Vol. 14, No. 11/12, pp. 1041-1051, 1990. Printedin Great Britain.

0145-2126/90 $3.00 + .00 PergamonPressplc

Leukemia Research

EXPRESSION OF p53 IN HUMAN LEUKEMIC CELL LINES STEFAN KRAISS,* REGINA E S P I G , t ULRICH V E T T E R , t WOLFGANG HARTMANNt a n d MATHIAS MONTENARH*

*Department of Biochemistry and tDepartment of Pediatrics, University of Ulm, P.O. Box 4066, D-7900 Ulm, F.R. Germany (Received 15 April 1990. Revision accepted 7 July 1990)

Abstract--The cell-encoded p53 antigen seems to be tightly associated with various human malignancies. We have analyzed biochemical properties of p53 in two different cell lines derived from patients with ALL or ANLL. p53 was found in elevated levels in both leukemic cell lines compared to unstimulated or stimulated normal lymphocytes. High levels of p53 in these cell lines are due to an extended stability of p53 protein rather than to different rates of synthesis, p53 from both cell lines formed low- and high-molecular weight oligomers which revealed that p53 exists in a heterogenous population in these tumor cells. The presence of immunologically different subsets of p53 was demonstrated by sequential immunoprecipitation experiments with different p53 specific monoclonal antibodies. Our results showed structural and immunological variabilities of p53 in cell lines derived from human tumors and may thus provide an insight into the role p53 may play in human malignancies. Key words: p53, leukemic cell lines, human leukemia, oligomerization, protein stability, structural subclasses of p53.

A variety of other human tumors, including mamma carcinomas and leukemias as well as human tumor cell lines, express high levels of the p53 protein [15-18]. Thus, deregulation of the p53 expression which results either in complete absence or massive overexpression of p53 seems to be a common feature of tumors or tumor derived cells or transformed cells (for a review see: [2]). The mechanisms by which overexpression of p53 is accomplished in transformed cells varies in different cell types, p53 has been shown to form tight complexes with viral proteins like SV40 large T antigen or the E l b gene product of adenovirus [19, 20] which are required for virus mediated cell transformation. This complex formation leads to a dramatic increase in the stability of the p53 protein which results in an accumulation to levels considerably above those found in normal cells [21]. It has recently been shown that p53 is also stabilized in non-SV40 or non-adenovirus transformed cells without complex formation with viral or cellular proteins [22, 23]. There is ample evidence that cell transformation is accompanied by the formation of high-molecular weight otigomers of p53 and that these aggregation properties are independent of complex formation with SV40 T antigen, E l b protein or the heat shock protein hsc 70 [24, Kraiss et al., in press]. However, nothing is known about biochemical properties of p53 in human tumors and human tumor cell lines.

INTRODUCTION A VARIETY o f cell-encoded proteins are involved in the malignant transformation of human cells [1]. The nuclear phosphoprotein p53 may represent such a transformation-associated cell-encoded protein [2]. It is expressed in minute concentrations in normal cells and tissues where it is associated with normal cellular proliferation and in particular in the G0/G1 transition [3, 4]. Transfection experiments revealed that p53 can immortalize and, in cooperation with an activated ras oncogene, transform primary cells [57]. Moreover, p53 is known to increase the tumorigenicity of established cell lines [8, 9] and to increase the metastatic capacity of murine carcinoma cells [10]. Recently, the gene coding for p53 was found to be either absent or mutated in a significant percentage of colorectal tumors [11[. This observation has led to the suggestion that p53 might function as a recessive oncogene, at least in certain human neoplasia [12]. Other human tumors, like osteosarcomas and blast cells of patients with CML in blast crisis, were characterized by rearrangements of the p53 gene [13, 14]. Abbreviations: A L L , acute lymphoblastic leukemia; ANLL, acute nonlymphoblastic leukemia; CML, chronic myeloid leukemia; SDS, sodium dodecysulfate. Correspondence to: Mathias Montenarh, Department of Biochemistry, University of Ulm, P.O. Box 4066, D-7900 Ulm, F.R. Germany. 1041

1042

S. KRAISSet al.

Therefore, we study p53 features in two different leukemia cell lines, one of t h e m is of lymphoblastic type (X308) and the o t h e r of myeloid (X376) origin. W e f o u n d elevated levels of p53 in both cell lines. This was due to an increased rate of synthesis as well as to an e x t e n d e d stability of the p53 protein, p53 from both cell lines f o r m e d high-molecular weight oligomers. Using different p53 specific m o n o c l o n a l antibodies we could detect at least two different immunological subsets o f p53.

