Role of intracellular degradation of epidermal growth factor in mitogenesis

Copyright 0 1982 by Academic Press. Inc. All rights of reproduction in any form reserved 0014-4827/82/11011 I-07$02.00/0

Experimental

ROLE

Cell Research

OF INTRACELLULAR GROWTH

DEGRADATION FACTOR

PETER L. GREENBERG’ Imperial

Cancer

Research

142 (1982) 11 l-l 17

Fund

Laboratories,

OF EPIDERMAL

IN MITOGENESIS and ENRIQUE Lincoln’s

ROZENGURT Inn Fields,

London

WC2,

UK

SUMMARY Addition of leupeptin, methylamine and the antitubulin agent nocodazole did not affect the initial rate of association of 1251-labelled epidermal growth factor (iZSI-EGF) to Swiss mouse 3T3 tibroblast cells in vitro, but continued incubation with these drugs (up to 24 h) led to an increase in cell-associated radioactivity in a time- and dose-dependent fashion. Combinations of these drugs caused additive increments in cell-associated and internalized radioactivity. Throughout the incubation period, 81-89% of the cell-associated rZSI-EGF was internalized. Upon incubation of ‘251-EGF with 3T3 cells in the presence or absence of the three inhibitors of degradation for periods of up to 24 h, and after removal of the surface-bound material, the internalized rz51-EGF was extracted and 42-53 % was found to be biochemically intact (by acid precipitation) and 56-65 % was antigenically similar to native EGF (using double antibody immunoprecipitation in an EGF radioimmunoassay). The extracted internalized lz51-EGF was capable of binding to fresh 3T3 cells. Despite causing a similar increase in intact internalized iZ51-EGF, leupeptin did not interfere with and nocodazole alone or in combination with leupeptin markedly enhanced EGF-stimulated DNA synthesis, whereas methylamine inhibited mitogenesis. These data indicate a dissociation between EGF degradation and DNA synthesis, and are not consistent with the hypothesis that intracellular degradation of EGF is necessary for its mitogenic effects.

Mouse EGF, a single chain polypeptide of 6045 D, stimulates DNA synthesis in quiescent fibroblasts and other cell types [l-3]. Following binding of the growth factor to specific cell surface receptor sites, the epidermal growth factor (EGF)-receptor complexes are internalized and both the hormone [4] and the receptor [5] are subsequently degraded, probably by lysosomal proteases [l-5]. A critical question is whether mitogenesis due to hormonal stimulation of cells is a consequence of events initiated at the plasma membrane or the result of processing of the internalized ligand-receptor complexes. The role of cellular degradation in the long-term biologic effects of EGF, such as stimulation of DNA synthesis remains unresolved. E-821812

Previous studies demonstrated that addition of methylamine, leupeptin or antitubulin agents to cultures of various cell types led to an enhancement of cell-associated 1251-labelled EGF (1251-EGF) due to interference with hormone degradation [69]. However, contrasting results have been reported regarding the effects of these agents on mitogenesis. Whereas cellular DNA synthesis was increased after treatment with antitubulin agents, such as colchicine and nocodazole [l&14], mitogenesis was decreased or increased when using lysomotropic amines [7-81, and either unchanged or increased after exposure to the -

’ Present address: Department of Medicine, Stanford University and Veterans Administration Medical Centers, Palo Alto, CA 94304, USA. Exp Cell Res 142 (1982)

112

Greenberg

and Rozengurt

lysosomal protease inhibitor leupeptin [9]. However, differing cell systems and culture conditions were used in these prior investigations. In order to determine whether intracellular degradation of EGF plays a role in mitogenesis, we evaluated the influence of these agents on cell-associated murine EGF and on mitogenesis, singly and in combination, in the same murine cell system to determine the relationship, if any, between degradation of EGF and its long-term metabolic changes (e.g. mitogenesis). Our data demonstrated that nocodazole alone and combinations of leupeptin and nocodazole blocked the degradation of lz51-EGF by 3T3 cells and markedly enhanced EGF-stimulated DNA synthesis. These findings indicate that EGF degradation is not necessary for mitogenesis.

MATERIALS

AND METHODS

Cell cultures Stock cultures of Swiss mouse 3T3 tibroblasts were propagated in 9&mm Nunc dishes as previously described [ 151. For experimental use, the cells were subcultured into 35-mm Nunc dishes in Dulbecco-Vogt modified Eagle’s (DME) medium with 10% fetal bovine serum (FBS). The cells were re-fed after 2 days growth and incubated for at least a further 5 days before use in the assays.

