Publicationof the International Union Against Cancer Publication d e I Union Internationale Contre le Cancer

Int. J. Cancer: 50,724-730 (1992) 0 1992 Wiley-Liss, Inc.

PROTEIN KINASE C ACTIVITY AND EXPRESSION IN NORMAL AND ADENOMATOUS HUMAN PITUITARIES V. ALVARO', Ph. TOURAINE', R. RAISMAN VOZARI~, F. BAI-GRENIER', P. BIRMAN' and D. JOUBERT(BRESSION)'.4 'INSERM U.223, Faculti de Midecine PitiP-SalpCtrikre, 105 Boulevard de I'HGpital, 7.5634Paris Cedex 13; 'INSERM U.289, Hijpital de la SalpgtriPre, Bdtirnent Nouvelle Pharrnacie, 75634 Paris Cedex 13; and 'Laboratoires Sandoz, 14 Boulevard Richelieu, 92500 Rueil-Malmaison, France. Protein kinase C (PKC) activity and expression were measured in 54 adenomas (prolactin (PRL)., growth hormone (GH)and non-secreting), I of them obtained from a patient treated with the dopamine agonist bromocriptine and 2 from patients treated with the somatostatin analog octreotide. They were also measured in normal human and rat pituitaries. Total PKC activity was measured by incorporation of 32Pinto histones, and PKC expression by dot blot immunoquantification using purified PKC as a standard. Both enzyme activity and expression were higher in adenomatous pituitaries than in normal human or rat pituitaries. PKC expression in GH-secreting and non-secreting tumors was significantly higher than that in PRL-secreting tumors. Furthermore, it was significantly higher in invasive tumors than in non-invasive tumors. In the 3 adenomas which were obtained from patients treated with bromocriptine or octreotide and which were used for PKC-activitymeasurement, particulate- and soluble-PKC activities were significantly lower than those measured in non-treated adenomas.

Human pituitary adenomas are tumors resulting from the proliferation of 1 or several pituitary-cell types, and this proliferation can lead to hormonal hypersecretion. These tumors have lost the typical cordonal architecture characterizing the normal pituitary, and basal membranes are either absent or fragmented; they can invade their surrounding dura mata but very seldom metastasize. A key enzyme, implicated in the control of cell proliferation and hormonal secretion, is protein kinase C (PKC), which phosphorylates serine and threonine residues. PKC is a Ca'+- and phospholipiddependent enzyme and represents a family of closely related isoenzymes (On0 et al., 1987; Takai et al., 1985; Huang, 1989), highly concentrated in the central nervous system. PKC is directly activated by diacylglycerol, generated during the turnover of phosphoinositides. This turnover is triggered in the pituitary by growth factors or by neuropeptides or neurotransmitters (O'Brian and Ward, 1989). Not only are neuropeptides and neurotransmitters implicated in the regulation of pituitary hormonal secretion, they may also regulate cell proliferation. Indeed, growth-hormone(GH)-releasing hormone is able to specifically stimulate proliferation of rat somatotrophs in culture, and this stimulation is antagonized by somatostatin (Billestrup et al., 1986). Corticotropin-releasing hormone increases corticotroph-cell proliferation (Gertz et al., 1987), and thyroliberin inhibits that of GH, cells (Ramsdell, 1990). It is now recognized that PKC plays a critical role in cellulargrowth regulation, as shown by several major lines of evidence. Phorbol ester tumor promoters and related agents are potent activators of PKC (Nishizuka, 1984; Weinstein, 1988); PKC transduces mitogenic signals from certain growth factors (Rozengurt, 1989), and over-expression of one PKC isoform can result in multiple growth abnormalities in rat fibroblasts (Housey et al., 1988). PKC activation has been implicated in the induction of mammary adenocarcinoma metastasis (Korczak et al., 1989), and PKC expression has been proposed as a potential tumoral marker for the early detection of breast cancers (O'Brian et al., 1989). In order to investigate whether PKC could be relevant to human pituitary tumor characterization, we measured its activity and expression in pituitary adenomas and in normal human and rat pituitaries. We also measured these parameters

in tumors obtained from 3 patients treated before surgery with bromocriptine [a dopamine (DA) agonist] or octreotide [a somatostatin (SRIH) analog] and in whom these treatments improved hormonal plasma levels and reduced tumor size. MATERIAL AND METHODS