MATERIALS AND METHODS Cell lines X308 is a suspension cell line which was established from peripheral mononuclear cells of a patient suffering from a CALLA-negative, Epstein-Barr nuclear antigen-negative Burkitt type B-ALL prior to cytostatic treatment [25]. X376 is a suspension cell line derived from peripheral mononuclear cells of a FAB4 classified ANLL in third relapse. The clonal origin of the cell lines were demonstrated by limiting dilution techniques [26]. All cells were grown in Dulbecco modified Eagle medium (DMEM) supplemented with 5% fetal calf serum in a 5% CO2 atmosphere at 37°C. Lymphocytes were obtained from peripheral blood of healthy human donors by Ficoll gradient centrifugation. For stimulation, cells were incubated with 0.2% phytohaemagglutinin P (DIFCO Laboratories, Detroit, U.S.A.) for three days. Antibodies We used the monoclonal antibody PAb421 which is known to precipitate human p53 [27]. PAb421 was purified and concentrated from hybridoma supernatants by protein A Sepharose chromatography. PAb1620 is a p53 specific monoclonal antibody which was originally isolated by Ball et al. [28] and further characterized by Milner et al. [29]. This antibody was kindly provided by G. Brandner, Freiburg. PAb1801, purchased from Oncogene Science (Mineola, U.S.A.), is a p53 specific monoclonal antibody originally isolated by Banks et al. [30]. Saturating amounts of monoclonal antibodies were used to ensure a complete immunoprecipitation of p53. Radiolabeling and extraction of cells One x 10 7 cells were washed three times with methionine-free DMEM and then labeled with 370 x 104 Bq of [35S]methionine for 2 h. For the pulse-chase labeling experiments cells were labeled for 30 min, washed twice with DMEM containing 5% fetal calf serum and then cultured in DMEM with 5% fetal calf serum and nonlabeled methionine in excess for various periods (chase) or extracted immediately (pulse). For Western Blot analysis cells were used unlabeled. Before extraction, cells were washed with ice-cold phosphate buffered saline (PBS), pelleted at 400 g for 2 min and lysed with 0.4 ml extraction buffer (0.5% Nonidet P40, 100 mM Tris-hydrochloride, pH 9.0, 0.1 M NaCI) for 30 min on ice. To protect proteins from proteolysis 1% Trasylol (Bayer AG, Leverkusen) and phenylmethylsulfonyl fluoride (final concentration 0.25 mg ml-1) were added to the extraction buffer. After pelleting for 2 min at 800g, the supernatants were centrifuged for 30 min at 105 000g in a type 50 Beckman rotor at 4°C. An

aliquot of the supernatants was taken to measure the protein content using the method of Lowry et al. [31]. For the subsequent immunoprecipitation equal amounts of total protein were used. Sucrose density gradient analysis A total of 0.4 ml cell extract corresponding to 1 x 10 7 cells was layered on top of a 10 ml linear 5-20% sucrose density gradient (100mM Tris-hydrochloride, pH7.3, 0.1 M NaCI, 0.5% Nonidet P40). The gradient was centrifuged for 14 h at 36 000 rpm in a Beckman SW41 rotor at 4°C. Thyroglobulin (18S), catalase (10S) and immunoglobulin G (7S) (Serva, Heidelberg) were run as markers in a parallel gradient. Fractions (0.5 ml) were collected from the bottom of the gradient. lmmunoprecipitation and SDS polyacrylamide gel electrophoresis The cell extracts or fractions of the gradient were precleared overnight at 4°C with 100 ~tl of a 10% suspension of heat-inactivated, formaldehyde fixed Staphylococcus aureus as previously described [32]. S. aureus was pelleted and the supernatant was further incubated for at least 2 h at 4°C with monoclonal antibody PAb421. By addition of S. aureus the thus formed highly specific immunocomplexes were pelleted. For sequential immunoprecipitation monoclonal antibody PAb1620 was used first and the supernatant of the first immunoprecipitation was subsequently incubated with PAb421 or vice versa. Immunocomplexes were washed five times with 50 mM Tris-hydrochloride, pH 7.4, 0.15 M NaCI, 5 mM E D T A containing 1% Nonidet P40 and then finally with 50mM NH4HCO3. Immunoprecipitates were eluted with 200 ~tl of elution buffer (50 mM NH4HCO3, 1% sodium dodecylsulfate, 5% fl-mercaptoethanol) for 45 min at 4°C, lyophilized and dissolved in 20 ~tl of sample buffer (65 mM Tris-hydrochloride, pH 6.8, 5% /3-mercaptoethanol, 10% glycerol, 0.01% bromophenol blue) by boiling for 10 min. After centrifugation for lmin at 10000g in an Eppendorf centrifuge, the samples were loaded onto 10 cm discontinuous 10% SDSpolyacrylamide gels and run at room temperature at a constant current of 18mA as described previously [33]. After electrophoresis, gels with [35S]methionine-labeled samples were prepared for fluorography on Kodak X-AR films as described elsewhere [34]. To quantitate the amount of labeled protein at the different time points of the pulsechase experiments and to further calculate the half-life of p53, the protein bands on the exposed films were analyzed by densitometric tracing. The pulse labeling time point was time point zero. The half-life was taken to be the time when 50% of the highest amount of immunoprecipitated labeled p53 was left. Western Blot analysis Western blotting was performed essentially as described [35]. Proteins were electrophoretically transferred from the SDS-polyacrylamide gels to nitrocellulose, p53 containing bands were blotted with 1:500 diluted monoclonal antibody PAb421 followed by [125I]-labeled protein A (specific radioactivity 5 x 105 Bq/~tg). Dried nitrocellulose sheets were autoradiographed using Kodak X-AR films. RESULTS L e v e l o f p53 in leukemic cells In order to examine biochemical properties of p53