1251-EGFbinding assay r

Cells were washed twice with binding medium (37°C) which consisted of DME medium containing 0.1% crystalline BSA, 0.1 PM potassium iodide, and 50 mM N, N-bis (2-hydroxyethyl)-2 aminoethane sulfonic acid (BES) adjusted to pH 7. The cells were incubated at 3PC with 1 ml of binding medium containing i*-YEGF at the required concentration. After incubation, unbound radioactivity was removed by washing the cells four times with cold (4°C) phosphate-buffered saline, pH 7.4, containing 0.1% BSA and 0.1 PM potassium iodide (PBS-BSA/KI) [6]. The cells were extracted (30 min at 4°C) with 1 ml of 0.1 N NaOH, and cell-associated radioactivity was determined in a y-counter. Under these conditions, non-specific bindmg, determined as cell-associated radioactivity in the presence of 2 pg/ml of unlabeled EGF, was less than 5% of the total. Cell numbers were determined by trypsinizing replicate dishes of cells, suspending the cells in Isoton, and counting them in a Coulter counter. Exp Cell Res 142 (1982)

Distribution of cell-associated 1251-EGF between the cell surface and cell interior Following incubation with lZ51-EGF, cultures were washed four times with PBS-BSA/KI at 4°C. Cultures were then treated with 1.0 ml of acetic acid (0.2 N, pH 2.5) containing 0.5 N NaCl for 6 min at 4°C to remove cell surface-bound ‘Y-EGF [16]. The liquid was then removed and the remainina cell-associated radioactivity was extracted by incubition for 30 min at 4°C with 0.1 N NaOH. The radioactivity seuarated by acid or NaOH was determined in a y-counter. As a comparative measure, the ability of trypsin to dissociate cell surface-associated ‘Y-EGF from 3T3 tibroblasts was assessed. After incubation of tibroblasts with iZ51-EGF, the cells were washed four times with PBS and incubated in trypsin in 0.02% versene, 250 &ml for 40 min at 3PC as previously described [II]. The reaction was stopped by adding 10% cold FBS. The supematant was harvested and the cells were pelleted by centrifugation. The radioactivity present in the supematant (cell-surface associated) was compared with that in both the cell pellet and that in the cells remaining in the dishes which were then extracted with 0.1 N NaOH.

Extraction of cell-internalized

12%EGF

The lZ51-EGF that remained within the cells after removal of cell surface-bound ‘“I-EGF was extracted by adding 1 ml of 0.1 N HCl containing 0.1% BSA to the monolayer of cells and incubating them at 23°C for 90 min [4]. The acid-extractable material, approx. 85% of the total cell-bound radioactivity. was sonicated. the pellet (containing approx. 1% of the radioactivity) removed by centrifugation, and the remaining supernatant was lyophilized, dialysed and reconstituted with binding medium for further evaluation, or dialysed for 24 h against cold PBS-BSA/KI and then against cold binding medium prior to rebinding studies (see below).

Fraction of intact lz5Z-EGF The acid-insoluble material in the samples of elution medium (from cell surface-bound or internalized YEGF) was precipitated by incubation with trichloroacetic acid (TCA). 5% final concentration. at 4°C for 30 min. The precipitate was washed once’with TCA and removed by centrifugation at 4°C. The acidinsoluble ‘Y-EGF in the precipitate and the acidsoluble Y-EGF in the sunematant were senaratelv measured, as was the radioactivity extracted inparallel dishes by 0.1 N NaOH. The percentage of acidinsoluble ‘251-labelled material represented intact EGF

[61. Radioimmunoassay of lz5Z-EGF The radioactive material removed from the cell surface or extracted from the cell interior as described above was brought to pH 7 with 0.1 N NaOH. Y-EGF was precipitated by double antibody immunoprecipitation. Ten ~1 of a 1: 200 dilution of ‘*Y-EGF (0.025 ng) or 100 ~1 of test material was incubated for 1 h at 3PC with

