Protein tyrosine phosphatase activity as a diagnostic parameter in breast cancer

Breast CancerResearchand Treatment33: 245-256, 1995. © 1995KluwerAcademic Publishers.Printedin the Netherlands'. Report

Protein tyrosine phosphatase activity as a diagnostic parameter in breast cancer

Astrid E. Ottenhoff-Kalff1, Brigitte A. van Oirschot 1, Adriaan Hennipman 3, Roel A. de Weger a, Gerard E.J. Staal 1and Gert Rijksen 1 1Department of Hematology, Laboratory of Medical Enzymology; 2 Department of Pathology; and 3 Department of Surgery, University Hospital Utrecht, PO. Box 85500, 3508 GA Utrecht, The Netherlands

Key words: protein tyrosine phosphatase, phosphotyrosine, breast cancer, histochemistry

Summary Cellular phosphotyrosine levels are regulated by the balance between protein tyrosine kinases (PTKs) and protein tyrosine phosphatases (PTPs). It is supposed that this balance is disturbed in tumour cells, making the increased or altered activity of PTKs and PTPs likely hallmarks of tumour tissues. Indeed it could be shown that the PTK activity was increased in breast cancer in correlation with prognosis (Hennipman et al., Cancer Res. 49, 516-522, 1989). In the present report we measured the PTP activities in breast cancer and normal breast tissues. An increase of approximately three- to four-fold was measured in the cytosolic tumour fractions compared to normal, whereas the solubilized membrane fraction PTP activity showed an increase in tumours of approximately 1.5-fold. Remarkably, the membrane PTP activity correlated with the presence of tumour positive axillary lymph nodes (p = 0.004), whereas the cytosolic PTP activity correlated with the mitotic index, a higher PTP activity occurring when the mitotic index was higher than 10 (p -- 0.0004). These results indicate that membrane PTP activity may be considered as an index of metastatic potential, whereas cytosolic PTP activity may be a measure of the growth capacity of the tumour. The increase of PTP activity in breast cancers was confirmed by enzyme-histochemical studies. In frozen sections of tumours a strong to moderate activity was found in both tumour cells and interstitial cells. In the interstitium membrane activity was most pronounced, whereas in the tumour cells diffuse staining of the cytoplasm together with a clear membrane staining was demonstrated. Immunoblotting with anti-phosphotyrosine antibodies also reveals differences between the tumours and normal tissues, confirming the disturbance of the balance between protein tyrosyl phosphorylation and dephosphorylation in the tumour cells.

Introduction The phosphorylation of proteins on tyrosine residues by protein tyrosine kinases is an important event in the control of cellular proliferation and differentiation. To attenuate the signals produced by the protein tyrosine kinases (PTKs), the presence

of counter-acting protein tyrosine phosphatases (PTPs) seems mandatory. Indeed, cellular phosphotyrosine levels are regulated by the activities of PTKs and PTPs, and the balance between PTKs and PTPs is important in maintaining the controlled growth of cells. On the other hand, PTPs may function not only by counteracting the effects of PTKs,

Addressfor offprints: G. Rijksen,Departmentof Hematology,Laboratoryof MedicalEnzymology,RO. Box 85500, 3508 GA Utrecht, The Netherlands

246

A E Ottenhoff-Kalff et al.

but also by enhancing tyrosine phosphorylation through activation of non-receptor tyrosine kinases, e.g. the Src-family tyrosine kinases [1-4]. Whereas protein tyrosine kinases have been extensively studied and constitute a well characterized family of proteins [5, 6], much less is known about the function of protein tyrosine phosphatases. PTPs have been recognized as a distinct family of enzymes with a high specificity for phosphotyrosine-containing substrates [7-11]. Molecular analysis of the PTPs resulted in the identification of a common homology domain (for reviews see [12,13]), containing a highly conserved Cys-residue which is absolutely required for PTP activity [7,11,14,15]. Both cytoplasmically localized (non-receptor) as well as transmembrane (receptor-like) PTPs, generally comprising one and two PTP enzymatic domains, respectively, have been identified [16-18]. In addition to this, phosphatases that contain homologous sequence to the SH2 domain of protein tyrosine kinases have been cloned [19, 20], which suggests transmodulation and possible interaction between PTKs and PTPs. In the past, we have shown that the protein tyrosine kinase activity of human breast cancer tissues was significantly increased compared to that in normal breast tissues [21, 22], and that this correlated with disease-free survival [21]. This might be expected considering the fact that many oncogenes encode PTKs, and that increased or altered oncogene expression may be reflected in an increased tyrosine kinase activity. Recently, we have identified the majority of the enhanced protein tyrosine kinase activity in the cytosolic fraction of the breast tumours as the protooncogene product c-Src [22]. In the present study we measured the protein phosphatase and tyrosine kinase activities in a series of breast cancer tissues and compared these with the activity found in normal breast tissues. We show that the PTP activity in breast tumour tissues is increased compared to normal tissue, both by direct enzymatic assay on homogenates as well as by enzyme-histochemistry. Also, once again, the PTK activity was found to be enhanced in the breast cancer tissues, compared to normal. In addition, immunoblots with anti-phosphotyrosine antibodies were performed to assess the bal-

ance of the endogenous PTK and PTP activity present in the samples. We show that the amount of tyrosyl phosphorylation in the tumours is extremely increased compared to that in the normal tissues.

