophobe RCCs contain only very few stromal cells; therefore, an admixture of normal and tumour DNA is not a technical problem for establishing the allelic status in tumour tissue. To confirm the results, we rehybridized the filters with probes retaining the heterozygosity, and compared the signal intensity of alleles in normal and tumour DNA as d e ~ c r i b e d . ' ~ To analyse mitochondrial DNA in tumours, we used normal mitochondrial genomic DNA (1 6.5 kb) as a probe. Mitochondria1 DNA was isoiated from normal kidney tissue as described." The mitochondrial DNA was linearized with PvuII digestion, electrophoresed in 1 per cent low melting agarose gel, excised, and labelled in agarose by the randomprimer method. The labelled DNA was then hybridized to total cellular DNA of normal kidney and chromphobe RCC digested with various restriction enzymes. The same experimental approach was used to analyse the mitochondrial DNA from ten

oncocytomas, ten papillary, and five non-papillary RCCs. RESULTS The results of Southern analysis of 11 chromophobe RCCs using DNA probes from chromosomes 3, 5, and 17 are shown in Table I. Loss of DNA sequences at chromosome 3p was observed in five of nine informative cases, while four retained constitutional heterozygosity. No duplication of alleles for loci on either the short or the long arms of chromosome 5 was detected. However, two of the five tumours informative for the chromosome 5q probe showed a loss of one allele in tumour tissue. RFLP analysis using chromosome 17-specificDNA probes failed to show allelic duplication, i.e., trisomy of chromosome 17. In contrast, six tumours showed a loss of constitutional heterozygosity at chromosome 17. Examples showing the loss of heterozygosity at chromosomes 5q and 17p are shown in Fig. 1. Southern analysis using various restriction enzymes revealed an alteration in the restriction pattern of mitochondrial DNA in three chromophobe RCCs (Table 1). Normally, the restriction enzyme PvuII has only one recognition site in the mitochondrial genome; i.e., after digestion, all normal tissue gives a hybridization signal at 16.5 kb, the size of mitochondrial DNA. However, in two chromophobe RCCs, two and three hybridization signals smaller than the 16.5 kb signal were seen. After digestion with endonuclease XbaI, normal mitochondrial DNA displays five fragments of 7.5, 4.5, 1.9, 1.7, and 0.8 kb when separated on agarose gel. In the two cases with altered PvuII bands, an altered XbaI restriction pattern with two or three additional fragments was seen. An example is shown in Fig. 2. In the third case, although the PvuII digestion does not indicate a rearrangement or deletion in the mitochondrial genome, the XbaI digestion shows an altered pattern of restriction similar to that of the other two tumours. Neither the ten non-papillary RCCs, the five papillary RCTs, nor the ten renal oncocytomas examined in this series experienced alterations after PvuII or XbaI digestion. DISCUSSION We have demonstrated in this study that chromophobe RCCs experience allelic loss at chromosomes 3p, 5q, and 17 at a variable frequency, but do not

275

DNA ALTERATIONS IN CHROMOPHOBE RCC

Table I-Allelic assignments on chromosomes 3, 5 , and 17 as determined by RFLP analysis, and alteration of mitochondrial D N A in chromophobe RCCs. D N A probes, their location, and the restriction enzymes used are indicated at the top. Alleles were designated 1 or 2 and refer to larger or smaller fragments, respectively. ‘-’ indicates that both normal and tumour tissue were homozygous. ‘ + ’ indicates an altered restriction pattern of the mitochondrial genome after PvuII o r XbaI digestion No.3p pBH302 HindIII