X376 X 3 0 8 L Na N a N

a

94_ 68_ -p53 4 3-

FIG. 1. Level of p53 in two different leukemic cell lines. Approximately 1 x 107 cells were extracted and cell extracts containing the same amount of total protein were incubated with the p53 specific monoclonal antibody PAb421. Immunocomplexes were analyzed on 10% SDSpolyacrylamide gels. For Western Blot analysis proteins were transferred to nitrocellulose and incubated with PAb421 followed by [12Sl]protein A. The nitrocellulose filter was subjected to autoradiography. The positions of the molecular weight markers phosphorylase a, 94 000 (94) bovine serum albumin, 68 000 (68) and ovalbumin 43 000 (43) are shown on the left of the figure. N, normal hamster control serum; a, PAb421; and L, normal lymphocytes.

1043

X376 X 3 0 8 nL NaN 94

aN

a

sL Na

m

68

m

-p53 43

i

FIG. 2. Metabolic labeling of p53. One x 107 cells were labeled with [35S]methionine for 2 h. Cell extracts with equal amounts of total protein were immunoprecipitated first with normal hamster control serum (N) followed by PAb421. Immunoprecipitates were analyzed on a 10c~ SDS-polyacrylamide gel. The positions of the molecular weight markers phosphorylase a, 94 000 (94), bovine serum albumin, 68 000 (68) and ovalbumin 43 000 (43) are shown on the left of the figure. N, normal hamster control serum', a, PAb421; nl, non-stimulated lymphocytes: and sl, stimulated lymphocytes.

1044

(A)

N n

m

m

43__

(B)

X376 Nbl~

a

X308 Nbl~

94_ 68_ - p53 eBIm

%

i

FIG. 4. Sequential immunoprecipitation of two immunological subclasses from X376 and X308 cells. One x 107 cells of each line were labeled with [35S]methionine for 2 h, extracted, and immunoprecipitated with normal hamster control serum (lane N) followed twice by monoclonal antibody PAb421 (a, a'). The supernatant of this first immunoprecipitation was incubated with PAb1620 (b) (A). Alternatively, the cell extract was first incubated twice with PAb1620 (b, b') and the supernatant subsequently with PAb421 (a) (B). Proteins were analyzed on a 10% SDS-polyacrylamide gel. The molecular weight markers are the same as those in Fig. 1. 1045

IAI

8S

1 1

1 5

10

946+8-p53

43-

(B)

18S 1

10S

5

10

7S 15

9468 _

~ -p53

43-

FIG. 5. Sucrose density gradient centrifugation of p53 from X376 (A) and X308 (B) cells. Two × 107 cells of each line were labeled with [35S]methionine for 2 h. Extracts were centrifuged through 5-20% sucrose density gradients, p53 in each fraction of the gradient was immunoprecipitated with monoclonal antibody PAb421. Proteins were analyzed on 10% SDSpolyacrylamide gels. Sedimentation was from right to left. Molecular weight markers for sucrose density gradients are thyroglobulin (18S), catalase (10S) and immunoglobulin G (7S). Molecular weight markers for SDS-polyacrylamide gel electrophoresis are the same are described in Fig. 1. 1046