Sample collection Pituitary adenomas (12 PRL-secreting, 21 GH-secreting and 21 non-secreting) were collected in the operating room after trans-sphenoidal adenomectomy. Normal human pituitaries (n = 7, age range 30 to 75 years) were obtained 3 to 5 hr post mortem from subjects with no evidence of neuroendocrine disorder. Fragments of adenomas and normal pituitary tissue (250 mg) were immediately used for enzyme-activity measurement. Other fragments were frozen in liquid nitrogen for quantification of PKC expression. The same sample could not be assayed for both activity and expression owing to the amount of tissue needed for measuring activity (not less than 150 mg). Before surgery, diagnosis was established according to clinical, biological and radiological criteria., Fragments of tumors were fixed for morphological and immunocytochemical studies to confirm diagnosis. Tumor size was evaluated in accordance with the classification of Derome et al. (1979); grade 0 corresponded to a microadenoma, grade I to a tumor that had reached the optic chiasma, grade I1 to a suprasellar extension obliterating the anterior recess of the third ventricle, and grade I11 to a tumor that had reached the foramen of Monro. Invasiveness was characterized by the neurosurgeon during surgical operation by dural infiltration by pituitary tumor cells. PKC activity was assayed in 3 prolactinomas (1 of them obtained from a patient treated with bromocriptine), 5 GHsecreting tumors and 6 non-secreting tumors, 2 of them obtained from patients treated with octreotide (Table I). PKC expression was assayed in 9 prolactinomas, 16 GH-secreting tumors and 15 non-secreting tumors (Tables 11, III and IV). Animals Eighteen female Wistar rats (Iffa-Credo, L'Arbresle, France) weighing 200 g were maintained with free access to water and standard rat chow. After decapitation, anterior pituitaries were immediately weighed and used for enzyme preparation, for measurement of either PKC activity or PKC expression, in the same conditions as those used for human specimens. In order to study the effect of post-mortem delay on PKC activity, liver fragments were collected from rats 3 and 5 hr after death. Measurement of PKC activity The method used was as described by Birman et al. (1989). Briefly, particulate and soluble fractions were prepared from a 4To whom correspondence and reprint requests should be ad-

dressed.

Received: August 30,1991 and in revised form October 25, 1991.

725

PROTEIN KINASE C IN HUMAN PITUITARIES

TABLE I - CLINICAL, BIOLOGICAL AND BIOCHEMICAL DATA ON THE 14 PATIENTS WHOSE ADENOMAS WERE USED FOR THE PKC-ACTIVITY MEASUREMENT

Number

GH-secreting adenomas

1 2 3 4 Non-secreting adenomas 5 6 7 8 9 PRL-secretine " adenomas 10 11 Treated oatients PRL adgnoma 124 Non-secreting adenoma 13' GH adenoma 14'

Imrnunocytochemistry (%positive cells)

Plasma (KgiL)' PRL GH

Age (Yd

Sex

56 40 46 35

F F M F

16 5 8 12

79 59 61 52 42

M M M M M

17 20 15 14 1

0.1 0.1 0.2 0.1 0.1

29 43

F F

44 404

44

F

50 60

Grade

lnvasiveness

GH 40 GH 50. aSu 30' GH 80; aSu 10 GH 80

I I I 111

no no no Yes

NO

IRC~ No IRC a ~ 8n u aSu 30 No IRC

I1 I1 I1 I1 I1

no Yes no no Yes

0.8 1.0

No IRC PRL 80

I I

Yes no

2.100

0.5

PRL 80

I

yes

M

15

0.5

No IRC

I1

no

M

15

GH 60

I

no

'Normal range PRL < 20 bg/l, GH < 5 pg/l.-*aaSu = mub-unit.-'No IRC months octreotide, decrease in tumoral volume.