Intracellular

t N

degradation

113

of EGF

10% DMSO was used as stock solution and further diluted to contain 0.1-l % DMSO urior to use. Colchitine, leupeptin, methylamine and bovine insulin (26 IU/mg) were purchased from Sigma Chemical Co. EGF and rabbit antimouse EGF antiserum were purchased from Collaborative Research, Inc. Goat antirabbit IgG was purchased from Miles-Yeda Ltd. Na, llsI and [3H]thymidine were purchased from New England Nuclear Co. The soluble lactoperoxidase procedure was used to label EGF with lz51 [6]. The specific activity of the rZ51-EGF at preparation was 140000-170000 cpmlng. 1

2

3

4

‘-4

Time (hours)

Fig. 1. Time course of ?-EGF, binding to Swiss 3T3 fibroblasts in the nresence or absence of inhibitors of degradation. Afte; 2-h pretreatment with 10 PM nocodazole and 16-h preincubation with 160 PM leupeptin at 37°C quiescent cultures of 3T3 cells were washed twice with binding medium at 3PC. The cells were then incubated in 1 ml of binding medium containing r29-EGF &ml; 90000 cpmlng 0, without; or A, with nocodazole 10 PM; 0, leupeptin 160 PM; or 0, 10 mM methylamine. The cultures were incubated at 37°C for the indicated time intervals, washed, and the cell-associated radioactivity was determined as described in Materials and Methods. Each point represents the mean of duplicate or triplicate determinations.

500 ~1 phosphate buffer (0.1 M, pH 7) and 10 ~1 of a 1: 150 dilution of EGF antiserum (rabbit antimouse) in the absence or presence of varying amounts of cold EGF (0.5-50 ng). The immune complex formed was precipitated by adding 100 ~1 of goat antirabbit IgG, 100 ~1 of 0.1 M EDTA, and 100 ~1 of 1% normal rabbit serum, and incubating this mixture for 16-24 h at 4°C. After incubation, 1 ml PBS was added, the precipitate was recovered by centrifugation, and the supematant was removed by decanting. The radioactivity (lZ51EGF) remaining in the pellet was counted in a ycounter.

Measurement of DNA synthesis Determinations of cellular DNA synthesis were performed in a 1: 1 mixture of DME medium and Waymouth medium, as previously described [lo, 18-J.The cultures were washed twice with DME medium at 3PC to remove residual serum immediately prior to assaying. The medium (2 ml) contained 1 PM (1 pCi/ml) [3H]thymidine (20 Ci/mM), insulin (1 pg/ml), EGF and drugs indicated. All values are averages of duplicate plates. The radioactivity of the acid-precipitable material was measured as previously described [ 181.

Materials Nocadazole (methyl-j-(2 thienyl-carbonyl-lH-bensimazol-2yl carbamate) (Aldrich Chemical Co.) 20 mM was dissolved in 100% DMSO. Two-hundred PM in

RESULTS Cell-associated 12Y-EGF

Addition of nocodazole, methylamine, or leupeptin to quiescent cultures of 3T3 cells cause a marked increase in the cell-associated 1251-EGF. Fig. 1 shows that the effect was detected as early as 2 h after the addition of 1251-EGF and the drugs. Inhibition of degradation by these agents persisted for 24 h, although at somewhat decreased levels. Optimal concentrations of colchicine, nocodazole, leupeptin and methylamine for decreasing 1251-EGF degradation were 0.1, 10, 160 (80 pg/ml) and 10 mM, respectively (fig. 2). After 3 h of incubation at 37°C for cells exposed to nocodazole, leupeptin and

350

0 Calchlcine . Nocodazole a Leupeptin q MeNH2

300

! 200 150 100 l.4lzl 10’

103 IIM

105 nh4

PM

Drug Concentration

Fig. 2. Dose-response curves for the effects of 0, nocodazole; 0, colchicine; A, leupeptin; and Cl, methylamine on inhibiting the degradation of lz51-EGF by 3T3 cells at 3-h incubation at 37°C. Pretreatment with the drugs and binding with ‘*SI-EGF were carried out as described in the caption to fig. 1. Data are expressed as percentage of the mean control binding values (1 191 cpm/dish) obtained in the absence of the drug. Exp

Cell

Res

142 (1982)

114

Greenberg and Rozengurt

Table 1. Effect of various inhibitors of hormone degradation on the binding and on the stimulation of DNA synthesis in quiescent 3T3 cells by EGF Cell-associated ‘2Sl-EGFa (% of control)

