Expression of bcl-2, bcl-x, bax and bak in Renal Parenchyma, Oncocytomas and Renal Cell Carcinomas

PATHOLOGY RESEARCH AND PRACTICE

© Gustav Fischer Verlag

Expression of bel-2, bel-x, bax and bak in Renal Parenchyma, Oncocytomas and Renal Cell Carcinomas Johannes Pammer, Markus Exner, Heinz Regele, Andrea Haitel, Wolfgang Weninger 1, Reinhard Horvat and Martin Susani Institut fOr Klinische Pathologie and 'Universitatsklinik fOr Dermatologie, Aligemeines Krankenhaus Wien, Osterreich

Summary Proteins of the bel-2 family are important regulators of programmed cell death. Alterations in the expression of these proteins may contribute to the progression of cancer. Expression ofbcl-2, bel-x, bax and bak was investigated by immunohistochemistry and Western-blotting of regular and alterated renal parenchyma as well as in 57 renal cell carcinomas. Bcl-2, bel-x and in part bax were found to be overexpressed in inflammed renal parenchyma, whereas atrophic tubuli predominantly stained for bel-2 and to a lesser degree for bel-x and bax. Only little bak expression was detected in alterated tubuli. Moderate to strong expression for bel-2, bel-x, bax and bak was found in 24,38,2 and 13 of 57 carcinomas, respectively. Bel-2, bel-x, bax and bak expression were correlated to tumor type. Chromophilic carcinomas stained stronger for bcl-2, bel-x and bax, whereas chromophobic carcinomas stained stronger for bel-x, bax and bak compared to elear cell carcinomas. Expression of bak correlated with that of bel-x and with an unfavorable histology as indicated by nuclear grading in these tumors. Our findings suggest that expression of bcl-2 and bel-x may be important for cell survival only in a subset of renal cell carcinomas, and that the antiapoptotic effect of these proteins appears to be frequently bypassed possibly by other factors impeding programmed cell death. Key words: Renal cell carcinoma - bel-2 - bel-x - bax -bak

Introduction Members of the bel-2 family are regulators of programmed cell death (PCD). These proteins determine Pathol. Res. Pract. 194: 837-845 (1998)

the fate of the cell by forming homo- and heterodimers at different relative concentrations causing apoptotic and anti-apoptotic effects [36J. Therefore expression of these genes is involved in the homeostasis of normal cell turnover mechanisms and may contribute both to the development of cancer as well as tumor responses to chemo- and radiotherapy [36J. The best investigated regulator of PCD is bcl-2. Expression of bcl-2 protein blocks many instances of PCD [28]. It is expressed in hematopoietic cell lineages and a limited number of nonlymphoid tissues. Immunoreactivity for the bel-2 product is present mainly in cell populations which are long lived and/or with high proliferative ability such as cells at the bottom of colonic crypts, duct cells of exocrine glands, basal cells of squamous epithelia, hematopoietic stem cells and several embryonic tissues [13, 20, 21]. Bel-2 was first identified in non-Hodgkin's follicular B-celllymphomas, where the balanced t (14;18) chromosomal translocation results in enhanced bel-2 protein expression. Overproduction of bel-2 also occurs in a wide variety of carcinomas (gastrointestinal cancer [3J, prostatic cancer [22J, lung cancer [14, 27 J, and breast cancer [8]), suggesting a possible role of this protein in the genesis of these neoplasms by suppressing normal cell turnover. Furthermore, overexpression of bel-2 provides protection against radiation and essentially all currently available chemotherapeutic drugs [28,29]. The bel-x gene encodes two distinct protein products, bel-xs (short, 170 aa) and bel-xL (long, 230 aa), Address for correspondence: Dr. Johannes Pammer, Institut fUr Klinische Pathologie, Allgemeines Krankenhaus, UniversiUit Wien, A - 1097 Vienna, Austria. Tel.: (43)-1-404 00-36 51, Fax: (43)-1-40400-3690 0344-0338/98/0194-0837$5.00/0

838 . J. Pammer et aI.