Materials and methods Patients and pathology

A consecutive series of 54 patients was collected during a short period in 1991. Patients were grouped according to histology. The specimens of normal breast tissue were from patients with juvenile hypertrophy, obtained when a reduction mammoplasty was performed. These specimens did not show any abnormality on histological examination and were therefore accepted to represent normal breast tissue. Specimens of malignant tumours of the breast were obtained after surgery, cut into small pieces, and immediately frozen in liquid nitrogen. All specimens were stored at - 70 ° C until further use. Histological diagnosis was made according to WHO criteria. The histologies were distributed as follows: 6 patients had infiltrating lobular carcinomas, 41 patients had ductal carcinomas not otherwise specified, 1 had medullary carcinoma, 2 had colloid carcinomas, and 4 patients had a combination of lobular and ductal histologies. The mitotic index is defined as the amount of mitoses per high power field (magnification 400 x). Stage grouping of the patients was done according to the criteria of the UICC-AJC 1977 and is shown in Table 1. At analysis the median follow-up had arrived at 17 months. Table 1, Stagegroupingof the patientsin the study

Stage groupingaccordingto UICC-AJC1977 n= relapses Stage i Stage 2 Stage 3 Stage 4

16 26 9 3

1 4* 2

Total

54

7

* One patient has died of disease.

Tyrosine phosphatase activity in breast cancer Three patients had disseminated disease, 2 were biopsed only, I was treated surgically with palliative intent. These 3 patients were excluded from calculations. The characteristics of this small patient population were distributed comparable to larger published series; 18 patients were pre-menopausal, 35 patients were post-menopausal, and of 1 patient the menopausal status was not known. Of the patients (stage 4 excluded) 23 had tumour involvement of the axillary lymph nodes. Four patients had bilateral tumours, either synchronous or metachronous. None of these patients had relapsed at the time of analysis; therefore, these 4 patients have not been excluded. Five of the 51 patients in Stage 1-3 have relapsed, 1 has died of disease. Patients were stratified on generally accepted prognostic parameters.

Materials" For the preparation of PTP substrate, tonsillar tissue was used. Tonsils were obtained from children when a tonsillectomy was performed, All chemicals used in buffers, poly (glutamic acid:tyrosine, 4:1), and levamisol were from Sigma (St. Louis, MO). BSA (bovine serum albumin) used in the immunoblotting procedure was essentially fatty acid free (Sigma). [33p]-ATP was from New England Nuclear (Berkely, CA), sodium tartrate was from Riedel de Hahn (Seelze, Germany), antiphosphotyrosine PY-20 was from ICN (Cleveland, OH), gold-labeled goat-anti-mouse IgG + IgM was from Amersham (Buckinghamshire, England) and polyvinylidene difluoride membranes were from Millipore (Bedford, MA).

Sample preparation Cytosolic and solubilized membrane fractions were prepared as described in [22]. The protein content was determined according to the method of Bradford [23]. To exclude the presence of membrane contaminations in our cytosolic fractions, both fractions were tested for the activity of the plasma membrane

247

marker 5' nucleotidase. Assays were conducted according to Hubbard et al. [24]. Contamination of the membrane fraction with cytosol was excluded by measuring lactate dehydrogenase (LDH) activity according to Beutler [25] in the solubilized membrane fractions.

Protein tyrosine kinase activity The protein tyrosine kinase activity was determined as described before, using a non-radioactive dot blot assay described in [26].

Preparation of tyrosine-phosphorylated PTP substrate To prepare substrate for the assay of protein tyrosine phosphatase (PTP) activity, 0.5 mg cytosolic and 0.5 mg solubilized membrane fraction from tonsil were combined to obtain a solution of 1 mg proteins per ml incubation buffer, which contained the following: 50 mM Hepes, pH 7.2; 10 mM MgC12; 2 mM MnC12; 0.2 mM EDTA-Na2; 0.8 mM EGTA; i mM dithiothreitol; 5 mM NaF; and 40 gM sodium-o-vanadate. One ml of this preparation was incubated overnight at room temperature with 5 mg of PGT (poly(glutamic acid:tyrosine, 4:1)) and 100 gM of [33p]-ATP with a specific activity of 0.15 Ci/mmol in a total volume of 7.0 ml. The phosphorylated PGT was precipitated for 1-2 hours with 10% TCA (trichloro acetic acid) and washed extensively with 10% TCA. The pellet was resuspended in 1 M NaOH. A buffer change was accomplished by passing the phosphorylated PGT over a Sephadex G-25 PD10 column (Pharmacia, Sweden) equilibrated in 50 mM Hepes, pH 7.0, containing 0.1 mM EDTA and i mM dithiothreitol (PTP buffer). The substrate fraction was eluted in 2 ml. The phosphorylated PGT was divided into small aliquots and stored at - 20 ° C until further use. Prior to the PTP assay, the PGT was diluted with PTP buffer and PGT to 0.25 mg/ml PGT and an incorporation of 5 mmol P-Tyr/mol tyrosine. To quantify the amount of incorporated phosphate, aliquots of the mixture were spotted on

248

A E Ottenhoff-Kalff et al.