Additions (cont.) Nocodazole CPM) Single

Leupeptin CPM)

of ‘“+V-EGF

24 h

[3H]thymidine incorp.* (% of control) 24 h

;.s 5

174 233 301 206 210

140 154 110 115 149 172 104 164

97 163 232 269 119 105 125 109 93 80

:oo

249 274

204

z

214

178

233

Methylamine (mW

3h

agents

0.2 1.0 2 10

15 30 60 160 -

98” 100 129 185 -

Mixtures

2

30

i 1:

30 160 30

;.s 2.5 2.5

321 172 339 319

157 156 498 219

159 116 260 129

160

10 10 10

309 307 300

588 338 654

60 196 211

10 10

160

’ Cells were pretreated with nocodazole for 2 h and leupeptin for 16 h at the indicated concentrations. Cellassociated radioactivity was measured after 3 and 24 h incubations at 37°C with binding medium containing l-2 ng of rz51-EGF (10000&160000 cpm/ng) per ml in the presence or absence of the indicated degradationinhibiting drugs. b Cells were pretreated with leupeptin for 16 h at the indicated concentration. [3H]Thymidine incorporation was measured after 24-h incubation at 37°C with a 1: 1 mixture of Waymouth’s : DME medium containing EGF (2.5 nglml) plus insulin (1 pg/ml) in the presence or absence of the indicated drugs. c Each value is the mean of 2-6 separate experiments and is expressed as a percentage of the control value obtained in the absence of added drugs (the diluents for the drugs were added). SDS were generally
methylamine, 185, 301 and 274%, respectively, cell-associated 12”I-EGF was found relative to control values (table 1). These findings are in accord with previous reports of this [6] and other laboratories [8, 9, 191 obtained in other cell types. Suboptimal concentrations of these drugs in combination caused additive effects on decreasing degradation of *251-EGF at 3 and 24 h. Combinations of these drugs at optimal concentrations had additive effects and caused a striking increase (338-654% relative to conEip CellRes

142 (1982)

trols) in cell-associated radioactivity 24 h of incubation (table 1).

after

Distribution of 12%EGF between the cell surface and the cell interior After exposure of the cell ; to 1251-EGF at 37°C for varying periods (45 min to 24 h) the 1251-EGF remaining on the cell surface, determined by the acetic acid elution procedure [16], was ll-19% of the cell-associated radioactivity for all the drugs and controls. A slight increase from 11 to 19 % sur-

Intracellular

degradation

of EGF

115

face-bound 1251-EGF occurred for the control cells, and a 3-16% decrease in these values was noted for the drug-treated cells throughout the incubation period. The cell surface-bound material was also assessed by independent means, using trypsin removal. After 1 and 3 h incubation, the radioactivity removed by trypsin when cells had been exposed to nocodazole was 13 and 20%, respectively. 200

Proportion of biochemically and antigenical1.y intact 12V-EGF

To characterize further the cell-associated ‘251-EGF, 3T3 cells were exposed to the radioactive peptide in the absence or presence of the inhibitors of degradation for various times [I, 2, 3, 4, 16, 24 h). Then the cultures were washed and the surfacebound and internalized 1251-EGF were obtained by sequential acetic acid elution [16] and HCI extraction [4]. As judged by TCA precipitability, 54-69% of the external material and 42-54 % of the internalized radioactive material were intact EGF throughout the incubation periods. Further, antigenic similarity of both the internalized and cell surface-associated ‘*“I-EGF with native EGF was demonstrated, as 60-90% of the external and 56-65 % of the internalized material was specifically immunoprecipitated by a rabbit anti-serum directed against mouse EGF, throughout the same incubation periods. Nearly all the radioactivity in test samples and 12”1-EGF standards was displaced by the addition of a large excess of unlabeled EGF. Both TCA and immunoprecipitation values were essentially unaffected by the addition of the various inhibitors of ‘““I-EGF degradation, singly or in combination. The native 12”I-EGF had 69% of its radioactivity precipitable by TCA and 74% precipitable by radioimmunoassay.