Expression of p53 in human leukemic cell lines in two different leukemic cell lines we were first interested in analyzing the level of p53 in these two cell lines in comparison to normal lymphocytes. Therefore, X308, X376, and normal lymphocytes were lysed and extracted with the non-ionic detergent Nonidet P40. Equal amounts of total protein in the cell extract were used for the immunoprecipitation with the p53 specific monoclonal antibody PAb421 which is known to recognize human p53 [27]. Immunoprecipitates were analyzed on a 10% SDS-polyacrylamide gel followed by Western blotting with [125I]protein A. As shown in Fig. 1 we only found a p53 specific protein in cell extracts from X308 and X376 cells but not from normal lymphocytes. However, there were differences in the amount of p53 in both leukemic cell lines which were quantitated by densitometry of the autoradiogram. X308 cells were characterized by a level of p53 which is a factor of four higher than in X376 cells. In normal lymphocytes p53 was undetectable by Western Blot analysis, suggesting that the level of p53 was extremely low.

Metabolic labeling of p53 Different amounts of p53 in different cell lines can be the result of a different rate of synthesis of p53. Therefore, incorporation of radiolabeled methionine into newly synthesized p53 was measured by metabolic labeling of X308, X376, or normal lymphocytes with [35S]methionine. In addition we have stimulated normal lymphocytes with phytohemagglutinin prior to labeling with [35S]methionine. Cells were lysed, extracted, and the cell extracts, with equal amount of total protein, incubated with the p53 specific monoclonal antibody PAb421 as described above. Immunoprecipitates were analyzed on a 10% SDSpolyacrylamide gel. The fluorogram of this polyacrylamide gel is shown in Fig. 2. There was only a faint band for p53 in normal lymphocytes (nL) whereas in stimulated lymphocytes p53 was clearly detectable. In both leukemic cell lines considerably more p53 was labeled with [35S]methionine than in normal and stimulated lymphocytes. X376 cells seemed to express two p53 protein species which differ slightly in their mobility on the SDS polyacrylamide gel. p53 bands from X376 cells were clearly stronger than the protein band for p53 from X308 cells. Since we used the same amount of total protein for the immunoprecipitation of p53 these results demonstrated different rates of synthesis in X376 compared to X308 cells and in comparison to unstimulated and stimulated lymphocytes. Stability of p53 in leukemic cells From the results shown in Fig, 2 one would expect

1047

a high level of p53 in X376 and a lower level in X308. However, as shown in Fig. 1 the contrary was found. This discrepancy can be explained if there were major differences in the stability of p53 in these two cell lines. Therefore, we analyzed the stability of p53 in X308 and X376 cells by pulse-chase labeling. Cells were labeled for 30 min with [35S]methionine and either lysed immediately or chased for 2, 6, and 14 h with medium containing unlabeled methionine. Cell lysis, immunoprecipitation, and SDS-polyacrylamide gel electrophoresis were as described above, p53 protein bands were scanned by laser densitometry. Figure 3 shows a diagram with the amount of radioactivity found for the p53 band at the various times. p53 from X376 cells was shown to possess a half-life of slightly less than 2 h whereas p53 from X308 cells turned out to be extremely stable with a half-life of more than 14h. We found no differences in the stability for the two different p53 protein species expressed in X376 cells. In the case of X308 cells we observed a reproducible increase in the radioactivity in the p53 band within the first 2 h of the chase. This could be due to an extremely slow incorporation of [35S]methionine into newly synthesized proteins. Alternatively, newly synthesized p53 from X308 cells might change its conformation during maturation and only the mature form would be recognized quantitatively by the particular monoclonal antibody used for immunoprecipitation of p53.

Immunological subclasses of p53 in leukemic cells In order to analyze the two different leukemic cell lines for the presence of immunological subclasses of p53 we labeled X308 cells and X376 cells with [35S]methionine for 2 h. After lysis of the cells p53 was first immunoprecipitated from the cell extract using monoclonal antibody PAb421. PAb421 is a monoclonal antibody which precipitates rodent and primate p53. The epitope recognized by PAb421 is located at the C-terminus of p53 between amino acids 370 and 378 [36]. The supernatant was repeatedly incubated with PAb421 to ensure quantitative precipitation and finally with monoclonal antibody PAb1620. PAb1620 is a p53 specific monoclonal antibody which was isolated by Ball et al. [28] and further characterized by Milner et al. [29, 37]. This monoclonal antibody is known to recognize murine and human p53 and crossblock the mouse specific monoclonal antibody PAb246. The epitope recognized by PAb1620 is conformation dependent similar to the epitope recognized by PAb246 [37]. Immunoprecipitates were analyzed on a 10% SDSpolyacrylamide gel. The fluorogram of this gel is shown in Fig. 4. p53 from both cell lines was quantitatively precipitated with PAb421. No p53 band was

1048

S. KRAISS et al.

% 100.