Tumour

No.5p No.5q J0209E-B J0205H-C MspI MspI

No.3p pH3H2 HindIII

No. 17p pYN22.1 MsPI

No. 17q pTHH59 TaqI

-

425 315 570 1529 349 693 819 289 796 229 1314

12

-

-

1 2 12

12 12 -

-

12 12 12

12 12

-

I

-

1

Mitochondria1 DNA PVUII XbaI

+

+

+ + +

-

display trisomy for chromosome 5q or 17. This allelotype has not been seen in other types of renal cell tumour. Although some of the papillary RCTs have two copies of apparently normal chromosome 17, such tumours never show an alteration of chromo1some 3p.8,’6Furthermore, no other types of kidney 2tumour showed a loss of heterozygosity for loci at chromosomes 5q and 17q analysed in this study. This combination of multiple allelic losses suggests J0209E-B J0205H-C that genetic alterations in chromophobe RCCs are distinct from those occurring in other types of kidney tumour. Until now, only three chromophobe N T N T RCCs had been analysed cytogenetically. ”,’* Each tumour showed karyotype instability, i.e., pulver1ization of chromosomes, telomeric association, and multiple chromosomal losses resulting in low chromosome numbers of 34-38. The data presented here, in concert with previous cytogenetic findings, 2suggest that extensive chromosomal loss may be a characteristic feature of chromophobe RCCs. J0205H-C pYNZ22.1 Ultrastructurally, chromophobe RCCs are charFig. I-Southern hybridization of chromosome 5p, 5q, and 17p acterized by a variable number of invaginated probes to Mspl-digested DNA from normal (N) and tumour (T) cytoplasmic vesicles and abnormal mitochontissue. For each RFLP, the largest allele is marked 1 and shortest dria.l’~’9~20 The present study shows that the mito2. (A) Tumour tissue retained the heterozygosity at chromosome chondrial DNA in chromophobe RCCs may be 5p (DNA probe J0209E-B). Rehybridization ofthe same filter to the DNA probe J0205H-C revealed the loss of allele 1 at chromo- extensively rearranged. Whether these changes some 5q in tumour tissue. (B) Tumour tissue retained the hetero- affect the function of mitochondrial genes, and if so, zygosity in the chromosome 5q region for the DNA probe what genes are involved, is not yet known. Recently, J0205H-C. Rehybridization of the same filter with the polymorphic DNA probe YNZ22.1, which is assigned to chromosome it has been shown that the replication of mito17p, shows the loss of heterozygosity. The presence of a faint chondrial DNA might be under the control of gene(s) encoded in the nucleus.21 Whether the band for allele I is due to the admixture of a few normal cells

A

N

T

N

T

276

A. K O V A C S E T A L .

B kb

N T

N T kb

7.5 16.5

4.5

1.9 1.7

0.8 Pvull

of collecting duct intercalated cells which are suggested to be progenitors for renal oncocytoma and chromopohobe RCC.23However, the relationship between renal oncocytomas and chromophobe RCCs, if any, and their cellular origin remains to be confirmed. In summary, we have identified a combination of allelic losses at chromosomes 3p, 5q, and 17 and an altered restriction pattern of the mitochondrial DNA which distinguishes chromophobe RCCs from other renal tumours. Our observation indicates that chromophobe RCC represents a distinct genetic entity among kidney cancers. Further molecular biological studies will be necessary in order to understand the complexity of genetic events affecting the chromosomal as well as the mitochondrial DNA in chromophobe RCCs. REFERENCES

Xbal

Fig. 2-Restriction p a t t e r n of m i t o c h o n d r i a l DNA from n o r m a l ( N ) and t u m o u r (T)tissue from case H A 3 15. T o t a l cellular D N A w a s digested w i t h P v u I I and XbaI, and hybridized to t h e entire 16.5 kb n o r m a l m i t o c h o n d r i a l g e n o m e a s described in the M e t h o d s . (A) A f t e r P v u I I digestion, n o r m a l k i d n e y tissue DNA displays a single hybridization signal a t 16.5 k b , while t h e DNA from tumour tissue s h o w s three smaller m a i n fragments. (B) Mitochondria1 DNA of t h e normal kidney tissue s h o w s five fragm e n t s after restriction with Xbal, corresponding t o t h e sequences of t h e mitochondrial genome. The DNA from t h e chromophobe RCC shows seven fragments; two of t h e m , the 1.9 and 1.7 kb fragments, are identical t o t h o s e of n o r m a l m i t o c h o n d r i a

alteration of chromosomal DNA in chromophobe RCCs results in genetic and morphological changes of the mitochondria remains to be established. Previous studies have shown an abnormal pattern of mitochondrial DNA in renal oncocytomas as ~ e l l . ~ 3The " most prominent morphological feature of oncocytic cells is an abundance of normal and 'abnormal' mitochondria which gives the tumour a characteristic gross and light microscopic appearance. A transition between cells with a high number of mitochondria and a low number of vesicles and cells with a low number of mitochondria and a high number of vesicles has been found in some chromophobe R C C S . There ~ ~ is a morphological similarity between the mitochondria-rich chromophobe RCC and renal oncocytoma. The cells of both tumours express the enzyme carbonic anhydrase C, but the band 3 protein, an anion exchange protein, has been found to be expressed only in o n c o c y t ~ r n a sThese .~~ phenotypic profiles correspond to two subtypes