generated by differential mRNA splicing [I]. Bel-xL inhibits peD even under conditions where bel-2 does not [2J, whereas bel-xs dominantly promotes cell death [1J. Bel-x is found in neural cells, several cell types in the bone marrow, reproductive tissues and a variety of epithelial cells. The pattern of bel-x expression differs from that ofbel-2 in many organs, suggesting that bel-x and bel-2 regulate apoptosis at different stages of cell differentiation [19]. Bel-x is overexpressed in colorectal tumors [17 J and in high grade prostate cancer [15 J. Bax dimerizes with bel-2, bel-xL and itself and is a death promoter [24, 30]. Bax is expressed at high levels in the murine nervous system, epithelia of uterus, ovary and the gastrointestinal tract [16]. Expression of bax is not significantly altered in colorectal tumors [17 J compared to normal mucosa, but frequently upregulated in thyroid carcinomas dependent on the histologic type [2]. Recently an additional bel-2 homologue with bel-2 antagonistlkiller activity, bak, has been identified, that binds both to bel-2 and bel-xL, but not to bel-xs, and can rigger apoptosis [7, 10]. Expression of bak is widespread in human tissues, but differs from that of other members of the bel-2 family. It is found in a variety of epithelial cells, granulocytes, cardiomyocytes and smooth musele cells [18]. Reduction in bak expression occurs early in colorectal tumor progression [17]. In normal human kidney, bel-2 is consistently expressed on glomerular parietal epithelial cells, Henle's loop and epithelial cells belonging to the distal tubules, but only on single epithelial cells of proximal tubules and collecting ducts [5, 25]. In contrast to elear cell carcinomas that are mostly negative or only weakly positive for bel-2, tubulopapillary carcinomas distinctly stain for bel-2 [25]. The most intense staining for bel-x in renal parenchyma was detected in Henle's loop, distal tubules and collecting ducts, whereas glomerular epithelia and proximal tubules were reported to be either negative or only weakly positive [19]. The expression pattern of bax in human renal parenchyma has not yet been described. Expression of bak has been reported to be strong in descending thin limbs and distal convoluted tubules [18J. In this study, we investigated the in-situ expression ofbel-2, bel-x, bax and bak in normal and inflammatory altered renal parenchyma, as well as in renal cell carcinomas by immunocytochemistry and immunoblotting. In-situ expression was assessed in its relation to morphology and grade of malignancy.

Material and Methods Tissue samples and classification

Fonnalin-fixed, paraffin-embedded tissue of renal carcinomas was obtained from the pathology files of the Institute of

Clinical Pathology at the University of Vienna. Cases consisted of 37 clear cell carcinomas, 10 chromophilic carcinomas, 10 chromophobic carcinomas, 4 oncocytomas and regular renal parenchyma. Nonnal and chronically damaged renal parenchyma, i.e. fibrotic tissue with atrophic tubuli and interstitial inflammatory infiltrate, was also found adjacent to the neoplasias. The carcinomas were classified according to Thoenes [33]. To confinn the diagnosis of chromophobic renal cell carcinomas, sections were stained with Hale's acid colloidal iron method modified by Mowry [4} and treated with antibodies to epithelial membrane antigen (EMA) [31]. These sections were processed by indirect immunohistochemical method using horse radish peroxidase-labeled avidin-biotin staining technique (Vector, Burlingame, CA). Immunohistochemistry

All samples were immunostained for bcl-2, vimentin, bak and bax expression using a monoclonal antibody to bcl-2 (bcl2-124, Dako, Glostrup, Denmark), vimentin (V9, Dako) and bak (TC-lOO, Calbiochem, Cambridge, MA) and a rabbit polyclonal Ab to bax (Ab-I, Calbiochem). TRIS-buffered saline was used as a washing-buffer. Dewaxed sections were subsequently boiled in 0.0l mollL citrate buffer (pH 6) at 450 W twice for 10 min in a microwave oven and then cooled to room temperature. Primary antibodies were applied at 1/50 (bcl-2), 1/400 (vimentin), 1/25 (bax) or 1/40 (bak) dilution. Staining for bcl-x (rabbit polyclonal Ab, Transduction Laboratories, Lexington, KY) was perfonned on untreated paraffin sections at a dilution of 1: 1000. Immunohistochemical staining was perfonned using an APAAP kit (Dako). The enzyme was developed with an alkaline phosphatase substrate kit (Vector). Negative controls, carried out on consecutive tissue sections of each tissue sample in which an isotype (IgG 1 and IgG2a, Coulter, Hialeah, FL) was used as a primary antibody or the primary antibody was omitted, resulted in no detectable staining. The specificity of the rabbit polyclonal Abs was also tested by Western blot analysis. Relative staining intensity for bcl-2, bak and bcl-x was arbitrarily designated 1+, 2+, 3+ (Table 1). Renal cell carcinomas were graded according to Fuhnnan [ll}. Staining intensities and grading were assessed independently by two clinical pathologists. Statistical analysis was performed by Mann Whitney Wilcoxon test, Speannan's rank test and Pearson's X2 test. The level of significance was set at p < 0.05. SDS-Gelelectrophoresis and immunoblotting

Renal tissue was freshly obtained from surgery and immediately frozen in liquid nitrogen. Tissue samples were lysed in NP-40 buffer (l % NP-40 in PBS) containing 1 mM PMSF. Table 1. Grading of the staining intensity of renal cell carcinomas

o 1

2 3

>2/3 of the carcinoma is negative >1/3 at least weakly positive >1/2 distinctly positive >1/3 strongly positive