908070-

60-

)

50J

403020 10 0 2

!

!

4

6

~

!

!

v

10

12

14

h Chaee

FIG. 3. Metabolic stability of p53 from X308 and X376 cells. For each single time point 1 × 107 cells were labeled with [35S]methionine for 30 min and chased for 2, 6, or 14 h. Cell extracts with equal amount of total protein were immunoprecipitated with PAb421. After analysis of the immunoprecipitated p53 on a 10% SDS-polyacrylamide gel and fluorography, the relative amounts of radiolabeled p53 were determined by densitometry. The graphic presentation of the stability shows a half-lifefor p53 from X308 (0) of more than 14 h and for X376 (A) of about 2 h.

found in the final immunoreaction with PAb1620 (Fig. 4A). The same experiment was repeated but instead PAb1620 was used first. After exhaustive incubation with PAb1620, PAb421 was employed in the final step. PAb1620 only precipitated a subset of p53 in both cell lines. The remaining p53 was recognized by monoclonal antibody PAb421. Again, we observed two bands for p53 from X376 which were precipitated similarly by both monoclonal antibodies. Thus, our results demonstrated the presence of at least two immunological subsets of p53 in both leukemic cell lines. Since we used PAb421 for the experiments described in Fig. 3 the increase in radioactivity during the first 2 h of the chase cannot be explained by the appearance of a new immunological subclass of p53 but more likely by slow uptake of radiolabeled methionine.

The cell extract was analyzed on a 5-20% sucrose density gradient. After centrifugation the gradient was fractionated and each fraction incubated with PAb421 to precipitate p53. Immunoprecipitates were analyzed on 10% SDS-polyacrylamide gels. The fluorograms of these gels are shown in Fig. 5A and B. p53 from both cell lines sedimented in a broad distribution over the entire gradient with major peaks between 7-10S and 14-20S. These data demonstrate that p53 appears in various forms in leukemic cells which differ in their quaternary structure. Again, we found two p53 protein species in X376 cells which showed the same sedimentation profile on the sucrose density gradient (Fig. 5A).

Quaternary structure of p53 in leukemic cell lines Sequential immunoprecipitation of p53 with two different monoclonal antibodies revealed the presence of at least two different subsets of p53. In order to pursue this idea and to analyze p53 for other structural subsets we determined the quaternary structure of p53 in these two cell lines. X376 and X308 cells were labeled with [35S]methionine and extracted with the non-ionic detergent Nonidet P40.

Activation of nuclear oncogenes is in many cases associated with the deregulated expression of the oncoproteins. This deregulation often results in an elevated level or in altered biochemical properties of the gene products, p53 is a cellular protein which seems to be associated with transformation in a variety of systems. Elevated levels of this protein were detected in many types of tumors and transformed cells of both rodent and primate origin

DISCUSSION

Expression of p53 in human leukemiccell lines [16, 17, 38, 39]. Interestingly, not only overexpression but, in some tumors, also the complete absence of p53 seems to correlate with cell transformation or tumor progression. This contradiction is explained by recent findings that only mutant p53 is able to promote cell transformation whereas wild-type p53 has a negative regulatory function for cell growth [40, 41]. p53 in normal cells is metabolically labile and levels are low. Many transformed cells have stable forms of p53 which accumulate to high steadystate levels [39]. So far most studies on the expression of p53 were performed with transformed cells of mouse or rat origin. Therefore, we were interested to analyze p53 features in cell lines derived from human tumors. We have chosen two leukemic cell lines since human leukemia and lymphomas are reported to express high levels of p53 [17, 18, 42]. By Western Blot analysis we found p53 levels in both tumor derived cell lines, X308 and X376, well above the levels found in normal human lymphocytes. By metabolic labeling p53 could be detected in low amounts in non-dividing lymphocytes and in elevated quantities in phytohemagglutinin stimulated lymphocytes which is in agreement with earlier observations with normal mouse lymphocytes [43]. Elevated levels of p53 were already demonstrated in fresh tumor material from patients with lymphoma or ALL or in cell lines of N-ALL, T-ALL or Burkitt lymphoma origin whereas low or even undetectable amounts of p53 were reported for tumor material from patients with CLL or AML or the corresponding cell lines [42]. The rate of synthesis for p53 was slightly lower in lymphocytes than in both leukemic cell lines. Since p53 was not detected by the sensitive Western blotting p53 seems to be rather unstable in lymphocytes, whereas it turned out to be stable in both tumor derived cell lines. This property of human p53 resembles p53 in transformed mouse, rat and monkey cell lines [22, 23] where an elevated stability was found for p53 with a half-life ranging from 2 to 14 h and more, whereas in their normal parental cells half-lives of about 20 min were found. We consistently observed two p53 protein species in the ANLL cell line X376 but not in the ALL cell line X308. Similar observations were reported for NALL, pre B-ALL or Burkitt lymphoma cell lines or for an SV40 transformed human cell, SVS0 [42]. This heterogeneity of p53 might be due to alternative splicing events [44] or to a heterocygosity on chromosome 17 [11]. However, both p53 species showed the same protein stability and sedimentation profile on sucrose density gradients and both were precipitated with two different monoclonal antibodies. There seem to exist differences in the metabolic activity between X308 and X376 cells. Rate of synthesis is low for p53 in X308 cells compared to X376