25 66 61 20

10 =

no immunoreactive cells.-'5 months bromocriptine, decrease in tumoral volume.-'3

TABLE I1 - CLINICAL, BIOLOGICAL AND BIOCHEMICAL DATA ON THE 16 PATIENTS WITH GH-SECRETING ADENOMAS USED FOR PKC EXPRESSION Age

(Yr)

43 31 67 45 40 69 35 61 29 39 37 46 24 38 39 57

Sex

M F F F M M F F M F M F F M F F

Plasma ( p g i l ) PRL GH

10 37 31

25 6 54 12 5 19 15 28 10 2 37 23 23

45 71 20 80 40 50 19 11 60 40 23 50 160 46 15 38

Immunocytochernistry' PRL GH

0 3 30 10 5 10 0 0 0 0 60 15 10 10 10 0

70 50 50 60 80 60 80 40 60 80 15 60 70 60 40 60

Grade

I

I1 I

I I I 111 I1 I1 11 I1 I1 I I1 11 I1

Invasiveness

PKC (ngimg prot.)

no 164.0 no 167.5 no 312.5 326.0 no no 393.0 no 416.0 ves 478.5 Yes 484.9 Yes 501.5 514.0 Yes Yes 580.0 1.258.0 ves ies 11339.0 1,475.0 Yes 1,665.0 Yes Yes 2,237.0 X k SEM = 769 4 150

'% positive cells

homogenate (1:lO wt/vol) and applied, after successivecentrif- Measurement of PKC expression ugations, on to a DEAE-cellulose (DE-52) column (Whatman, Westernblot. Pituitary and liver tissues (70 mg) were homogMaidstone, UK). Enzyme activity was eluted with 0.10 mol/l enized (14:l wt/vol) in 10 mmol/l Tris-HC1 buffer (pH 7.4) NaCl. Activity was measured using the method of Takai et al. containing 2 mmol/l EDTA and 0.1 mmol/l phenylmethyl (1985), by measuring the incorporation of 32Pfrom [Y~~P]-ATPsulfonyl fluoride (PMSF) (Sigma). After centrifugation at 800 (SA 30 Ci/mmol) (Amersham, Aylesbury, UK) into type 111-S g for 4 min, 40 p1of the supernatant were added to 25 pI of 80 histones (Sigma, St. Louis, MO). The reaction mixture (250 pl) mM Tris-HCI buffer (pH 6.8) containing 15% P-mercaptoethcontained 0.1 pg/p1 histones, 25 nmol [y3'P]-ATP, 0.1 pg/pl of anol, 2% SDS (Merck, Darmstadt, Germany) 1% glycerol and diolein, 3.2 ng/pl phosphatidylserine, 13.2 ng/pl of leupeptine to 50 pl of 0.5 M Tris-HC1 buffer (pH 6.8) containing 2% (Sigma), 0.75 rnM calcium chloride (CaCI,), 10 mM magne- P-mercaptoethanol, 10% SDS and 0.05% bromophenol blue. sium acetate (MgAc,) and the enzyme preparation (50 pl). This mixture was boiled for 5 min and immediately used for immunoblotting. Basal PKC activity was estimated in the absence of CaCI,. Samples (10 p1containing 5 to 14 pg proteins per lane) were Protein concentration was measured by the method of Lowry et al. (1951) in particulate and soluble fractions before subjected to 12%-polyacrylamide electrophoresis according to application to the DEAE-cellulose column. No more than 10 Laernmli (1970), and proteins were electrophoretically transmg proteins were ever applied to the DEAE-cellulose column, ferred to a nitrocellulose (NC) sheet for 16 hr at constant and specific PKC activity was calculated for 10 mg of proteins. voltage (30 V). Each NC sheet was then incubated for 1 hr at

726

ALVARO ET AL. TABLE 111 -CLINICAL, BIOLOGlCAL AND BIOCHEMICAL DATA ON THE 15 PATIENTS WITH NON-SECRETING ADENOMAS USED FOR PKC EXPRESSION

2;

Sex

65 43 36 63 45 49 49 32 31 44 43 37 78 31 44

M M M F M M F M M M F M M M M

Plasma (bgil) PRL GH

9 31 15 72 18 16 26 17 7 7 15 7 17 16 17

Immunocytochernistry (% positive cells)

Grade

asu 10' No IRC2 PRL 10 No IRC No IRC No IRC No IRC No IRC uSu 15 No IRC No IRC GH 30 No IRC No IRC a s u 45

I1 111 I1 I11 I1 I1 I1 I1 111 I1 I1 I11 I1 I1 I1

0.2 0.6 0.5 0.4 0.3 0.5 0.9 0.6 0.2 0.1 0.4 0.1 0.1 0.5 0.2

Invasiveness

PKC (ngimg prot.)