1. Kovacs G. Application of molecular cytogenetic techniques to the evaluation of renal parenchymal Lumours. Guest Editorial. J Cancer Res Clm Oncul1990; 116: 318-323. 2. Kovacs G , Erlandsson R, Boldog F. el a/. Consistent chromosome 3p deletion and loss of heterorygosity in renal cell carcinomas. Proc Nut/ AcndSci USA 1988; 85: 1571-1575. 3. Kovacs G, Frisch S. Clonal chromosome abnormalities in tumor cells from patients with sporadic renal cell carcinomas. Cancer Re.? 1989; 49651-659. 4. Kovacs G, Emanuel A, Neumann HPH, Kung H. Cytogenetics of renal cell carcinomas associated with von Hippel Lindau disease. Gene,\ Chromo.somesCmicrr 199 1 ; 3 256-262. 5. Bergenheim U, Nordenskjold M, Collins VP. Deletion mapping in renal cell carcinoma. Cancer Res 1989; 4 9 1390-1396. 6 . Dal Cin P, Gaeta J, Huben R, Li FP, Prout G R , Sandberg AA. Renal cortical tumors. Am J Clin Parhol1989; 9 2 408-414. 7. Kovacs G. Papillary renal cell carcinoma: a morphologic and cytogenetic study ofeleven cases. Am J Pathol 1989; 134: 27-34. 8. Kovacs G, Fuzesi L, Emanuel A. Kung H. Cytogenetics of papillary renal cell tumors. Genes Chromosomes Cancer 1991; 3 249-255. 9. Kovacs G, Welter C, Wilkeiis L, Blin N, DeRiese W. Renal oncocytoma: a phenotypic and genotypic entity of renal parenchymal tumors. Am J Patho/ 19x9; 134 967-971 10. Welter K , Kovacs G, Seitz G, Blin N. Alteration of mitochondrial DNA in human oncocytomas. Genes Chromo.some.r Cancer 1989; 1: 79-82. I I . Thoenes W, Storkel S, Rumpelt HF, Moll H, Baum HP, Werner S. Chromophobe cell renal carcinoma and its variants-report on 32 cases. J Parhol1988; 155: 271-287. 12. Thoenes W. Storkel S, Rumpelt H F . Histopathology and classification of renal cell tumors (adenomas, oncoyctomas and carcinomas). Parho[ Re.r Procr 1989; 181: 125-143. 13. Bannasch P, Schacht V, Storch E. Morphogenese und Mikromorphologie epithelialer Nierentumoren bei Nitrosomorpholin vergifteten Ratten. 1. Induktion und Histologie der Tumoren. Z . Krebsforsch 1974; 81: 31 1-33 I . 14. Human GeneMapping 10.Tenth International Workshopon Human Gene Mapping. Cyfogenel Cell Grnet 1989; 51: 688-802. I S . Kovacs G, Kung H. Nonhomologous chromatid exchange in hereditary and sporadicrenalcell carcinomas. Proc Nut/ AcadSci U S A 1991; 88: 194-198. 16. Kovacs G. Wilkens L, Papp T. DeRiese W. Differentiation between papillary and nonpapillary renal cell carcinomas by DNA analysis. J NurlCancer Inst 1989; 81: 527-530.

D N A ALTERATIONS I N CHROMOPHOBE RCC 17. Kovacs G, Soudah B, Hoene E. Binucleated cells in a human renal cell carcinoma with 34 chromosomes. Cuncer Genrt Cylogenet 198X; 31: 21 1-216.

18. Kovacs A, Kovacs G. Low chromosome number in chromophobe renal cell carcinomas. Genes Chromosomes Cancer 1992; 4 267-268. 19. Storkel S , Steart PV, Drenckhahn D, Thoenes W. The human chromophobecell carcinoma: its probable relation to intercalated cells of the collecting duct. Yirchows A w h B [Cell Pathol] 1989; 5 6 237-245. 20. Bonsib SM. Lager DF. Chromophobe cell carcinoma: analysis o f five cases. Am J Surg Pnthol1990; 14: 260-265.

277

21. Zeviani M, Servidei S, Gellera C, Bertini E, DiMauro S , DiDonato S .

An autosomal dominant disorder with multiple deletion of mitochondrial D N A starting a t the D-loop region. Nature 1989; 339 309-3 I I . 22. Thoenes W, Storkel S , Rumpelt HJ, Moll R. Cytomorphologicdl typing of renal cell carcinoma-a new approach. Eur Urol 1990; 1 8 6-9. 23. Storkel S , Pannen B, Thoenes W, Steart PV, Wagner S , Drenckhahn

D. Intercalated cells as a possible source for the development of renal oncocytoma. Virchow Arch B [Cell Pathol] 1988; 5 6 185-189.

Mitochondrial and chromosomal DNA alterations in human chromophobe renal cell carcinomas.

Renal cell tumours are characterized by the loss of chromosome 3p and trisomy of 5q segments (common, non-papillary renal cell carcinoma), or by triso...
381KB Sizes 0 Downloads 0 Views