1049

cells and this might be responsible for the increase in radioactivity in p53 within the first 2 h of the chase. The extremely high stability of p53 in X308 turned out to be responsible for the high level of p53 in this cell line. These data suggest that the level of p53 within a human tumor cell line can be regulated by a variable rate of synthesis as well as by the stabilization of the protein. So far nothing is known about mechanisms which are responsible for the regulation of the synthesis of p53 or its stabilization in tumor cells. Sucrose density gradient analysis revealed that p53 from both tumor derived human cell lines was found in low and various high-molecular weight forms. Since it has been shown that the degradation of p53 is an ATP energy dependent process [45] it seems to be plausible that the enzymes which cause degradation of p53 cannot degrade oligomeric forms of p53 as efficiently in the two cell lines. However, there seems to be no gross difference in the sedimentation properties between p53 from X308 and X376 cells which might have explained the differences in the stability of p53 from both cell lines. It was recently shown for p53 from mouse and rat cells that the formation of high-molecular weight oligomers of p53 correlates perfectly with the transformed phenotype of the cells [24, 46, Kraiss et al., in press]. The sucrose density gradient analysis revealed that p53 in these two tumor derived cell lines is represented by heterogenous subclasses of p53. Different subclasses of p53 were also demonstrated by the use of two different monoclonal antibodies. Sequential immunoprecipitation studies revealed that antibody PAb1620 precipitated only a subset of p53 whereas PAb421 and another monoclonal antibody, PAbl801 [30], precipitated p53 quantitatively (data not shown). Similar observations were reported for mouse p53 where a subset of p53 with a particular conformation was immunoprecipitated by the mouse p53 specific monoclonal antibody PAb246 [24, 47]. Preliminary experiments provided evidence that in mouse cells PAb1620 may recognize the same subset of p53 as PAb246 (data not shown) which would support recent results [29, 37]. Biochemical properties as well as probably different biological functions of these two immunologically defined subclasses of p53 remain to be elucidated. Our results have demonstrated a structural variability of p53 in cell lines derived from human tumors. This structural variability may be a key to the understanding of the role p53 seems to play in human malignancies. Acknowledgements--The authors want to thank Karin Dengler-Wupperfeld and Claudia Imhof for expert technical assistance, Alison Gatrill for editorial help. Mono-

1050

S. KRAISSet al.

clonal antibody PAb1620 was kindly donated by G. Brandner, Freiburg, F.R.G. This work was supported by grants Mo309/2-1 and SFB 322 A1 from the Deutsche Forschungsgemeinschaft to M.M. and from the Wilhelm Sander-Stiftung (88.023) to W.H.