no 172.0 no 174.0 no 368.0 no 403.2 no 410.0 no 423.0 no 507.1 no 561.0 Yes 577.2 Yes 592.0 Yes 631.0 Yes 685.0 no 716.0 Yes 932.5 Yes 1,008.7 X f SEM = 544.0 f 59.2

'aSu = asub-unit.-*No IRC = no immunoreactive cells. TABLE IV - CLINICAL, BIOLOGICAL AND BIOCHEMICAL DATA ON THE 9 PATIENTS WITH PRL-SECRETING

ADENOMAS USED FOR PKC EXPRESSION

8

Sex

97 51 30 35 31 28 40 33 43

M F M F M F F F F

Plasma (pgil) PRL GH

105 159 2,000 1.500 190 43 350 105 404

5 0.3 13 8 5 6 10 0.2 1

Irnmunocytochemistry' PRL GH

80 20 20 80 80 20 60 80 80

0 0 0 0 0 0 0 0 0

Grade

I I I I I I 0 I 0

Invasiveness

PKC (ngimg prot.)

no 97.0 no 107.1 no 110.3 no 133.3 no 146.0 no 200.0 no 286.5 no 434.0 no 568.3 X 2 SEM = 231.4 2 52

'% positive cells.

37°C with 30% defatted milk in TBS (50 mM Tris, 9% NaCl), pH 7.4, in order to prevent non-specific background binding, followed by incubation for 1 hr with a monoclonal mouse antibody (MAb) for PKC (Amersham) at 1:400 dilution in TBS 1% BSA (Sigma) in a sealed plastic bag at room temperature. Unbound MAb was removed by washing the NC sheets 4 times for 15 min in TBS 1% BSA. Bound MAbs were detected after incubation with an [ILZS]-labelled second antibody (anti-mouse G immunoglobulin, SA 10 bCi/mmol) (Amersham) for 2 hr at room temperature, diluted in TBS 1% BSA to give 150,000 cpm/lOO ~ 1The . NC sheets were then washed several times with TBS until the radioactivity measured in the wash solution reached the counter background. The NC sheets were then air-dried and exposed for 4 days to pmax film (Amersham) in exposure cassettes (Sigma) at -80°C.

Dot-blot immunoquantiJcation Pituitaries (70 mg) were homogenized (14:l wt/vol) in 10 mmol/l Tris-HC1 buffer (pH 7.4) containing 2 mmol/l of EDTA and 0.1 mmol/l of PMSF. After centrifugation at 800g for 4 min, the supernatant (1.6 to 4 mglml of proteins) was diluted in 7 serial dilutions, and 10 ~1 of each dilution were applied in duplicate to an NC sheet using a dot-blot apparatus (BioRad, Richmond, CA). The NC sheet had previously been marked out in squares (9 X 9 mm) and soaked in TBS. After passive diffusion of the proteins at 0°C through the NC, immunodetection was carried out in the same way as for Western blotting. The NC sheets were then air-dried, the squares cut out and the radioactivity of each square measured.

The standard used was purified PKC (Calbiochem, La Jolla, CA). It was processed in the same way as the homogenate supernatants. The standard curve (Fig. l a ) was plotted from 10 to 240 ng, and was linear between 10 and 120 ng; 1 ng PKC gave a signal of 126 cpm. For each extract, only the linear part of the dilution curve (Fig. lb, typical experiment) was used for PKC quantification. A minimum of 4 dilutions were used in all cases. Data were analyzed using Mann-Whitney's U-test or Student's t-test. RESULTS

PKC activity Normal tissue. Particulate- and soluble-PKC activities (Table V, Fig. 2) measured in 3 normal pituitaries obtained 3,4, and 5 hrpost mortem were, respectively, 1.7 ? 0.7 and 11.7 t 5.3 pmol 32P/mn/10mg proteins (X k SEM; n = 3). These values were lower than those measured in the normal rat pituitary, which were 4.8 2 1.0 and 25.0 2 4.8 (n = 4) in the particulate and soluble fraction, respectively. The recovery of rat liver PKC activity, 3 and 5 hr after death, was 85 and 75%, respectively, in the particulate fraction and 71 and 50%, respectively, in the soluble fraction. Adenomas. Irrespective of adenoma type (Table V, Fig. 2), both particulate- and soluble-PKC activities were significantly higher (p < 0.01) in adenomas than in the normal human tissue [particulate: 15.6 k 2.6 mol 32P/mn/10mg proteins (Z ? SEM) vs. 1.7 2 0.7 pmol 'P/mn/lO mg proteins in the