3 MM paper (Whatman), precipitated prior to dilution to the proper concentration, and washed extensively with 10% TCA and 100 mM PPi. Papers were washed in ethanol/ether (1/1) and ether, and airdried. Incorporation of label was quantified by liquid scintillation counting. The total incorporation of 33p was calculated by subtracting the background phosphorylation of endogenous proteins, measured as described above in the absence of PGT.

Assay of PTP activity Cytosolic fractions were diluted to 0.1 mg sample protein/ml with PTP buffer containing 0.5 mg BSA, and solubilized membrane fractions were diluted to 0.1 mg/ml with PTP buffer containing 0.5 mg BSA and 0.5% Nonidet P-40. Dephosphorylation of [33p]-PGT was carried out at 37 ° C in a total volume of 40 gl containing 50 mM Hepes, pH 7.0, 0.1 mM EDTA, and 1 mM dithiothreitol. The reaction was started by adding cytosolic or solubilized membrane protein (2 gg/assay). After 10 minutes the reaction was stopped by the addition of 200 gl 12% TCA. Proteins and excess 33p-PGT were precipitated for 60 min at room temperature, and pelleted by centrifugation at 14,000 x g for 10 rain. Radioactivity remaining in the supernatant (soluble 33p-PO4) was measured by liquid scintillation counting. PTP activities were calculated after correction for blank values (incubation of substrate with buffer) and are given in pmol PO4/min/mg protein. To ensure the specificity of our protein tyrosine phosphatase activity, we tested several inhibitors of type II serine/threonine, acid- or alkali-specific phosphatases. Two inhibitors of type II serine/threonine phosphatases, NaF and EDTA, added to the reaction mixture in a concentration range of respectively 020 and 0-5 mM, did not inhibit the PTP activity of our breast cancer cytosol or solubilized membrane fractions. There was also no inhibition of the PTP activity by both sodium tartrate (0-10 mM) and levamisol (0-2 raM), inhibitors of respectively acid and alkaline phosphatases. In contrast, the PTP inhibitor sodium orthovana-

date was able to inhibit both the cytosolic and solubilized membrane PTPs (see Results).

Western blotting To measure the balance between endogenous tyrosine kinase and phosphatase activities, tissues were homogenized as described in [22], in the following extraction buffer: 20 mM Hepes, pH 7.2; 0.2 mM EDTA-Na2; 0.8 mM EGTA; 1 mM dithiothreitol; 20 mM Mg-acetate; 5 mM NaF; 10% glycerol; 1 mM phenylmethylsulfonyl fluoride; and 0.055 TIU/ml (trypsin inhibitor unit/ml) aprotinin. When proteins from the membrane fraction were extracted, 0.5% Nonidet-P40 was added to this buffer. 100 gg cytosolic or membrane proteins were incubated for 60 min at 37 ° C in the presence of 500 gM ATP in a total volume of 60 gl. Phosphorylations were also carried out in the presence of 1 mM sodium-o-vanadate to inhibit the protein tyrosine phosphatases. The reaction was stopped by the addition of 20 gl sample buffer, which contained 60 mM Tris-HC1 pH 6.8, 50% glycerol, 2% SDS, and 0.005% bromophenol blue. Subsequently, the sample was boiled for 5 min and loaded onto a 10% SDS-polyacrylamide gel. After electrophoresis and transfer of the proteins to a PVDF (polyvinylidene difiuoride) membrane, the membrane was blocked for 60 min at 37 ° C with 5% BSA in PBS (phosphate-buffered saline). After washing three times 5 min in 0.1% BSA-PBS (PBSB; PBS containing 0.1% BSA), the membrane was incubated with 2 gg/ml PY20 antiphosphotyrosine antibodies in PBSB containing 1% normal goat serum for 2-16 hours. As a control for a specific binding of the antibody, a duplicate membrane was incubated with the same solution also containing 20 mM phenylphosphate, which had been preincubated for 2 hours. After washing the membrane three times 5 min with PBSB, a secondary antibody solution consisting of gold-labelled goat-anti-mouse IgG + IgM 1:100,1:20 v/v fish gelatin in PBSB was added to the membranes and incubated for at least two hours. Proteins were made visible by immuno-gold silver enhancement as described in [26].

Tyrosine phosphatase activity in breast cancer Histochernical protein tyrosine phosphatase assay T h e detection of the activity of P T P s by histochemical staining was p e r f o r m e d essentially as described by Kidd et al. [27]. Briefly, 6-8 g m tissue sections were air-dried, and incubated for 4 hours at 37 ° C in 400 gl of reaction solution. This reaction solution contained 10 m M P-Tyr (or P-Ser or P-Thr) and 1.75 m M lead nitrate in 50 m M M E S (2-[N-morpholino] ethanesulfonic acid) buffer, p H 6.0. In s o m e cases 10 m M s o d i u m - o - v a n a d a t e was a d d e d to inhibit P T P activity. T h e slides were then w a s h e d four times for 5 min each in deionized w a t e r and dev e l o p e d for I m i n in 0.2% a m m o n i u m sulfide solution (freshly m a d e f r o m 20% solution), followed by washing four times for 5 rain each in deionized water. T h e slides were counterstained with h a e m a t o x ylin ( M a y e r ) for 5 min, w a s h e d for 10 min in water, d e h y d r a t e d in alcohol and xylol, and e m b e d d e d in D e P e X (Kiln±path, Z e v e n a a r , T h e Netherlands).

