medium (EMEM; Gibco Ltd, Trident House, Paisley, Renfrewshire, Scotland, supplemented with 10% fetal calf serum and antibiotics, penicillin 50 IU/ml and.
Bioscience Reports, Vol. 9, No. 1, 1989
Lateral Diffusion of Plasma Membrane Receptors Labelled with Either PlateletDerived Growth Factor (PDGF) or Wheat Germ Agglutinin (WGA) in Human Polymorphonuclear Leukocytes and Fibroblasts 1 Pia Ljungquist 2, ~ke Wasteson and Karl-Eric Magnusson Received June 30, 1988 The aims of the present investigation were (a) to compare the lateral mobility of membrane receptors of human fibroblasts and polymorphonuclear leukocytes (PMNL) labelled with either platelet-derived growth factor (PDGF), or the lectin wheat germ agglutinin (WGA), and (b) to study effects of serum or PDGF on the mobility of these receptor molecules in human fibroblasts. Human foreskin fibroblasts (AG 1523) were grown on coverslips either under standard (10%) or under serum-free conditions yielding "normal" and "starved" cells, respectively. The receptor mobility was studied in response to exposure to PDGF, or serum, in short time or prolonged incubations. Human polymorphonuclear leukocytes (PMNL) were adhered to microscope slides by clotting drops of blood. They were stained with rhodaminated PDGF or fluoresceinated WGA. The diffusion of labelled receptors was assessed with fluorescence recovery after photobleaching (FRAP). It was found that (a) fibroblasts grown at normal serum concentration had a lower diffusion coefficient ( D = 3 x 10-i~ s -1) for the PDGF-receptor and a slightly lower mobile fraction (R = 60%) than starved cells (D = 5 x 10 10 cm z s - i and R = 73%), (b) addition of serum to starved cells increased both D and R for the PDGF receptor to 12 x 10-1~ s -~ and 96%, respectively, (c) a similar pattern was obtained for WGA-labelled glycoconjugates indicating general membrane effects of serum-induced cell stimulation, and (d) in PMNL the PDGF receptor displayed motility characteristics (D = 3 - 4 x 10-1~ 1 and R = 59%) similar to those in fibroblasts, possibly suggesting equivalent anchorage mechanisms in the membrane. KEY WORDS: diffusion; membrane receptors; platelet-derived growth factor (PDGF); wheat germ agglutinin (WGA); polymorphonuclear leukocytes (PMNL); fibroblasts.
Departments of Cell Biology and Medical Microbiology, Faculty of Health Sciences, University of Link6ping, S-581 85 Link6ping, Sweden. This report was presented in brief at Bioscience 88, April 17-20, 1988, Maim6, Sweden. 2 To whom correspondence should be addressed (Department of Cell Biology). 63
0144-8463/89/0200-0063506.00/09 1989PlenumPublishingCorporation
64
Ljungquist, Wasteson and Magnusson INTRODUCTION
Platelet-derived growth factor (PDGF) is a growth promoting factor stored in the alpha granules of platelets (Bowen-Pope and Ross, 1984; Heldin et al., 1985). By specific interaction with different target cells, e.g. fibroblasts (Bowen-Pope and Ross, 1984), PDGF has been shown to induce cell receptor expression, cytoskeleton effects (Bockus and Stiles, 1984), cell degranulation (Bowen-Pope and Ross, 1984; Heldin et al., 1985) and locomotion (Deuel et al., 1982). PDGF has also been found to be a chemotactic inducer of granulocytes (Grotendorst et al., 1981; Deuel et al., 1982; Williams, 1983). Moreover, it has been shown to induce competence genes for intracellular mediators of the mitogenic response in different types of fibroblasts (Callahan et al., 1985). Glial cells, fibroblasts and fibroblast-like cell lines like 3T3 require PDGF for optimal growth (Stiles, 1983; Westermark et al., 1983), but PDGF has in some instances proved to be insufficient by itself (Pledger et al., 1977; Stiles et al., 1979). However, it has later been found that PDGF can indeed by itself induce a mitogenic response in human diploid fibroblasts (Westermark et al., 1985). Several studies have indicated that receptor clustering induced by epidermal growth factor (EGF) leads to receptor mediated endocytosis (Zidovetzki et al., 1981). Receptor aggregation also seems to play a crucial role in the early stages of the activation process leading to a mitogenic response (Schreiber et al., 1983; Yarden et al., 1985). In this context, allosteric