727

PROTEIN KINASE C IN HUMAN PITUITARIES TABLE v - PKC SPECIFIC ACTIVITY (PMOL OF '~PIMNIIOMG OF PROTEINS) IN ADENOMAS

18.-

Soluble fraction

Particulate fraction

69.2 40.9 67.2 64.3 60.4 r 5.7

15.6 10.6 14.4 9.6 12.6 2 1.3

57.4 24.4 26.1 96.1 47.3 50.3 & 11.7

10.6 11.6 7.1 28.7 10.2 13.7 2 3.4

56.0 37.6 46.8 k 6.5

37.6 15.8 26.7 k 7.7

53.3 & 6.1

15.6

Non-treated adenomas Somatotropes 1 2 3 4

X 2 SEM Non-secreting 5 6 7 8 9

X t SEM Prolactinomas 10 11

X

SEM

X t SEM Treated adenomas Prolactinoma + bromocriptine non-secreting + octreotide Somatotrope + octreotide X k SEM Normal pituitaries Human (n = 3) Rat (n = 4)'

2.6

4.0

2.2

13.2

8.9

17.3

1.3

11.5 r 3.2

4.1 k 1.9

11.7 k 5.3 25.0 f 4.8

1.7 f 0.7 4.8 k 1.0

In, number of experiments; 20 rat pituitaries were used in each experiment.

FIGURE 1- PKC expression after immunoquantificationby dotblot analysis: (a) Standard curve with purified PKC. (b) Typical protein-dilutioncurve for an adenoma homogenate. The adenoma used in this experiment was adenoma number 12 in Table I, a GH adenoma having a PKC concentration of 1.258 ng/mg protein.

normal tissue; soluble: 53.3 2 6.1 pmol 32P/mn/10mg of proteins vs. 11.7 2 5.3 pmol of 32P/mn/10mg proteins in the normal tissue]. There was no statistical difference in PKC activity between GH-secreting, PRL-secreting and nonsecreting tumors, and no significant correlation between PKC activity and plasma hormone levels, tumor size or invasiveness.

There was no correlation between PKC expression and GH or PRL plasma levels. However, PKC expression was higher ( p < 0.001; Mann-Whitney's U-test) in invasive tumors (934.9 2 126.3 ng of PKC/mg of proteins; n = 25) than in non-invasive tumors (316.5 2 34.3 ng of PKC/mg of proteins; n = 15) (Tables 11,111, IV). Effects of dopamine agonist and somatostatin analog treatments on PKC activity In the 3 adenomas obtained from patients treated before surgery with bromocriptine or octreotide, particulate- and soluble-PKC activity were, respectively, 4.1 t 1.9 and 11.5 ? 3.2 pmol 32P/mn/10mg proteins (Table V). These values were significantly lower than those for the non-treated adenomas.

PKC expression As shown by Western-blot analysis, the anti-serum recognized PKC in all types of adenoma, and in the liver used as the control, at a molecular weight of 80 kDa (Fig. 3). PKC expression in the different types of adenomas (Tables DISCUSSION 11, III, IV; Fig. 4) was as follows: Our results show that both PKC activity and PKC expression 769.0 -C 150 ng of PKCimg of proteins (X 5 SEM; range 164 are higher in human pituitary tumors than in normal human to 2,237; n = 16) for GH-secreting adenomas; pituitaries obtainedpost mortem and than in rat pituitaries. 544.0 2 59.2 ng of PKC/mg of proteins (range 172 to 1,008; Levels of PKC activity (particulate and soluble) are elevated n = 15) for non-secreting adenomas; in adenomas, whatever the type of adenoma, when compared 231.4 2 52 ng of PKC/mg of proteins (range 97 to 568; with those in normal human and rat pituitaries. The fact that n = 9) for prolactinomas. all normal human samples were obtained after several hours PKC expression in GH-secreting and non-secreting adeno- post-mortem delay may artefactually account for the difference mas was statistically higher (p < 0.005; Student's t-test) than in enzyme activity observed between normal and tumoral that in PRL-secreting tumors. However, there was no signifi- human tissues. However, 2 arguments militate against such an cant difference between GH-secreting and non-secreting ade- interpretation: (1) when PKC activity was analyzed in the rat nomas. In all cases, mean values were significantly higher liver, a 5-hr post-mortem delay induced a decrease of only 25 (p < 0.01) than those measured in the normal rat pituitary and 50%, respectively, in the particulate- and soluble-PKC (90.3 t 5.8 ng of PKC/mg of proteins; n = 6) and in the activity, and all normal human samples were obtained with normal human pituitary (129.5 2 17.0 ng of PKC/mg of post-mortem delays equal to or lower than 5 hr; (2) when corrected for the post-mortem delay recovery, PKC activity in proteins; n = 4).