Results

Determination of P T P activities Protein tyrosine p h o s p h a t a s e activities ( m e a n activities + S E M (standard error of the m e a n ) ) of b o t h cytosolic and solubilized m e m b r a n e fractions were d e t e r m i n e d (Table 2). Cytosolic P T P activity a p p e a r s to be increased three- to four-fold, while in the solubilized m e m b r a n e fractions P T P activity is increased only a p p r o x i m a t e l y 1.5-fold. We wond e r e d w h e t h e r there was any correlation of these activities with k n o w n prognostic p a r a m e t e r s . Strikingly, there a p p e a r e d to be a significant correlation b e t w e e n m e m b r a n e P T P activity and the p r e s e n c e of tumour-positive l y m p h nodes (Table 3), with higher P T P activities occurring w h e n there was l y m p h n o d e i n v o l v e m e n t (p = 0.004). In contrast, cytosolic P T P activity did not correlate with l y m p h n o d e involvement, but did correlate with the mitotic index; if the mitotic index was _ 10. In addition, there was a t e n d e n c y for correlation of the cytosolic P T P activity with the B l o o m and R i c h a r d s o n grade of malig-

249

Table2. Average PTP and PTK activities of normal and tumour tissues expressed in pmol P-Tyr/min/mg or POg37min/mg and given as mean + SEM

Normal tissue cytosol membranes Tumour tissue cytosol membranes

PTP ± SEM

PTK ± SEM

68 _+27 (n = 8) 182 _+53 (n = 8)

22 ± 12 (n = 7) N.D.

253 ± 25 (n = 43) 278 _+35 (n = 49)

196 ± 22 (n = 52) N.D.

nancy, with a higher P T P activity in patients with grade II or I I I t u m o u r s (p = 0.056). We also m e a s u r e d the P T K activities in the cytosolic fractions of breast cancer and n o r m a l tissues (Table 2) and c o n f i r m e d our earlier o b s e r v a t i o n [21] that this activity had a prognostic significance. For the patients who had relapsed during the follow-up period, a correlation b e t w e e n the cytosolic P T K activity and the disease-free survival existed; patients who had e x p e r i e n c e d a relapse had a significantly higher P T K activity in the cytosol than those patients who were still disease-free (344 _+ 63 versus 183 + 23; p = 0.02). Besides this relationship, a correlation also exists b e t w e e n l y m p h n o d e status and cytosolic P T K activity. T h e P T K activity was sign±f-

Table3. Correlation of PTP values with prognostic parameters. The mean + SEM of PTP in the cytosol and PTP membranebound were calculated for the various groups of patients Factor

(n)

PTP (cytosol)

PTP (membrane)

Disease-free Relapsed p Node positive Node negative p Mitotic index > 10 Mitotic index < 10 p Grade I Grade II/III p

(41") (4)

266 _+29 183 _+33 0.4 253 ± 38 253 _+35 0.9 398 _+49 161 _+29 0.0004 180 ± 40 274 _+48 0.056

286 _+37 183 + 125 0.12 387 _+56 87 _+34 0.004 361 ± 85 217 ± 47 0.2 251 _+57 290 ± 48 0.8

(25) (27) (12) (16) (12) (28)

* Different n-values of the various groups are caused by the inaccessability of certain prognostic parameters for some of the patients.

A E Ottenhoff-Kalff et aL

250 120-

7~ I00

so

2O

0

i o.1

i

ii,llll

i 1

i

ILILIH

I

I

orthovanadate

I ~l~lH

I

I

I ~ll~H

100

10

1000

(/~M)

Fig. 1. Inhibition of PTP activity by sodium orthovanadate. PTP activity was measured in cytosolic (*) and solubilized membrane (m) fractions of breast cancer tissues in the presence of increasing amounts of sodium orthovanadate. PTP activity is given as a percentage of the activity without orthovanadate.

icantly lower in patients who did not have lymph node involvement, than in those patients who were node positive (241 + 26 versus 158 + 32; p = 0.003). No correlation was found between estrogen and progesterone receptors and any of the measured enzyme activities. Also, the menopausal status of the patients did not correlate with the PTK or PTP activities measured. Apart from the correlations with the prognostic parameters, we also looked for a correlation between the PTK activity and the PTP activity, as it can be imagined that the PTK and PTP activities are increased cooperatively. However, no correlation between the cytosolic PTK and PTP values was found (correlation coefficient r = 0.243), nor was there a correlation between the cytosolic PTK and the PTP from the membrane fraction (correlation coefficient r = 0.138).