(negative) (weak) (moderate) (strong)

Regulatory Apoptotic Proteins in Renal Cell Carcinomas . 839 Protein concentrations were measured by a Micro BCA Protein Assay Reagent (Pierce, Rockford, IL). Lysates were subjected to SDS-PAGE in 12% gels. Prior to SDS-PAGE, lysates were mixed with reducing SDS-loading buffer (7.2% SDS, 9 mmol/L EDTA, 20% glycerol, 10 mmol/L dithiothreitol, 13 mmol/L Tris-phosphate buffer, pH 6.8), thereby adjusted to equal concentrations and boiled for 5 min. Separated proteins were then electroblotted onto a nitrocellulose membrane (Schleicher & Schull, Germany) at 0.8 mAlcm2 for 2 hours. The membranes containing the complete set of samples were dried and blocked in blocking buffer (5% non-fat dry milk in PBS) followed by immune overlay with the following antibodies: bcl-2 (0.5 Ilg/ml), bak (1.0 Ilg/ml), bax (1.6 Ilg/ml) and bcl-x (l Ilg/ml). After washing, bound antibodies were detected with HRP-labeled sheep anti mouse-IgG (Amersham Life Science, UK) or HRP-labeled goat anti-rabbit IgG (Pierce, Rockford, IL), respectively. After rinsing in Western blotting detection reagent ECL (Amersham) membranes were exposed to a X-OMAT-AR film (Eastman Kodak, NY) for 5 min, 30 min and 60 min each.

bcl-2

31 21 1 2 3

4

5 6

7

8

9

10 11 12 1314

4

5 6

7

8

9

10 11

bcl·2

31

21 1 2 3

43 -

12 13 14

bcl-x

31 21

1 2

3

4

5 6

7

4 5

6

7

8

9

10 11 12 1314

bax

Results Analysis of bel-2. bel-x. bax and bak Expression by Western Blotting

To confirm the specificity of the Abs used for immunohistochemistry and to study the biochemical properties of the antigen expression in renal tissue and renal cell carcinomas, Western blot analysis was performed. As presented in Fig. 1, the Abs for bcl-2, bcl-x, bax and bak detected proteins of approximately 26, 31, 21 and 26 kd, respectively, corresponding to the molecular weight of these antigens [18, 19, 24J. Monoclonal Abs to bcl-2 specifically recognized a single band of 26 kd, the expected size for bcl-2a. Longer exposure revealed an additional band of 22 kd corresponding either to the shorter splice variant bcl-2~ [35J or a proteolytically cleaved form of bcl-2 with pro-apoptotic activity [6]. The 31 kD protein exhibits the established size of the bcl-xL protein, whereas the smaller splice form, bel-xs, was not detected in renal tissue and renal cell carcinomas. This indicates that the immunohistochemical staining for bel-x in renal tissue predominantly detects bel-xL. Immunoblots of bel-2, bcl-x and bak showed double bands probably due to phosphorylation of proteins. The expression of bel-2, bel-x, bax and bak in tissue lysates of chromophilic carcinomas (Fig. 1, lanes 1 to 3) and elear cell carcinomas (lanes 4, 5, 7, 8, 9, 11) was heterogenous. The lysate of one chromophobic renal cell carcinoma (Fig. 1, lane 6) was strongly positive for bel-x and bak. Bel-2 and bax (not shown) could be detected in this tumor after longer exposure of the film. Regular renal tissue (Fig. 1, lanes 10, 12-14) contained relatively high amounts of bel-2, bcl-x and bak. Bel-2 was distinctly expressed in all 4 Iysates of regular renal

21 14 1 2 3

8

9

10

11 12 13 14

8

9

10 11 12 1314

bak

31

21 1 2

3

4

5

6

7

Fig. 1. Immunoblot analysis of bcl-2, bcl-x, bax and bak protein in standardized lysates of renal cell carcinomas (chromophilic renal cell carcinomas: lanes 1, 2, 3; chromophobic renal cell carcinoma: lane 6; clear cell carcinomas: lanes 4 [G2], 5 [G3], 7 [GI], 8 [G3], 9 [G2], II [G3]) and normal renal cortex (lanes 10, 12-14). Monoclonal Abs to bcl-2 specifically recognize a single band of 26 kd, the expected size for bcl-2a. Longer exposure revealed an additional band of 22 kd corresponding either to the shorter splice variant bcl-2~ or cleaved bcl-2. Antibodies to bcl-x, bax and bak recognize specific bands of 31, 21 and 26 kd, the expected molecular weight of these proteins.

tissue, but only in lysates of 5 of 10 tumors (lanes 1, 5, 7, 8, 11). Large amounts of bel-x were found in 3 of 4 lysates of regular renal tissue (lanes 10, 13, 14), but only in 4 carcinomas (lanes 3-6). Similarly bak was expressed strongly in 3 of 4 lysates of renal tissue (lanes 10, 13, 14), but in only one renal cell carcinoma (lane 2). Quantitative analysis of bax expression of Western blots was not possible with the antibody used.