activation of intramolecular auto-phosphorylation by the EGF kinase is assumed to form the key step (Yarden et al., 1985). For EGF binding to human epidermal carcinoma A 431 cells, high and low affinity receptors with different lateral diffusion characteristics have been observed (Rees et al., 1984). The PDGF receptor was recently purified to homogeneity (R6nnstrand et al., 1987; Ek and Heldin, 1982), but in this case also there appears to be receptors with high and low affinity for PDGF molecules consisting of A-A, B-B homodimers, or A-B heterodimers (Heldin, unpublished observations). No biophysical characterisation of the PDGF receptor has been done either in fibroblasts, or in polymorphonuclear leukocytes (Deuel et al,, 1982; Williams, 1983). Nor has the effect of PDGF alone or serum been evaluated along these lines. Several properties of surface receptors depend upon their lateral mobility in the membrane. Although the lipid bilayer in which many surface proteins move is fluid, great variation has been observed in the mobilities of various receptors, even within a specific receptor class (Schlessinger et al., 1976a,b, 1977). In order to characterise the lateral mobility of the PDGF receptor, we labelled the human foreskin fibroblast cell line AG 1523 and peripheral blood PMNL with rhodaminated PDGF, or fluoresceinated lectin wheat germ aggultinin (WGA) and studied the lateral diffusion with fluorescence recovery after photobleaching (FRAP) (Jacobson et al., 1976, Johansson et al., 1987).
MATERIALS AND METHODS Preparation of Fibroblasts and Polymorphonuclear Leukocytes (PMNL) The human foreskin fibroblast cell line AG 1523 (Human Genetic Cell Repository, Camden, NJ) was used in these assays. It was maintained at 37~ in a
Lateral Diffusionof Membrane Receptors
65
humidified incubator with 5% CO2 (Type CO2-Auto-Zero, Heraeus, Hanau, FRG) and subcultivated into Petri-dishes, either in Eagle's minimum essential medium (EMEM; Gibco Ltd, Trident House, Paisley, Renfrewshire, Scotland, supplemented with 10% fetal calf serum and antibiotics, penicillin 50 IU/ml and streptomycin 50/~g/ml), or under serum-free conditions (MCDB 105 medium, Flow Laboratories, Woodcock Hill, England, with i mg/ml serum albumin), yielding "normal" and "starved" cells, respectively. The cells were used for an experiment 3 days after subcultivation. The PMNL cells were taken from peripheral blood by fingertip puncture (2-3 drops). They were incubated for 30min at 37~ on a microscope slide, the clot was then detached from the coverslip at the periphery with a Pasteur pipette, and the PMNL cells attached to the glass rinsed with Krebs-Ringer's phosphate buffer with 10 mM glucose, i mM Ca z+ and Mg z+, pH 7.3 (KRG), which also removed non-adherent erythrocytes. This yielded a well-dispersed population of attached cells, corresponding roughly to a cell density obtained by applying 0.1 ml separated PMNL (2 x 106 per ml) (B0yum, 1968) on a coverslip.
Cell Labelling Fibroblasts or PMNL were incubated either with 10/~1 WGA, conjugated with fluoresceinisothiocyanate (FITC; E-Y Labs, San Mateo, CA) and incubated at 4~ for 5 min, or with 10/~1 P D G F conjugated with rhodamine and incubated at 4~ for 30 rain. In both cases 50/zl K R G was added with the respective ligand and the cells were then rinsed in KRG. At this point K R G with or without serum was added. A wet chamber was made by putting two strips of Parafilm "M ''| (American Can. Co., Greenwich, CT) as spacers and a coverslip from above; the chamber was then sealed with a hot mixture of wax and Vaseline (1:1). Less than 5 rain elapsed between cell labelling and the start of the photobleaching experiment, which was completed within 60 rain.
Labelling of PDGF Two hundred ~1 rhodamine B-isothiocyanate solution (mixed isomers R-1755, Sigma Chemical Co., St Louis, MO; 1 #g/ml in PBS), 0.9ml PBS (phosphate-buffered saline, pH 7.3) and 2 ml PDGF (1000 ng/ml in PBS) were mixed overnight in a tube by end-over-end rotation at 4~ Unreacted rhodamineisothiocyanate was separated by dialysis against PBS; the PBS was changed daily for three days. The PBS was stirred with a magnetic stirrer. The final concentration of labelled PDGF was 250 ng/ml.