4w

500

Km

1000

12cm

Fig. 3. Binding of extracted cell-processed ‘WEGF to-Swiss 3T3 &Us. Following incubation of cells (5x 106 cells ner 90-mm tissue culture dish, 1.5 ml binding medium per dish) with ‘*%EGF (10.4 r&ml; 100006 cpmlng) for 3 h at 37°C in the absence or presence of nocodazole 10 PM, the cell surface-associated and cell-internalized materials were separately extracted as described in Materials and Methods. The cells exposed to nocodazole were pretreated with this agent for 2 h. The separate extracts from cells within 7-11 culture dishes were pooled, sonicated, and lyophilized. The lyophilized materials were separately resuspended in binding medium and dialysed for 24 h against PBSBSA/KI and then against binding medium. The dialysed extracts containing *251-EGF-like activity (9.3 pgjlo” cpm) which was cell surface-associated after incubation 0, with or A, without nocodazole, or internalized 0, with or A, without nocodazole, were incubated with fresh 3T3 cells (in 30-mm tissue culture dishes) in 0.5 ml binding medium for 45 min at 3X. Native *Z51-EGF (150000 cpmlng) which was similarly handled but which had not been exposed to cells (O), and different concentrations of native ‘*“I-EGF (150000 cpm/ng) (0) were also incubated in parallel dishes under the same experimental conditions. Specific binding was demonstrated for all extracts by disnlacement in parallel dishes with 500 &ml cold EGF. The cell-associated radioactivity was determined as described in Material and Methods. Inset, Expanded scale of ‘*“I-EGF binding with extracted materials.

Binding of extracted cell-processed 12%EGF

In view of the fact that a considerable fraction of cell-associated 1251-EGF was intact, it was of interest to determine whether this material retained the ability to bind to cells not previously exposed to EGF. To this end, external and internalized ‘251-EGF were separately extracted from 3T3 cells after 3 h incubation in the presence or abExpCell

Res 142 (1982)

116

Greenberg

and Rozengurt

sence of nocodazole as described in Materials and Methods. As shown in a representative experiment in fig. 3, internalized ‘*“I-EGF extracts showed substantial but somewhat diminished binding ability. Specific binding was demonstrated by displacement with unlabelled EGF. When the data were corrected to take into consideration the percentage of internalized ‘*“I-EGF which was biochemically or antigenically intact, the rebinding values more closely approximated the binding obtained with native or cell surface-bound ‘*“I-EGF (data not shown).

dence of cell toxicity was noted during incubation in the cells exposed to these drugs. DISCUSSION

After binding of lzsI-EGF to specific plasma membrane receptors, the EGF-receptor complex is internalized and degraded through a lysosomal pathway which takes place after fusion of endocytic vesicles with lysosomes [l-5]. The functional significance of hormone degradation in eliciting the mitogenic response has been controversial [8-14, 191. In order to determine the role of intracellular degradation of EGF for DNA synthesis stimulating DNA synthesis, we utilized a Table 1 shows that addition of nocodazole variety of pharmacologic agents which inThree drugs (2-10 PM) to quiescent cultures of Swiss hibited peptide degradation. 3T3 cells exposed to EGF and insulin en- with different mechanisms of action-nocohanced [3H]thymidine incorporation into dazole, which inhibited microtubule polyacid-insoluble material. In other experimerization [ 143, leupeptin, which inhibited ments we found that the amounts of increthe lysosomal protease cathepsin B [20, 211 mental cell-associated i2”I-EGF were correand methylamine, which altered lysosomal lated with enhanced DNA synthesis in cells and intracellular pH [2 l-231, interfered with exposed to various concentrations of nocodegradation of EGF when added singly, and dazole (results not shown). Leupeptin ef- more potently, when combined at high confected no consistent change, whereas addicentrations. These effects persisted for 24 tion of methylamine resulted in a dose- h, the period when DNA synthesis was related inhibition of mitogenesis (to 22% measured. of control). A particularly salient feature in Following the specific binding of ‘2sItable 1 is that addition of combinations of EGF with surface receptors on 3T3 cells nocodazole and leupeptin to 3T3 cells throughout the 24 h incubation period, 81markedly enhanced DNA synthesis at 89 % of the cell-associated hormone was inmaximal concentrations (260%), and ternalized and 42-54% of this material recaused a striking increase (498%) in 12sI- mained biochemically intact (TCA-precipiEGF cell-associated radioactivity. In con- table) and 5665% was antigenically intact trast, methylamine, which was much less (EGF radioimmunoassay) in the presence efficient than a combination of leupeptin or absence of the degradation-inhibiting and nocodazole in blocking ‘2sI-EGF de- drugs. The internalized ‘Y-EGF retained gradation, inhibited [3H]thymicline incorpobiologic activity after 3 h incubation with ration stimulated by EGF. Thus, the effect 3T3 cells upon its extraction, as shown by of methylamine on EGF degradation cannot its ability to rebind to fresh 3T3 cells. Total account for the inhibition of DNA synthesis cell-associated lzsI-EGF (not internalized caused by this drug. No morphologic evi- material alone) after 2 h exposure with leuExp Cell Res 142 (1982)