840 . J. Pammer et al. Analysis of bcl-2, bel-x, bax and bak Expression by Immunocytochemistry

Normal Renal Tissue (Table 2) bcl-2 (Fig. 2a): Parietal epithelial cells of glomerula were strongly positive for bcl-2, whereas visceral epithelial cells were negative. Weak staining could be de-

tected in some proximal tubules in normal renal tissue. Most of the cells of Henle's loop were either negative or labelled weakly for bcl-2; only few were distinctly positive. Furthermore, staining of ascending limbs of Henle's loop was usually stronger than that of the distal convoluted tubules and staining of distal convoluted tubules, that could be identified by treating serial sec-

Fig. 2. a: By immunocytochemistry, bcl-2 is found in parietal Bowman's capsule, single cells of proximal tubules (arrows) and in distal tubules. b: Bcl-x is expressed weakly in proximal and strongly in distal tubules. c: Bax is weakly expressed in some distal tubules and in smooth muscle cells of vessels (arrows). d: These collecting ducts stain moderately for bax. The staining is most prominent located at the luminal portion of the cells. e: Bak expression is very faint in proximal and moderate in distal tubules (arrows). Weak bak expression in parietal Bowman's capsule (arrowhead) is unusual in regular renal parenchyma. f: Single cells of collecting ducts stain strongly for bak.

Regulatory Apoptotic Proteins in Renal Cell Carcinomas . 841

tions with an anti-CK19 moAb (data not shown), varied from negative to distinctly positive. Collecting ducts were negative to moderately positive for bel-2 and frequently single cells in the majority of these ducts expressed bel-2 strongly. Basal cells of the renal pelvis were focally positive for bel-2. bcl-x (Fig. 2b): Glomerular cells were negative for bel-x. Epithelia of proximal tubules of regular tissue

were either negative or weakly positive; those of thin segments of Henle's loops were negative to moderately positive. The vast majority of distal convoluted tubules stained strongly for bel-x. Bel-x expression of collecting ducts was homogeneous and weak. Staining both of tubuli and of ducts was in outlines granular. Transitional epithelium stained weakly to moderately for bel-x being most prominent in basal cells.

Fig. 3. a: Expression of bcl-x is strong in chromophilic renal cell carcinomas. b: A diffus or perimembraneous cytoplasmic staining for bcl-x is found in this chromophobic renal cell carcinoma. c: This chromophilic renal cell carcinoma focally stains strongly for bax. d: A chromophobic renal cell carcinoma stains moderately for bax. Note strong staining of few atrophic tubuli (arrows) and neutrophil granulocyte (arrowhead). e: Staining for bak in chromophobic renal cell carcinomas is cytoplasmic. f: Tumor cells with marked nuclear atypia stain strongly for bak in this clear cell carcinoma.

842 . J. Pammer et al.

bax (Fig. 2c, d): Proximal tubuli expressed bax only focally and weakly. Weak staining for bax was also detected in most ascending limbs of Henle's loop and some distal convoluted tubules. Staining intensity of collecting ducts was homogenous and weak to moderate, that of transitional epithelium weak. bak (Fig. 2e, f): Glomerular cells did not express bak. The staining of proximal tubules was negative to weak. Positive proximal tubular cells often exhibited a granular staining pattern. Henle's loops were negative or weakly positive. The staining intensity of distal tubules varied from negative to moderate and the staining of the ascending limb was stronger than that of distal convoluted tubules. Collecting ducts stained moderately, single cells were strongly positive. The collecting ducts were the structures strongest positive for bak within renal parenchyma. Staining of the transitional epithelium was weak to moderate and homogenous. Focally basal cells and luminal umbrella cells were stronger positive. Alterated Renal Parenchyma

Bcl-x could be found weakly in few podocytes and strongly in parietal epithelial cells in inflammation.