ALVARO ET AL.

728 PKC speclflc actlvlty (prnol 32Plrnn110 rng

PKC expression

(nglmg prot.)

prot.)

80

0

T

6C

0

soluble activity 0

particulate activity

0

A 0 4(

0

0

T

O

8

0

A

0 0

21

T %

O

n

I

II

Ill

I I

CH

PRL

adenoma

Ns type

treated

human

normal

rat

pltultarles

F I G U R E2 - Particulate- and soluble-PKC activities (mean k SEM) measured in normal (human and rat) pituitaries, in non-treated [PRL-secreting, GH-secreting and non-secreting (NS)] adenomas and in treated adenomas.

non invasive

invasive

FIGURE 4 - Relation between PKC expression in non-secreting, PRL-secreting and GH-secreting adenomas, tumor grade and invasiveness. Non-invasive tumors have significantly ( p < 0.001) lower amounts of PKC than do invasive tumors.

the normal human pituitaries did not differ from that measured in the rat pituitaries, which was significantly lower than the values measured in the adenomas. Moreover, in other studies using normal human pituitaries collected after similar post-mortem delays, we have shown that in vitro GH or PRL release is regulated by hypothalamic factors (Joubert et al., 1989),for which the receptors have been identified (Bression ef al., 1980). This means that the normal pituitary cells are indeed functional. Finally, the fact that similar PKC-activity and PKC-expression values were found in the various normal human pituitaries used in our study suggests that non-specific PKC degradation due to cell death and tissue necrosis does not significantly affect the values obtained. FIGURE 3 - Western-blot analysis of PKC. PKC was analyzed by There was no statistical difference between secreting and immunodetection with a PKC antibody and an [I'zs]-labelled second antibody. A major 80 kDa band was identified whatever the non-secreting tumors, and no relation was observed between tissue used. Lane 1, non-secreting adenorna; lane 2, PRL-secreting the PKC activity measured in the adenomas and the volume or adenoma; lane 3, GH-secreting adenoma; lane 4,rat liver; lane 5 , invasiveness of the tumor, or the hormonal plasma levels of the patients. However, the number of tumors analyzed in each standard PKC.

729

PROTEIN KINASE C IN HUMAN PITUITARIES

group of adenomas may not have been sufficient to allow any clear conclusion to be drawn on the relationship between PKC activity and these 3 parameters. PKC activity has also been found to be higher in surgical samples from human breast tumors as compared to that found in samples of normal breast tissue obtained from the same patients (O’Brian et al., 1989). This observation made in breast tissue is in agreement with ours in pituitary tissue in that the tumoral tissue expresses higher PKC activity than does the normal tissue. However, in that colon carcinomas, CaZfdependent protein kinase activity has been found to be lower than that in the adjacent normal colon mucosa (Huang, 1989; Guillem et al., 1987a,b). Likewise, skin biopsies from patients with the proliferation skin disorder psoriasis have reduced levels of PKC activity compared to skin biopsies from control persons (Horn et al., 1987). Thus, when alterations in the level of PKC activity are observed in tumor cells, they may represent either an increase or a decrease in enzyme activity. Our results also indicate an elevated level of PKC, measured by immunoquantification, in human pituitary tumors when compared with that in normal rat and human pituitaries. The increase in the amount of PKC was greatest in invasive tumors, but there was no relationship with the hormonal plasma level or the size of the tumor. The absence of a relationship between PKC expression and the size of the tumor was particularly apparent in GH-secreting and non-secreting tumors, which are large tumors and where non-invasive tumors have lower levels of PKC than do invasive ones. However, in the case of prolactinomas, which are all detected early due to the clinical symptoms, it cannot be concluded that, in this particular type of adenoma, there is no relationship between PKC expression and tumor size, as all tumors were small and non-invasive. Modified PKC expression has also been observed by Reifenberger in human brain tumors. Indeed, using immunohistochemical determination of PKC expression, Reifenberger et al. (1989) have shown that measurement of PKC expression may be a useful aid to histopathological tumor classification in neuro-oncology. The antibody we used for immunoquantification recognizes a and p PKC isotypes only indistinctly. Our results, therefore, do not give any information as to which isotype is responsible