Inhibition of PTP activity; integrity of the cell fractions The PTP activity in our assay could not be inhibited by NaE EDTA, sodium tartrate, or levamisol, arguing that this activity is not even partly accountable to non-tyrosine specific phosphatases. However, in our assay, there is a clear inhibition of phosphotyrosine dephosphorylation by sodium orthovanadate (Fig. 1). Markedly, the inhibition of mammary tu-

mor cytosolic fractions differs from that of solubilized membrane fractions; the K i (50% inhibition) of the cytosolic PTP is 14 gM orthovanadate, versus 0.48 gM orthovanadate for the membrane PTR This phenomenon is an indication that the PTP's present in these two fractions are actually distinct enzymes. This also underlines the integrity of the membrane and cytosol preparations. The purity of these fractions was tested by measuring the 5' nucleotidase activity (a marker of the plasma membrane) and LDH activity (a cytosolic marker) in both fractions. Even using 135 times more cytosolic protein than membrane protein, no 5' nucleotidase activity could be detected in the cytosol, whereas in the membrane fractions an average activity of 1.73 gmol/min/mg protein was measured. The LDH activity in the solubilized membrane fractions was less than 3% of the total LDH activity. Thus, relative PTP activities in membrane versus cytosol can be considered true representatives of these cell fractions.

Immunoblotting with anti-phosphotyrosine The balance between phosphorylation and dephosphorylation in the cell is important to maintain controlled cell growth. To investigate whether this balance was disturbed in tumour cells versus normal cells, we performed immunoblot analysis with antiphosphotyrosine antibodies. During storage of the tissues, and the procedure of isolating cytosolic and membrane fractions, cellular ATP is depleted. Therefore, we incubated the cytosolic or membrane fractions with ATP for 1 hour, either in the presence or absence of 1 mM vanadate. The results are shown in Fig. 2. Four tumours and four normal tissues were compared. Even lanes represent incubations with vanadate. It is evident that the amount of phosphorylation in the tumour tissue in the presence of vanadate is extremely high compared to the normal tissue. In both tumour and normal tissues, phosphorylation in the membrane fractions seems to be slightly higher than in the cytosolic fractions. However, when the tumour fractions are incubated in the absence of vanadate, there is hardly any phosphorylation of proteins left in the tumour tissue. In

Tyrosine phosphatase activity in breast cancer

!00

T -+-100

T

N

+ 100

N + -+-

64

100

N

N +-+

100

T '

100

T + -

92

36

200

16 97

116 97

66

66

45

45

31

31

Tissue +

Io

vanadate

~g/lane

251

T

T

--+-e4

N

+ 64

N

N

+ -+-22

17

N +--+

12

11

T

T

-- + - - + 47

22

Tissue vanadate

pg/lane

Fig. 2. Western blots probed with anti-phosphotyrosine, a). Cytosolic fractions of tumour (T) or normal (N) tissues were incubated with ATP (500 gM) _+I mM vanadate for t h at 37 ° C. Atiquots were subjected to electrophoresis and immunoblotting with antibodies to phosphotyrosine, b) Solubilized membrane fractions of tumour (T) or normal (N) tissues. Samples were treated as described under 2a.

the normal tissue, however, there is not much difference between incubations with or without vanadate. The majority of the substrates which are phosphorylated in the normal tissue are of high molecular weight (around 100-200 kD), whereas the tumour substrates range from 13-200 kD.

Enzyrne-histochemical detection ofPTP activity To detect the PTPs in situ, we performed an enzyme-histochemical method described recently [27] to detect tyrosine phosphatase activity directly on tissue sections of tumours. Ten tumours and normal mammary tissue were used to localize the PTP activity. In normal mammary tissue a weak reactivity was observed in the epithelium and in the interstitium (Fig. 3a). In tumours a strong to moderate activity was found in interstitial cells and tumour cells. The enzyme-histochemical reaction correlated well with the biochemically detected PTP activity. In the interstitium the reaction gave a membrane staining. In the tumour cells the reaction was more diffuse in the cytoplasm but clear on the membranes (Fig. 3c, e). Incubation in the presence of vanadate reduced or inhibited the reaction strongly (Fig. 3f). Membrane reactivity was lost but in some cases a weak cytoplasmic reactivity re-

mained. Incubation without substrate P-tyrosine or with P-serine as substrate did not result in a visible reaction (Fig. 3b, d).

Discussion

Oncogenes activated by a variety of mechanisms frequently have been shown to encode growth factors, receptor and non-receptor tyrosine kinases, or other enzymes that participate in mitogenic signalling. Thus, the increased or altered expression of oncogenes in tumour cells would expect to give rise to a change in tyrosine kinase activity in these cells. Indeed, in previous reports [21, 22], we showed that in breast cancer tissues, compared to normal tissues, the PTK activity was elevated and even contained prognostic significance [21]. It has often been shown that the content of tyrosine-phosphorylated proteins in transformed or tumour cells is elevated compared to normal cells [28], pointing to a disturbance in tyrosine phosphorylation. It is conceivable however that, besides changes in the tyrosine kinase activity in tumour cells, tyrosine phosphatase activities are also affected. In the presenI study we show that PTP activity in breast cancer is increased in comparison with normal breast tissue. As far as we know this is the first

252

A E Ottenhoff-Kalff et al.