Bcl-2 and bcl-x were distinctly upregulated in morphologically normal tubuli and collecting ducts within inflammed areas and adjacent to tumors. Bcl-2 was expressed strongly in the majority of atrophic tubuli, whereas bcl-x expression varied from negative to moderate in these tubules. Bax was weakly upregulated in tubuli within inflamed areas. Staining of atrophic tubuli for bax varied from negative to strongly positive. Both visceral and parietal epithelial cells of some alterated glomerula adjacent to tumors became weakly to moderately positive for bak. Weak to moderate expression of bak was also seen in a minority of atrophic tubuli. Renal Cell Carcinomas and Oncocytomas (Table 3)

bel-2: In the majority of all types of carcinomas, more than 50% of the tumor cells expressed bcl-2. Staining was both intracytoplasmic and membranous. The staining for bcl-2 was significantly stronger in chromophilic (tubulopapillary) renal cell carcinomas compared to clear cell carcinomas. bel-x: Expression of bcl-x was significantly stronger in chromophilic (Fig. 3a) and chromophobic (Fig. 3b)

Table 2. Expression ofbel-2, bel-x, bax and bak in renal parenchyma

Visceral Bowman's capsule Parietal Bowman's capsule Proximal convoluted tubule Thin segment of Henle's loop Ascending limb of Henle's loop Distal convoluted tubule Collecting duct Transitional epithelium

bel-2

bel-x

bax

bak

o

o o

o o

o o

2-3 0(-1)* 0-3 1-3 (0-)2 0-2* 0-2**

0-1 0-2 1-2 (1-)3

1-2 1-2**

0(-1)

o

(0-)1 0-1 1(-2) 1(-2)

0-1 0-1 0-2 0-2 2* 1-2*

Staining intensities as described in Table I. * Single cells were strongly positive for bcl-2 or bak. ** Staining of the transitional epithelium was stronger basally. Table 3. Expression pattern ofbel-2, bel-x, bax and bak in renal cell carcinomas

bel-2 bel-x bax bak

Clear cell type n=37

Chromophilic n= 10

Chromophobic n= 10

Oncocytomas n=4

12/12119113" 12/6113/6b. c 2711 0/0/0", c 20110/512d

1/2/2/5"·b 0/1/3/6b 3/7/0/0" 2/51211

3/3/4/0b 0/0/4/6c 2/512/0<* 1/61211 d

1/1/2/0 0/11112 0/4/0/0 1/1/1/1

Number of respective tumor cases, correlating to staining intensities as described in Table 1 (negative/weak/moderate/strong expression). Significancies of differences in staining intensities of different histologic types as assessed by Mann Whitney Wilcoxon test: ": p < 0.02; b: p < 0.004; c: p < 0.002; d: p > 0.04. * Only 9 carcinomas were investigated.

Regulatory Apoptotic Proteins in Renal Cell Carcinomas . 843 Table 4. Correlations between nuclear grading and expression ofbcl-2, bcl-x, bax and bak in clear cell carcinomas

bcl-2 bcl-x bax bak

grading

bcl-2

n.s. 0.35 (p < 0.034)' n.s. 0.59 (p < 0.0001)5

n.s. n.s. n.s.

bcl-x

bak

n.s.

n.s. 0.77 (p < 0.0001)2 n.s. 4

0.47 (p < O.OOw 0.77 (p < 0.0(01)2

data: Spearman's rank correlation: r sub s (significance) Pearson X3 test: 1n.s., 2p < 0.0001, 3p < 0.003, 4p < 0.025, 5p < 0.019, n.s. =not significant.

compared to clear cell carcinomas. Similar to bcl-2 staining bcl-x reactivity was predominantly intracytoplasmic and in some instances membranous (not shown). Single tumor cells exhibited also a weak nuclear staining. bax: Staining for bax was intracytoplasmic and significantly stronger in chromophilic (Fig. 3c) and chromophobic (Fig. 3d) compared to clear cell carcinomas. A weak nuclear staining in single tumor cells was also detected for bax. bak: Bak was significantly stronger expressed in chromophobic (Fig. 3e) compared to clear cell carcinomas (Fig. 3f). There was a tendency of stronger bak staining in chromophilic renal carcinomas. In contrast to bcl-2 and bcl-x, reactivity for bak was predominantly intracytoplasmic. When the staining intensities of clear cell carcinomas was compared with their commonly supposed histogenetic origin, i.e. proximal tubules [32, 37 J, we found stronger expression of bcl-2 in the majority (2/3) of these tumors, which is in accordance with a previous report [25J. Overexpression of bcl-x could be found in half of the cases compared with proximal tubules in non-inflamed tissue; bak only in a minority of cases. In contrast, chromophilic carcinomas stained for bcl-2 and bcl-x distinctly stronger in 9 of 10 cases and for bak in 8 of 10 cases compared to normal proximal tubules. In chromophobic carcinomas, bcl-2, bcl-x and bak expression was stronger in 5, 7 and 2 of 10 tumors, respectively, whereas expression of these antigens was weaker in 4, 1 and 6 tumors compared to the average staining intensity of collecting ducts. No overexpression of bax in renal cell carcinomas could be found compared to their supposed histogenetic origins, since strong staining for bax was only detected in small areas of 3 clear cell carcinomas. In summary, overexpression of bcl-2, bcl-x and bak was detected in 39, 35 and 27 carcinomas compared to tubules, which are their presumable histogenetic origin, i.e., proximal tubules and collecting ducts, and moderate to strong expression for these proteins was found in 24, 38 and 13 of these 57 tumors, respectively. 22 of 37 clear cell carcinomas and all chromophilic and chromo-

phobic carcinomas expressed either bcl-2 or bcl-x moderately to strongly. In oncocytomas, staining for bcl-2, bcl-x and bak was very heterogenous and varied from negative to strongly positive, whereas all 4 oncocytomas stained weakly for bax. Correlation ofbcl-2, bcl-x, bax, bak, and Nuclear Grading (Table 4)