Intracellular peptin, has previously been shown to have antigenic similarity to native EGF [9]. Despite causing similar increases in the amount of intact internalized lz51-EGF, the three drugs caused different mitogenic effects. Nocodazole markedly enhanced and leupeptin did not appreciably alter EGFstimulated DNA synthesis, whereas methylamine inhibited mitogenesis. When added in combinations, methylamine interfered with the nocodazole-induced augmentation of EGF-stimulated mitogenesis, whereas leupeptin did not. Crucially, nocodazole and leupeptin, added together, caused a potent inhibition of EGF degradation, but these drugs did not interfere with the stimulation of DNA synthesis produced by this peptide. Since the amount of intact EGF which persists within cells treated with nocodazole and leupeptin is considerably higher than that obtained in the presence of methylamine, it seems reasonable to conclude that the inhibition of the mitogenic response by methylamine is unrelated to its effect on EGF degradation. Whether methylamine inhibits the proliferative response to EGF at a different stage than that blocked by nocodazole and leupeptin [8, 19, 241 or because the drug perturbs proton and related ion movements that occur in mitogen-stimulated cells [23] remains to be investigated. On the basis of the findings presented here, we conclude that EGF degradation is not a necessary intracellular process leading to the induction of mitogenesis.

Printed

in Sweden

degradation

of EGF

117

For this research Dr Greenberg was partially supported by an ACS Eleanor Roosevelt Cancer Fellowship, through the International Union Against Cancer.

REFERENCES 1. Carpenter, Cl & Cohen, S, Ann rev biochem 48 (1979) 193. 2. Catt, K, Harwood, J. Aguilera. G & Dufau, M, Nature 280 (1979) 109: King, A C & Cuatrecasas, P, New Eng j med 305 (1981) 77. Carpenter, G &Cohen, S, J cell biol71(1976) 159. Das, M & Fox, C F, Proc natl acad sci US 75 (1978) 2644. Brown, K, Friedkin, M & Rozengurt, E, Proc natl acad sci US 77 (1980) 480. Maxlield, F, Davies, P, Klempner, L, Willingham, M & Pastan, I, Proc natl acad sci US 76 (1979) 5731. 8. Kina. A C. Hemaez-Davis. L & Cuatrecasas. P. Procnatl acad sci US 78 (1981) 717. 9. Savion, N, Vlodavsky, I & Gospodarowicz, D, Proc natl acad sci US 77 (1980) 1466. 10. Friedkin, M, Legg, A & Rozengurt, E, Proc natl acad sci US 77 (1979) 3909. 11. - Exp cell res 129 (1980) 23. 12. Otto, A, Zumbec, A, Gibson, L, Kubler, A M & Jimenez de Asua, L, Proc natl acad sci US 76 (1979) 5731. 13. McClain, D A & Edelman, G M, Proc natl acad sci US 77 (1980) 2748. 14. Friedkin, M & Rozengurt, E, Adv enzyme reg 19 (1981) 39. 1.5. Dicker, P & Rozengurt, E, Nature 276 ( 16. Haigler, H, Maxfield, F, Willingham, M I, Jbiol them 255 (1980) 1239. 17. Marshall, S & Olefsky, J M, J biol them 254 (1979) 10153. 18. Rozengurt, E & Heppel, L, Proc natl acad sci US 72 (1975) 4492. 19. King, A C, Hernaez-Davis, L & Cuatrecasas, P, Proc natl acad sci US 77 (1980) 3283. 20. Umezawa, H, Methods enzymol55 (1976) 678. 21. Seglen, P 0, Biochem j 174 (1978) 469. 22. Okhuma, S & Poole, B, Proc natl acad sci US 75 (1978) 3327. 23. kozengurt, E, Adv enzyme reg 19 (1981) 61. 24. McKanna, J, Haigler, H & Cohen, S, Proc natl acad sci US 76 (lY79) 5689. Received February 23, 1982 Accepted May 27, 1982

ExpCell

Res

142 (1982)