for the observed increase in PKC expression. Moreover, expression of one isotype may have decreased while expression of another may have increased. Further work is thus needed to clarify this particular point. In this regard, pituitary tumors may represent a useful tool, as they are all characterized by increased cell proliferation but not always by hormonal hypersecretion. In the rat pituitary, the a and @ isotypes have been detected and both have been shown to be involved in the regulation of hormonal secretion (Naor et al., 1989; Naor, 1990). No information is available so far as to which isotype is implicated in pituitary-cell proliferation. In rat fibroblast, the over-expression of the PI isotype of PKC results in multiple growth abnormalities (Korczak et al., 1989), and in a related study, when NIH-3T3 cells were transfected with y PKC isotype encoding cDNA, altered growth regulation was observed (Persons et al., 1988). Lastly, in the 3 tumors obtained from patients treated with the DA agonist or the SRIH analog, PKC activity was much lower than in the other tumors. This result is in agreement with a previous study showing that in vivo treatment with the DA agonist CV 205-502 induces a decrease in pituitary PKC activity both in rats treated with estradiol and in those not treated with estradiol (Birman et al., 1989). This study, which needs to be extended to a larger number of observations, suggests that these molecules exert their in vivo biological effects (inhibition of hormone release, reduction of tumor size) in part by regulating PKC activity. Taken together, these results suggest that, in human pituitary tumors, PKC activity and expression are increased in comparison with normal pituitary tissue (human and rat). This indicates that PKC may participate in or parallel the mechanism leading to the abnormal control of pituitary-cell proliferation and hormone secretion. ACKNOWLEDGEMENTS

We thank the neurosurgeons of Hapita1 Foch (Suresnes) for providing the pituitary fragments, Mrs. M. Kujas for immunocytochemical studies, Mrs. M. Roche for technical assistance and Mrs. M. Le Guennec for the preparation of the manuscript.

REFERENCES BILLESTRUP, N., SWANSON, L.W. and VALE, W., Growth-hormonereleasing factor stimulates proliferation of somatotrophs in vitro. Proc. nut. Acad. Sci. (Wash.), 83,6854-6857 (1986). BIRMAN,P., TOURAINE, Ph., BAI-GRENIER, F., DUBRAY,C., KAABACHE,T., PEILLON,F. and JOUBERT-BRESSION, D., The stimulated C-kinase activity in estradiol-treated rat pituitaries is reduced by a chronic treatment with the dopamine agonist CV 205-502. Acta endocrinol., 121,489494 (1989). BRESSION, D., BRANDI, A.M., MARTRES, M.P., NOUSBAUM, A,, CESSELIN, F., RACADOT,J. and PEILLON,F., Dopaminergic receptors in human prolactin-secreting adenomas: a quantitative study. J. clin. endaainol. Metab., 51,1037-1043 (1980). DEROME,P.J., PEILLON, F., BARD,R.H., JEYDINAK, C.P., RACADOT, J. and GUIOT,G., RCsultats chirurgicaux dans 150 cas d’adhomes a prolactine. Nouv. Presse Med., 8,577-581 (1979). GERTZ,B.J., CONTRERAS, L.N., MCCOMB,D.J., KOVACS,K., BLAKETYRELL, J. and DALLMAN,M.F., Chronic administration of corticitropinreleasing factor increases pituitary corticotroph number. Endocrinology, 120,381-388 (1987). GUILLEM, J.G., OBRIAN,C.A., FITZER,C.J., FORDE,K.A., Lo GERFO, P., TREAT,M. and WEINSTEIN, I.B., Altered levels of protein kinase C and Ca”-dependent protein kinases in human colon carcinomas. Cancer Res., 47,2036-2039 ( 1 9 8 7 ~ ) . GUILLEM, J.G., O’BRIAN,C.A., FITZER,C.J., JOHNSON, M.D., FORDE, K.A., Lo GERFO,P. and WEINSTEIN, I.B., Studies on protein kinase C and colon carcinogenesis. Arch. Surg., 122,1475-1478 (1987b).