Coexpression of bak and bcl-x in clear cell carcinomas was highly significant, whereas the correlation of staining for bax and bcl-x was only weak. Expression of bax and bak was significantly correlated in clear cell carcinomas only by Pearson's X2 test. Nuclear grading correlated with bak and to a lesser degree with bcl-x expression as assessed by Spearman's rank test. No significant correlations were detected within the other types of renal cell carcinomas.

Discussion In this report, we investigated the expression pattern

of bcl-2, bcl-x, bax and bak in renal parenchyma and

renal cell carcinomas using immunocytochemistry and immunoblotting. Our data on bcl-2 expression differ in some details from previous reports [5, 25J. Similar to Lu et al. [21], we did not detect bcl-2 in all distal tubules. This finding is not in accordance with other reports of consistent bcl-2 expression in distal tubules and has been related to differences in tissue fixation and immunohistochemical techniques [5 J. To assess tissue preservation and accessibility of antigens, we performed staining for vimentin in all cases and for cytokeratin 19 in few cases. Vimentin was found in all renal cell carcinomas except the chromophobic subtype and cytokeratin 19 was constitutively expressed in distal tubules as described by Moll et al. [23, 26J. These results suggest that failure of bcl-2 staining in these tubules is not caused by a technical problem but indicates a lack of in-situ expression. The meaning of differences in the expression level of members of the bcl-2 family in renal cell carcinomas is

844 . 1. Pammer et al.

not evident at the moment. Whereas clear cell carcinomas are supposed to be derived from proximal tubules [32, 37 J, chromophilic renal cell carcinomas show signs of maturation both toward proximal and distal tubules [9, 32J, so that some of these tumours may originate from distal tubules. Therefore stronger expression of bcl-2 in chromophilic renal cell carcinomas as compared to clear cell carcinomas has been related to the histogenetic origin of these tumours [25]. Similarly stronger expression of bcl-x in chromophilic renal cell carcinomas compared to clear cell carcinomas and the rather strong expression of bcl-x and bak in chromophobic renal cell carcinomas, that are derived from intercalated cells of collecting ducts, may be related to the origin of these tumours. To test this hypothesis, we compared the variable expression of bcl-2, bak and EMA within collecting ducts by staining of serial sections of regular renal tissue. However, we found no correlation between stronger bcl-2 and bak expression and cytoplasmic EMA-staining, that characterized intercalated cells (data not shown) [32]. Thus, so far the varying expression of apoptosis-regulating proteins in renal cell carcinomas cannot be explained by their different histogenesis. When we compared the staining intensities of bcl-2, bcl-x, bax and bak in clear cell carcinomas, chromophilic and chromophobic carcinomas with their supposed histogenetic origins, i.g. proximal tubules and collecting ducts [31, 32, 37 J, we found overexpression of bcl-2, bcl-x and bak, but not bax, in 39, 35 and 27 of the 57 tumors. 24 tumors stained moderately to strongly for bcl-2 and 38 tumors for bcl-x. Like bcl-2, the bcl-xL protein has been shown to inhibit apoptosis induced by growth factor withdrawal, antimetabolites, and many anticancer drugs. The upregulation of bcl-2 and bcl-x may contribute to the development and progression of these tumors, as well as to the resistance to chemotherapy [28, 29, 36J. As only 22 of 37 clear cell carcinomas stained moderately to strongly for either bcl-2 and bclx, the role of these anti-apoptotic effectors appears to be frequently bypassed in these neoplasms by other factors impeding programmed cell death. In addition to our immunohistochemical data, Western blots of regular renal tissue showed high amounts of bak and also to a minor degree of bcl-2 and bcl-x in contrast to renal carcinomas. As we compared clear cell carcinomas and chromophilic carcinomas with proximal but not distal tubules, this could be explained partly by stronger expression of bcl-2, bcl-x and bak by distal tubules present within total parenchymal lysates. Another reason for that discrepancy may be larger amounts of epithelia in regular parenchyma compared to tumor tissue containg variable quantities of tumor stroma. Furthermore, antigens detected in the blots are not necessarily exclusively derived from tumor cells. Total tumor samples do not only contain unknown amounts