Tyrosine phosphatase activity in breast cancer study reporting an increase in PTP activity in breast cancer, measured in a direct enzyme assay. The PTP activity in the cytosolic fractions was increased three- to four-fold, whereas that in the solubilized membrane fraction was increased only approximately 1.5-fold. Also, there were several correlations with known prognostic parameters. The membrane-bound PTP activity correlated with the presence of axillary lymph nodes; when no lymph nodes were involved, the PTP activities were significantly lower. Furthermore, the cytosolic PTP activity correlated very well with the mitotic index measured: a mitotic index higher than 10 correlated with a high PTP activity. The fact that the PTP activities in cytosol and membrane fractions correlate with different parameters is intriguing. We have to realize that membrane and cytosolic phosphatases are different enzymes and may have different functions. The presence of tumour positive lymph nodes points to metastasis, while a high mitotic index is indicative of a high rate of proliferation of the tumour cells, two events that need not necessarily be correlated. It is tempting to speculate that the PTP activity in the membrane fraction is more involved with the process of metastasis. One can imagine that this may alter cell-to-cell contacts or adhesion of the cells. In contrast, the activity of cytosolic PTPs may be more concerned with the generation of signals leading to the cell nucleus, and cell division. The past few years, much work has been focused on the tyrosine phosphatases, and many novel PTPs have been discovered (reviewed in [12, 13]). However, not much is yet known about the specific function of the different phosphatases. Despite the correlation between the PTP activity and the parameters mentioned above, there was no correlation of PTP activities with disease-free survival. The mean follow-up of the patients in our study was 17 months, whereas usually a median follow-up duration of 4-5 years is required for more definitive conclusions [29]. It will be of interest to reevaluate the relationship of sur-

253

vival and the PTP and PTK values with longer follow-up. With respect to the data on PTK activity, we could confirm our previous observations [21, 22], that the cytosolic PTK activity in breast tumours is increased compared to normal breast tissues. In keeping with our earlier report [21], the cytosolic PTK activity correlated with disease-free survival; patients suffering a relapse had a significantly higher PTK activity, showing that the cytosolic PTK activity has a prognostic significance. Also, the cytosolic PTK activity correlated with the involvement of axillary lymph nodes, with significantly higher PTK activities when lymph nodes were involved. From the immunoblot analyses with anti-phosphotyrosine a number of conclusions could be drawn. First, an obvious difference between the tumour and normal tissues was that in the presence of vanadate, there was much more tyrosine phosphorylation in the tumours. This probably reflects the higher PTK activity measured in tumour cells compared to normal cells. Secondly, whereas there was hardly any difference in phosphorylation of proteins in the normal tissue with or without vanadate, phosphorylation of proteins was largely absent in the tumour tissue if vanadate was omitted from the incubations. This implies that in normal cells, at least in vitro, there is a balance between tyrosine kinases and phosphatases such that the phosphatases are less important for tyrosine phosphorylation. However, in the tumour cells, it appears that the activity of the PTPs is strongly prevalent over that of the tyrosine kinases at least in vitro. This means that in the physiological situation, there must be a very tight regulation of the PTPs, leading to dephosphorylation of phosphorylated tyrosine residues only when the PTPs are specifically 'turned on'. This points to a much more sophisticated regulatory role for the protein tyrosine phosphatases than simply that of housekeeping enzymes which happen to dephosphorylate proteins counterbalancing tyrosine

<

Fig. 3. Enzyme histochemical detection of protein tyrosine phosphatases. Normal breast tissue (a, b) or malignant mammary tissue (c-f) were subjected to an enzyme-histochemical assay to detect protein tyrosine phosphatase activity, using phosphotyrosine (a, c, e), or phosphoserine (b, d) as a substrate. The reaction with phosphotyrosine was inhibited by incubation of the substrate in the presence of 10 mM Na3VO 4 (f). Magnification: a, b: 400 x; c, d: 100 x; e, f: 350 x.

254

A E Ottenhoff-Kalff et al.

kinases. The fact that recently, two tyrosine phosphatases were cloned which possess SH2 domains [19, 20], points to the fact that there may be cross talk between tyrosine-phosphorylated growth factor receptors, for example, and PTPs. Also, protein tyrosine phosphatases have been reported to be inhibited by intracellular factors [30], and, for example, by very low physiological concentrations of zinc ions [31]. Using an enzyme-histochemical assay, we also detected phosphotyrosine phosphatase activity in normal mammary and mammary carcinoma tissues. The tumour cells showed a diffuse activity in the cytoplasm and a clear activity on the membranes, and a strong membrane staining in the interstitium of the tumour tissues was also seen. In the normal cells, the staining in mammary cells and in the interstitium were markedly lower than in the tumour tissue. The fact that the staining in the interstitium of the tumour tissue was so high, may reflect an interaction or communication between the malignant cells and the surrounding tissue. In this regard, the strong membrane staining seen in the interstitium may be suggestive for the metastatic capacity of the turnout tissue. In a recent paper, the likelihood of metastasis was correlated to the presence of angiogenesis, or the blood vessel count in the tumour tissue [32]. This again suggests the possibility that the strong staining in the interstitium, which harbors the blood vessels, is a reflection of the metastatic capacity of the breast tumours, particularly because there also seems to be a correlation between the membrane PTP activity measured in the biochemical assay, and the presence of metastases in the axillary lymph nodes. Two previous reports have shown by enzyme-histochemicalmethods that the protein tyrosine phosphatase activity in breast cancer is increased with respect to normal tissue [27, 33]. In the report by Kidd etal. [27], the increase in PTP activity did not correlate with the presence of c-erbB-2 overexpression, nor with the grade and stage of the disease. However, the detection of PTP activity by enzyme-histochemical means can only give semi-quantitative impressions of the PTP activity. The observation that the PTP activity in the breast tumours is increased compared to normal tis-