REFERENCES 1. Slamon D. J., deKernion J. B., Verma I. M. & Cline M. J. (1984) Expression of cellular oncogenes in human malignancies. Science 224, 256. 2. Lane D. P. & Benchimol S. (1990) p53: Oncogene or anti-oncogene. Genes Dev. 4, 1. 3. Mercer W. E., Nelson D., DeLeo A, B., Old L. J. & Baserga R. (1982) Microinjection of monoclonal antibody to protein p53 inhibits serum-induced DNA synthesis in 3T3 cells. Proc. Nat. Acad. Sci. U.S.A. 79, 6309 4. Shohat O., Greenberg M., Reisman D., Oren M. & Rotter V. (1987) Inhibition of cell growth mediated by plasmids encoding p53 antisense. Oncogene 1, 277. 5. Eliyahu D., Raz A., Gruss P. & Oren M. (1984) Participation of p53 cellular tumor antigen in transformation of normal embryonic cells. Nature 312, 646. 6. Jenkins J. R., Rudge K. & Currie G. A. (1984) Cellular immortalization by a cDNA clone encoding the transformation-associated phosphoprotein p53. Nature 312, 65l. 7. Parada L. F., Land H., Weinberg R. A. & Rotter V. (1984) Cooperation between the gene encoding p53 tumor antigen and ras in cellular transformation. Nature 312, 649. 8. Kelekar A. & Cole M. D. (1986) Tumorigenicity of fibroblast lines expressing the adenovirus Ela, cellular p53, or normal c-myc genes. Mol. Cell. Biol. 6, 7, 9. Tuck S. P. & Crawford L. (1989) Overexpression of normal human p53 in established fibroblasts leads to their tumorigenic conversion. Oncogene Res. 4, 81. 10. Pohl J., Goldfinger N., Radler-Phol A., Rotter V. & Schirrmacher V. (1988) p53 increases experimental metastatic capacity of murine carcinoma cells. Mol. Cell. Biol. 8, 2078. 11. Baker S. J., Fearon E. R., Nigro 1. M., Hamilton S. R., Preisinger A. C., Jessup J. M., van Tuinen P., Ledbetter D. H., Barker D. F., Nakamura Y., White R. & Vogelstein B. (1989) Chromosome 17 deletions and p53 gene mutations in colorectal carcinomas. Science 244, 217. 12. Nigro J. M., Baker S. J., Preisinger A. C., Jessup J. M., Hostetter R., Cleary K., Bigner S. H., Davidson N., Baylin S., Devilee P., Glover T., Collins F. S., Weston A., Modali R., Harris C. C. & Vogelstein B. (1989) Mutations in the p53 gene occur in diverse human tumour types. Nature 343, 705. 13. Ahuja H., Bar-Eli M., Advani S. H., Benchimol S. & Cline M. J, (1989) Alterations in the p53 gene and the clonal evolution of the blast crisis of chronic myelocytic leukemia. Proc. Nat. Acad. Sci. U.S.A. 86, 6783. 14. Masuda H., Miller C.~ Koeffler H. P., Battiford H. & Cline M. J. (1987) Rearrangement of the p53 gene in human osteogenic sarcomas. Proc. Nat. Acad. Sci. U.S.A. 84, 1116. 15. Cattoretti G., Andreola S., Clemente C., D'Amato L.

& Rilke F. (1988) Vimentin and p53 expression on epidermal growth factor receptor-positive, oestrogen receptor-negative breast carcinomas. Br. J. Cancer 57, 353. 16. Crawford L. V., Pim D. C. & Lamb P. (1984) The cellular protein p53 in human tumors. Mol. Biol. Med. 2, 261. 17. Koeffler H. P., Miller C., Nicolson M. A., Ranyard J. & Bosselman R. A. (1986) Increased expression of p53 protein in human leukemia cells. Proc. Nat. Acad. Sci. U.S.A. 83, 4035. 18. Smith L. J., McGulloch E. A. & Benchimol S. (1986) Expression of the p53 oncogene in acute myeloblastic leukemia. J. Exp. Med. 164, 751. 19. Lane D. P. & Crawford L. (1979) T antigen is bound to a host protein in SV40-transformed cells. Nature 278, 261. 20. Sarnow P., Ho Y. S., Williams J. & Levine A. J. (1982) Adenovirus Elb-58kd tumor antigen and SV40 large tumor antigen are physically associated with the same 54kd cellular protein in transformed cells. Cell 28~ 387. 21. Thomas R., Kaplan L., Reich N., Lane D. P. & Levine A. J. (1983) Characterization of human p53 antigens employing primate specific monoclonal antibodies. Virology 131,502. 22. Halevy O., Hall A. & Oren M. (1989) Stabilization of the p53 transformation-related protein in mouse fibrosarcoma cell lines: effects of protein sequence and intracellular environment. Mol. Cell. Biol. 9, 3385. 23. Reihsaus E., Kohler M., Kraiss S., Oren M. & Montenarh M. (1990) Regulation of the level of the oncoprotein p53 in non-transformed and transformed cells. Oncogene 5, 137. 24. Kraiss S., Ouaiser A., Oren M. & Montenarh M. (1988) Oligomerization of oncoprotein p53. J. Virol. 62, 4737. 25. Hartmann W., Hitzler H., Schlickenrieder J. H. M., Zapf J., Heit W., Gaedicke G. & Vetter U. (1988) Heterogeneity of insulin and insulin-like growth factor I binding in a human Burkitt type ALL cell line during the cell cycle and in three Burkitt type ALL sublines. Leukemia 2, 241. 26. Vetter U., Schlickenrieder J. H. M., Zapf J., Hartmann W., Heit W., Hitzler H., Byrne P., Gaedicke G., Heinze E. & Teller W. M. (1986) Human leukemic cells: Receptor binding and biological effects of insulin and insulin-like growth factors. Leukemia Res. 10, 1201. 27. Harlow E., Crawford L. V., Pim D. C. & Williamson N. M. (1981) Monoclonal antibodies specific for the SV40 tumor antigens. J. Virol. 39, 861. 28. Ball R. K., Siegl B., Quellhorst S., Brandner G. & Braun D. G. (1984) Monoclonal antibodies against simian virus 40 nuclear large tumor antigen: epitope mapping, papova virus cross-reaction and cell surface staining. EMBO J. 3, 1485. 29. Milner J. A., Cook A. & Sheldon M. (1987) A new anti-p53 monoclonal antibody, previously reported to be directed against the large T antigen of simian virus 40. Oncogene 1,453. 30. Banks L., Matlashewski G. & Crawford L. V. (1986) Isolation of human p53 specific monoclonal antibodies and their use in the studies of human p53 expression. Eur. J. Biochem. 159, 529. 31. Lowry O. H., Rosebrough N. J., Farr A. L. & Randall R. J. (1951) Protein measurement with the folin phenol reagent. J. Biol. Chem. 193, 265.