of connective tissue and possible areas of necrosis and hemorrhages, but also inflammatory infiltrates. This makes the comparison among different lysates difficult. In the current report, we found a strong correlation between expression of bcl-x and bak and a weak correlation between nuclear grading and expression of bak in clear cell carcinomas. Immunoblotting clearly showed that bak and bcl-x expression is authentic and not due to corrections of the antibodies used. Bak binds to bcl-xL [lOJ and, to a minor degree, to bcl-2 [7]. Thus the correlation of bcl-x and bak expression may be related to a functional coexpression of these two proteins in renal cell carcinomas. The correlation between bak-staining and nuclear atypia proposes that upregulated bak expression may be a poor prognostic marker in these tumors. Strong expression of a proapoptotic protein in tumors of higher malignancy seems surprising, however, a correlation of unfavorable histology, mitotic rate and the number of apoptotic cells has been described in renal cell carcinomas [34]. In conclusion, this report documents direct correlation of bcl-2, bcl-x, bax and bak expression to tumor type insofar that chromophilic tumors stain stronger for bcl-2, bcl-x and bax, whereas chromophobic tumors stain stronger for bcl-x, bax and bak compared to clear cell carcinomas. Moderate to strong expression for bcl-2 was found in 24, for bcl-x in 38, for bax in 2 and for bak in 13 of 57 carcinomas. Expression of bak correlates with that ofbcl-x and with an unfavorable histology indicative of poor clinical prognosis in these tumors.

References 1. Boise LH, Gonzalez-Garcia M, Postena CE, Ding L, Lindsten T, Turka LA, Mao X, Nunez G, Thompson CB (1993) Bel-x, a bel-2 related gene that functions as a dominant regulator of cell death. Cell 74: 597--608 2. Branet F, Brousset P, Krajewski S, Schlaifer D, Selves 1, Reed IC, Charon P (1996) Expression of the cell deathinducing gene bax in carcinomas developed from the follicular cells of the thyroid gland. 1 Clin Endocrinol Metab 81:2726-2730 3. Bronner MP, Culin C, Reed IC, Furth EE (1995) The bel2 proto-oncogene and the gastrointestinal epithelial tumor progression model. Am 1 Pathol146: 20-26 4. Burck HC (1983) Histologische Technik. Thieme Verlag, Stuttgart 5. Chandler D, EI-Naggar AK, Brisbay S, Redline RW, McDonnel TJ (1994) Apoptosis and the expression of the bel-2 proto-oncogene in the fetal and adult human kidney. Hum Patho125: 789-796 6. Cheng E, Kirsch DG, Clem RI, Ravi R, Kastan MB, Bedi A, Ueno K, Hardwick 1M (1997) Convention of bel-2 to a bax-like death effector by caspases. Science 278: 1966-1968

Regulatory Apoptotic Proteins in Renal Cell Carcinomas . 845 7. Chittenden T, Harrington EA, O'Connor R, Flemington C, Lutz RJ, Evan GI (1995) Induction of apoptosis by the bcl-2 homologue bak. Nature 374: 733-736 8. Doglioni C, Cei Tos AP, Laurino L, Chiarelli C, Barbareschi M, Viale G (1994) The prevalence of immunoreactivity in breast carcinomas and its clinicopathological correlates, with particular reference to oestrogen receptor status. VirchowsArchiv 424: 47-51 9. Droz D, Patey N, Paraf F, Chretien Y, Gogusev J (1995) Composition of extracellular matrix and distribution of cell adhesion molecules in renal cell tumors. Lab Invest 7l:71~718

10. Farrow SN, White JHM, Martinou I, Raven T, Pun KT, Grinham CJ, Martinou JC, Brown A (1995) Cloning of a bcl-2 homologue by interaction with adenovirus E18 19K. Nature 374: 731-733 11. Fuhrmann SA, Lasky LC, Limas C (1982) Prognostic significance of morphologic parameters in renal cell carcinomas. Am J Surg Pathol 6: 655-663 12. Gottschalk AR, Boise LH, Thompson CB, Quintans J (1994) Identification of immunosuppressant-induced apoptosis in a murine B-cell line and its prevention by bcl-x but not bcl-2. Proc Natl Acad Sci USA 91:

23. 24. 25. 26.

27.

28. 29.