sue, is hard to reconcile with the idea that aberrant phosphorylation of tyrosyl residues in proteins can lead to oncogenic transformation through the underexpression of a PTR For instance, treatment of NRK cells with vanadate increased the amount of phosphotyrosine in the cells and generated a transformed phenotype [34, 35]. On the other hand, there is also evidence for other scenarios of aberrant tyrosyl phosphorylation than the aforementioned. First of all, in a recent report by Zhai et al., it was shown that the expression of certain phosphatases was substantially increased in human breast epithelial cells (LAR and PTP1B) and in rat mammary epithelial cells (LAR) as a result of n e u - P T K induced neoplastic transformation [36]. The authors suggested that the increase in PTPs might actually enhance the tumorigenicity of the cells under study, as the cell lines which had the highest level of p185he" expression also had the highest expression of LAR and PTPIB [36]. The report by Zhai et al. might also give an indication as to which PTPs are responsible for the elevation in PTP activity which we report here. It will be interesting to determine this in future studies using type-specific PTP antibodies. In addition, one must keep in mind that protein tyrosine phosphatases can be positively involved in the regulation of tyrosine kinase activity, e.g. the Src family of tyrosine kinases are negatively regulated by the phosphorylation of a carboxyterminal tyrosine and become activated by dephosphorylation [1-3]. For example, the membrane receptor tyrosine phosphatase CD45 causes a dephosphorylation of Tyr 505 (the carboxyterminal tyrosine in Lck) of the Lck enzyme, which activates the enzymatic activity of Lck [4]. Also, the overexpression of PTPc~ in rat embryo fibroblasts has been shown to cause cell transformation and activation of pp60 c-sr¢ [37]. The authors showed that the endogenously present Src protein was not phosphorylated on the carboxyterminal tyrosine residue. In this regard, the results of our earlier work may have some interesting implications. The fact that most of the enhanced PTK activity in the breast turnouts is due to the presence of the Src kinase, and the PTP activities are also increased in breast tissues, prompts us to speculate that activation of the

Tyrosine phosphatase activity in breast cancer Src kinase in the tumours may be caused at least partly by the enhanced activity of the PTPs. Future studies on the regulation of the Src kinase, its substrates, and regulation of the PTPs in breast cancer will be necessary to test this hypothesis.

12.

13.

Acknowledgements The Netherlands Cancer Foundation is acknowledged for the financial support of the work presented here. Furthermore, the authors would like to thank N.H. Steenbergen-Nakken and C.D.H. van Basten for performing the enzyme histochemical assays on the breast tissues.

References 1. Kmiecik TE, Shalloway D: Activation and suppression of pp60c-src transforming ability by mutation of its primary sites of tyrosine phosphorylation. Cell 49: 65-73, 1987 2. Piwnica-Worms H, Saunders KB, Roberts TM, Smith AE, Cheng SH: Tyrosine phosphorylation regulates the biochemical and biological properties of pp60c-src. Cell 49: 7582, 1987 3. Cartwright CA, Eckhart W, Simon S, Kaplan PL: Cell transformation by pp60c-src mutated in the carboxy-terminal regulatory domain. Cell 49: 83-91, 1987 4. Veillette A, Bookman MA, Horak EM, Samelson LE, Bofen JB: Signal transduction through the CD4 receptor involves the activation of the internal membrane tyrosineprotein kinase p561ck. Nature 338: 257-259, 1989 5. Ullrich A, Schlessinger J: Signal transduction by receptors with tyrosine kinase activity. Cell 61: 203-212,1990 6. Schlessinger J, Ullrich A: Growth factor signaling by receptor tyrosine kinases. Neuron 9: 383-391,1992 7. Streuli M, Krueger NX, Tsai AYM, Saito H: A family of receptor-linked protein tyrosine phosphatases in humans and Drosophila. Proc Natl Acad Sci USA 86: 8698-8702, 1989 8. Charbonneau H, Tonks NK, Kumar S, Diltz CD, Harrylock M, Cool DE, Krebs EG, Fischer EH, Walsh KA: Human placenta protein-tyrosine-phosphatase: Amino acid sequence and relationship to a family of receptor-like proteins. Proc Natl Acad Sci USA 86: 5252-5256, 1989 9. Tonks NK, Diltz CD, Fischer EH: Characterization of the major protein-tyrosine-phosphatases of human placenta. J Biol Chem 263: 6731-6737, 1988 10. Tonks NK, Diltz CD, Fischer EH: Purification of the major protein-tyrosine-phosphatases of human placenta. J Biol Chem 263: 6722-6730, 1988 11. Streuli M, Krueger NX, Thai T, Tang M, Saito H: Distinct

14.

15. 16.

17.

18.

19.

20.

21.

22.

23.

24.

25. 26.

27.