Expression of p53 in human leukemic cell lines 32. Kessler S. W. (1975) Rapid isolation of antigens from cells with a staphylococcal protein A-antibody adsorbent: Parameters of the interaction of antibody-antigen complexes with protein A. J, Immunol. 115, 1617. 33. Montenarh M. & Henning R. (1983) Self-assembly of simian virus 40 large T antigen oligomers by divalent cations. J. Virol. 45, 531. 34. Bonnet W. M. & Laskey R. A. (1974) A film detection method for tritium-labeled proteins and nucleic acids in polyacrylamide gels. Eur. J. Biochem. 46, 83. 35. Burnette W. N. (1981) Electrophoretic transfer of proteins from sodium dodecyl sulfate-polyacrylamide gels to unmodified nitrocellulose and radiographic detection with antibody and radioiodinated protein A. Anal. Biochem. 112, 195. 36. Yewdell J. W., Gannon J. B. & Lane D. P. (1986) Monoclonal antibody analysis of p53 expression in normal and transformed cells. J. Virol. 59, 444. 37. Cook A. & Milner J. (1990) Evidence for allosteric variants of wild-type p53, a tumour suppressor protein. Br. J. Cancer 61, 548. 38. Crawford L. V., Pim D. C., Gurney E. G., Goodfellow P. & Taylor-Papadimitriou J. (1981) Detection of a common feature in several human tumour cell lines-a 53000 dalton protein. Proc. Nat. Acad. Sci. U.S.A. 78, 41. 39. Oren M. (1985) The cellular tumor antigen: gene structure, expression and protein properties. Biochim. Biophys. Acta 823, 67.

1051

40. Eliyahu D., Michalovitz D., Eliyahu S., Pinhasi-Kimhi O. & Oren M. (1989) Wild-type p53 can inhibit oncogene-mediated focus formation. Proc. Nat. Acad. Sci. U,S.A. 86, 8763. 41. Finlay C. A., Hinds P. W. & Levine A. J. (1989) The p53 proto-oncogene can act as a suppressor of transformation. Cell 57, 1083. 42. Prokocimer M., Shaklai M., Bassat B., Wolf D., Goldfinger N. & Rotter V. (1986) Expression of p53 in human leukemia and lymphoma. Blood 68, 113. 43. Milner J. (1984) Different forms of p53 detected by monoclonal antibodies in non-dividing and dividing lymphocytes. Nature 310, 143. 44. Matlashewski G., Pim D., Banks L. & Crawford L. (1987) Alternative splicing of human p53 transcripts, Oncogene Res. 1, 77. 45. Gronostajski R. M., Goldberg A. L. & Pardee A. B. (1984) Energy requirement for degradation of tumorassociated protein p53. Mol. Cell. Biol. 4, 442. 46. Shiroki K., Segawa K., Koita T. & Shibuya M. (1986) Neoplastic transformation of rat 3Y1 cells by a transcriptionally activated human c-myc gene and stabilization of p53 cellular tumor antigen in the transformed cells. Mol. Cell. Biol. 6, 4379. 47. Sti~rzbecher H.-W., Chumakow P., Welch W. J. & Jenkins J. R. (1987) Mutant p53 proteins bind hsp 72/ 73 cellular heat shock-related proteins in SV40 transformed monkey cells. Oncogene 1,201.