735~7354

13. Hockenberry DM, Zutter M, Hickey W, Nahm M, Korsmeyer SJ (1991) BcI-2 protein is topographically restricted in tissues characterized by apoptotic cell death. Proc. Natl Acad Sci USA 88: 6961--6965 14. Ikegaki N, Katsumata M, Minna J, Tsujimoto Y (1994) Expression of bcl-2 in small cell lung carcinoma cells. Cancer Res 54: 6-8 15. Krajewska M, Krajewski S, Epstein JI, Shabaik A, Sauvageot J, Song K, Kitada S, Reed JC (1996) Immunohistochemical analysis of bcl-2, bax, bcl-x, and mcl-I expression in prostate cancers. Am J Pathol148: 1567-1576 16. Krajewska S, Krajewski M, Shabaik A, Wang HG, Reed JC (1994) Immunohistochemical determination of in vivo distribution of bax, a dominant inhibitor of bcl-2. Am J Patho1145: 1323-1336 17. Krajewska M, Moss SF, Krajewski S, Song K, Holt PR, Reed JC (1996) Elevated expression ofbcl-x and reduced bak in primary colorectal adenocarcinomas. Cancer Res 56:2422-2427 18. Krajewski S, Krajewska M, Reed JC (1996) Immunohistochemical analysis of in vivo patterns of bak expression, a proapoptotic member of the bcl-2 protein family. Cancer Res 56: 2849-2855 19. Krajewski S, Krajewska M, Shabaik A, Wang HG, Shinji I, Fong L, Reed JC (1994) Immunohistochemical analysis of in vivo patterns of bcl-x expression. Cancer Res 54: 5501-5507 20. leBrun DP, Warnke RA, Cleary ML (1993) Expression of bcl-2 in fetal tissues suggests a role in morphogenesis. Am J Path 142: 743-753 21. Lu Q, Poulsom R, Wong L, Hanby AH (1993) BcI-2 expression in adult and embryonic non-hematopoietic tissues. J Path 169: 431-437 22. McDonnel TJ, Troncoso P, Brisbay SM, Logothetis C, Chung LWK, Hsieh J, Tu S, Campbell ML (1992) Ex-

30.

31.

32. 33.

34.

35.

36. 37.

pression of the protooncogene bcl-2 in the prostate and its association with emergence of androgen-independent prostate cancer. Cancer Rec 52: 6940--6944 Moll R, Hage C, Thoenes W (1991) Expression of intermediate filament proteins in fetal and adult human kidney. Lab Invest 65: 74-86 Oltvai ZN, Milliman CL, Korsmeyer SJ (1993) BcI-2 heterodimerizes in vivo with a conserved homolog, bax, that accelerates programmed cell death. Cell 74: 609--619 Paraf F, Gogusev J, Chretien Y, Droz D (1995) Expression of bcl-2 oncoprotein in renal cell tumors. J Pathol 177:247-252 Pitz S, moll R, Storkel S, Thones W (1987) Expression of intermediate filament proteins in subtypes of renal cell carcinomas and in renal oncocytomas. Lab Invest 56: 642--653 Rao SK, Krishna M, Woda BA, Savas L, Fraire AE (1996) Immunohistochemical detection of bcl-2 protein in adenocarcinomas and non-neoplastic cellular compartments of thelung. Mod Mathol 9: 555-559 Reed JC (1994) BcI-2 and the regulation of programmed cell death. J Cell Bioi 124: 1--6 Reed JC (1995) BcI-2: prevention of apoptosis as a mechanism of drug resistance. Hematol Oncol Clin N Am 9: 451-474 Sedlak TW, Oltvai ZN, Yang E, Wang K, Boise LH, Thompson CB, Korsmeyer SJ (995) Multiple bcl-2 family members demonstrate selective dimerizations with bax. Proc Natl Acad Sci USA 92: 7838 Storkel S, Steart PV, Drenkhahn D, Thoenes W (1989) The human chromophobe cell renal carcinoma: its probable relation to intercalated cells of the collecting duct. Virchows Archiv B Cell Pathol56: 237-245 StOrkel S (1993) Karzinome und Onkocytome der Niere. 132, Gustav Fischer, Stuttgart - Jena - New York Thoenes W, Stoerkel S, Rumpelt HJ (1986) Histopathology and classification of renal cell tumors (adenomas, oncocytomas and carcinomas). Pathol Res Pract 181: 125-143 Todd D, Yang G, Brown RW, Cao J, D' Agati V, Thompson TS, Truong LD (1996) Apoptosis in renal cell carcinoma: detection by in-situ end labeling of fragmented DNA and correlation with other prognostic factors. Hum Pathol27: 1012-1017 Tsujimoto Y, Croce CM (1986) Analysis of the structure, transcripts, and protein products of bcl-2, the gene involved in human follicular lymphoma. Proc Natl Acad Sci USA 83: 5214-5218 Yang E, Korsmeyer SJ (1996) Molecular thanatopis: a discourse on the bcl-2 family and cell death. Blood 8: 386-401 Yoshida SO, Imam A, Olson CA, Taylor CR (1986) proximal renal tubular surface membrane antigens identified in primary and metastatic renal cell carcinomas. Arch Pathol Lab Med 110: 825-832

Received: December 20, 1997 Accepted in revised form: September 3, 1998