255

functional roles of the two intracellular phosphatase like domains of the receptor-linked protein tyrosine phosphatases LCA and LAR. EMBO J 9: 2399-2407, 1990 Fischer EH, Charbonneau H, Tonks NK: Protein tyrosine phosphatases: A diverse family of intracellular and transmembrane enzymes. Science 253: 401-406, 1991 Brautigan DL: Great expectations: protein tyrosine phosphatases in cell regulation. Biochim Biophys Acta 1114: 6377, 1992 Guan K, Dixon JE: Protein tyrosine phosphatase activity of an essential virulence determinant in Yersinia. Science 249: 553-556, 1990 Guan K, Broyles SS, Dixon JE: A Tyr/Ser protein phosphatase encoded by vaccinia virus. Nature 350: 359-362, 1991 Sap J, D'Eustachio R Givol D, Schlessinger J: Cloning and expression of a widely expressed receptor tyrosine phosphatase. Proc Natl Acad Sci USA 87: 6112-6116,1990 Charbonneau H, Tonks NK, Walsh KA, Fischer EH: The leukocyte common antigen (CD45): A putative receptorlinked protein tyrosine phosphatase. Proc Natl Acad Sci USA 85: 7182-7186, 1988 Cool DE, Tonks NK, Charbonneau H, Walsh KA, Fischer EH, Krebs EG: cDNA isolated from a human T-cell library encodes a member of the protein-tyrosine-phosphatase family. Proc Natl Acad Sci USA 86: 5257-5261, 1989 Plutzky J, Neel BG, Rosenberg RD: Isolation of a src homology 2-containing tyrosine phosphatase. Proc Natl Acad Sci USA 89: 1123-1127, 1992 Shen S-H, Bastien L, Posner BI, Chrdtien P: A protein-tyrosine phosphatase with sequence similarity to the SH2 domain of the protein-tyrosine kinases. Nature 352: 736-739, 1991 Hennipman A, van Oirschot BA, Smits J, Rijksen G, Staal GEJ: Tyrosine kinase activity in breast cancer, benign breast disease, and normal breast tissue. Canc Res 49:516-521,1989 Ottenhoff-Kalff AE, Rijksen G, van Beurden EACM, Hennipman A, Michels AA, Staal GEJ: Characterization of protein tyrosine kinases from human breast cancer: Involvement of the c-src oncogene product. Cancer Res 52: 47734778, 1992 Bradford MM: A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72: 248-254, 1976 Hubbard AL, Wall DA, Ma A: Isolation of rat hepatocyte plasma membranes. I. Presence of the three major domains. J Cell Biol 96: 217-229, 1983 Beutler E: In: Red Cell Metabolism: A Manual of Biochemical Methods. Grune & Stratton, New York, 1975, pp 38-70 Rijksen G, van Oirschot BA, Staal GEJ: Nonradioactive assays of protein-tyrosine kinase activity using anti-phosphotyrosine antibodies. In: Hunter T, Sefton BM (eds) Methods in Enzymology, Volume 200, Protein Phosphorylation. Academic Press Inc., San Diego, California, 1991, pp 98-107 Kidd KR, Kerns B-JM, Dodge RK, Wiener JR: Histochemical staining of protein-tyrosine phosphatase activity in pri-

256

28.

29.

30.

31.

32. 33.

A E Ottenhoff-Kalff et al. mary human mammary carcinoma: Relationship with established prognostic indicators. J Histochem Cytochem 40: 729-735, 1992 Hunter T, Sefton BM: Transforming gene product of Rous sarcoma virus phosphorylates tyrosine. Proc Natl Acad Sci " USA 77: 1311-1315, 1980 Slamon D J, Clark GM, Wong SG, Levin WJ, Ullrich A, McGuire WL: Human breast cancer: Correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 235: 177-182, 1987 Ingebritsen TS: Phosphotyrosyl-protein phosphatases II. Identification and characterization of two heat-stable protein inhibitors. J Biol Chem 264: 7754-7759, 1989 Brautigan DL, Bornstein P, Gallis B: Phosphotyrosyl-protein phosphatase specific inhibition by Zn2 +. J Biol Chem 256: 6519-6522, 1981 Marx J: Cellular changes on the route to metastasis. Science 259: 626-629, 1993 Partenen S, Pekonen F: Histochemically demonstrable

34.

35.

36.

37.

phosphotyrosyl-protein phosphatase in normal human breast, in benign breast diseases, and in breast cancer. Anticancer Research 9:667-1989 Klarlund JK: Transformation of cells by an inhibitor of phosphatases acting on phosphotyrosine in proteins. Cell 41: 707-717, 1985 Rijksen G, V611er MCW, Van Zoelen EJJ: Orthovanadate both mimics and antagonizes the transforming growth factor [3 action on normal rat kidney cells. J Cell Phys 154: 393401, 1993 Zhai Y-E Beittenmiller H, Wang B, Gould MN, Oakley C, Esselman W J, Welsch CW: Increased expression of specific protein tyrosine phosphatases in human breast epithelial cells neoplastically transformed by the neu oncogene. Cancer Res 53: 2272-2278, 1993 Zheng XM, Wang Y, Pallen CJ: Cell transformation and activation of pp60 e-s~cby overexpression of a protein tyrosine phosphatase. Nature 359: 336-339, 1992