HORN,F., MARKS,F., FISHER,G., MARCELO, C.L. and VOORHEES, J.J., Decreased protein kinase C activity in psoriatic versus normal epidermis. J. invest. Dermatol., 88,220-222 (1987). HOUSEY,G.M., JOHNSON,M.D., HSIAO,W.L.W., O’BRIAN,C.A., MURPHY, J.P., KIRSCHMEIER, P. and WEINSTEIN, I.B., Overproduction of rotein kinase C causes disordered growth control in rat fibroblasts. Cei, 52,343-344 (1988). HUANG,K.P., The mechanism of protein kinase C activation. Trends Neurosci., 12,425-431 (1989). JOUBERT (BRESSION), D., BENLOT,C., LAGOGUEY, A,, GARNIER, P., BRANDI,A.M., GAUTRON,J.P., LEGRAND,J.C. and PEILLON,F., Normal and growth hormone (GH)-secreting adenomatous human pituitaries release somatostatin and GH-releasing hormone. J. clin. endocrinol. Metab., 68,572-577 (1989). KORCZAK, B., WHALE,C. and KERBEL,R.S., Possible involvement of Ca2+mobilization and protein kinase C activation in the induction of spontaneous metastasis by mouse mammary adenocarcinoma cells. Cancer Res., 49,2597-2602 (1989). LAEMMLI, U.K., Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (Lond.), 227,680-685 (1970). LOWRY,O.H., ROSEBROUGH, N.J., FARR,A.L. and RANDALL,R., Protein measurement with the Fohn phenol reagent. J. biol. Chem., 193,265-271 (1951). NAOR,Z . , Further characterization of protein kinase-C subspecies in the hypothalamo-pituitary axis: differential activation by phorbol esters. Endocrinology, 126,1521-1526 (1990).

730

ALVARO ET AL.

NAOR,Z., DAN-COHEN, H., HERMON, J. and LIMOR,R., Induction of exocytosis in permeabilized pituitary cells by a- and p-type protein kinase C. Proc. nut. Acad. Scz. (Wash.), 86,4501-4504 (1989). NISHIZUKA, Y., The role of protein kinase C in cell-surface signal transduction and tumour promotion. Nature (Lond.), 308, 693-697 (1984). O’BRIAN,C.A., VOGEL,V.G., SINGLETARY, S.E. and WARD, N.E., Elevated protein kinase C expression in human breast tumor biopsies relative to normal breast tissue. CancerRes., 49,3215-3217 (1989). O’BRIAN,C.A. and WARD, N.E., Biology of the protein kinase C family. Cancer Metastasis Rev., 8, 199-214 (1989). ONO, Y., FUJII,T., OGITA,K., KIKKAUVA,U., IGARASHI, K. and NISHIZUKA, Y., Identification of three additional members of rat protein kinase C family: a-,p- and y-sub-species. FEBS Lett., 226, 125-128 (1987). PERSONS, D.A., WILKINSON, W.O., BELL,R.M. and FINN,O.J., Altered

growth regulation and enhanced tumorigenicity of NIH 3T3 fibroblasts transfected with protein kinase C-1 cDNA. Cell, 52,343-354 (1988). RAMSDELL, J.S., Thyrotropin-releasing hormone inhibits GH4 pituitary cell proliferation by blocking entry into S-phase. Endocrinology, 126,472479 (1990). . . REIFENBERGER, G., DECKERT,M. and WECHSLER, W., Immunohistochemical determination of protein kinase C expression and proliferative activity in human brain tumors. Acfa neuropathof., 78, 166-175 (1989). ROZENGURT, E., Signal transduction pathways in mitogenesis. Brit. med. Bull., 45,515-528 (1989). TAKAI,Y., KAIBUCHI, K., TSUDA,T. and HOSHIJIMA, M., Role of protein kinase C in transmembrane signaling. J. cell. Biochem., 29, 143-155 (1985). WEINSTEIN, B., The origins of human cancer: molecular mechanisms of carcinogenesis and their implications for cancer prevention and treatment. CancerRes., 48,41354143 (1988).