Fluorescence Recovery After Photobleaching The mobility of the P D G F receptor, presumably by lateral diffusion in the plane of the membrane, was assessed by the fluorescence recovery after
66
Ljungquist, W a s t e s o n and M a g n u s s o n
photobleaching (FRAP; Peters et al., 1974; Axelrod et aL, 1976; Jacobson et al., 1976; Dahlgren et al., 1980; Peters, 1981; Magnusson et al., 1987). The equipment used to study lateral diffusion has recently been described in detail elsewhere (Johansson et al., 1987). Briefly, data aquisition and handling is done with a small computer, and the mode of measurement chosen from a menu (SLOW-FRAP), allowing sampling at regular or at gradually increasing intervals (At). For each cell studied, At was increased from 2 to 15 s, and the final point F= taken after 120 s. This means that about 20 cells could be studied per hour. A cell sample is illuminated and bleached through a fixed, exchangeable slit, in epifluorescence through a Zeiss Universal microscope, equipped for fluorescein or rhodamine activation. Using a • objective and a round, 160-/~m diameter slit, the 1/e 2 radius (W) of the bleached spot was estimated to be 0.89/~m (Johansson et al., 1987). Using this experimental set up, the fluorescence from the illuminated spot was determined with a photomultiplier tube (S F Photometer, Carl Zeiss, Oberkocken, FRG). The diffusion constant and percent receptor recover)' after photobleaching were calculated according to Axelrod et al. (1976) and Jacobson et al. (1976). The laser used was a Spectra Physics Argon laser (Type 2020-3, Spectra Physics, Mountain View, CA) run in current mode at 35 amps, and at the 488 nm-line.
RESULTS The labelling of human A G 1523 fibroblasts and PMNL with rhodaminePDGF is shown in Fig. 1 and Fig. 2, respectively. Only when the fibroblasts had been grown in the absence of serum (starved cells) was the binding of P D G F strong enough to allow photography without exceedingly long exposure times. The data concerning lateral mobility of PDGF- and WGA-receptors is presented in Table 1 and Fig. 3. The fibroblasts grown in the presence of serum in EMEM medium showed a lower diffusion coefficient (D) and a slightly lower mobile fraction (R) for the PDGF-receptor than fibroblasts grown in the absence of serum (MCDB 105 medium), i.e. starved cells. When serum was present in the Table 1.
Mobility characteristics ~ of P D G F receptors and W G A - l a b e l l e d glycoconjugates in fibroblasts and P M N L
Cell Fibroblast
Fibroblasts
PMNL PMNL
Receptor normal starved starved + s e r u m normal starved starved + s e r u m starved + P D G F
PDGF PDGF PDGF WGA WGA WGA WGA PDGF WGA
Diffusion coefficient, D (10 - l ~ cm 2 s - i )
Mobile fraction, R (%)
No. of exps, n
3.2 5:0.5 5.1 • 0.1 12.4 • 0.3 2.0 + 0.2 1.5 • 0.1 3.9 + 0.6 1.8 • 0.2 3.7 • 0.4 1.1 + 0.1
60 + 7 73 i 3 96 • 2 31 • 3 49 • 4 50 -I- 3 37 • 4 59 :t: 2 32 + 2
12 29 20 16 22 12 11 35 25
Values are given as m e a n + standard error of the m e a n (SEM).
Lateral Diffusion of Membrane Receptors
67
(a)
(b)
Fig. 1. Human AG 1523 fibroblasts: (a) stained with rhodamine-PDGF and (b) phase contrast image. The cells had been grown without serum. chamber with starved cells, D and R increased for the P D G F receptor. Moreover, the cells appeared synchronized, as evidenced by narrow distributions of D and R (Fig. 3). A similar pattern was obtained for WGA-labelled glycoconjugates, indicating general membrane effects of serum-induced cell stimulation (Fig. 4). In the P M N L the P D G F receptor displayed motility characteristics similar to those observed in fibroblasts, both for the P D G F receptor (Fig. 5; Tables 1 and 2)
68
Ljungquist, Wasteson and Magnusson
(a)
(b)
Fig. 2. HumanPMNL: (a) stained with rhodamine-PDGFand (b) phase contrast image. and WGA-labelled glycoconjugates (Fig. 6, Table 1). This could possibly suggest similar anchorage mechanisms of the P D G F receptor in the membrane of fibroblasts and PMNL. When the concentration of rhodamine-PDGF at labelling was reduced from 100 to 20 ng/ml, a negligable effect on receptor mobility was observed (Table 2), possibly suggesting that P D G F p e r se did not change receptor properties.
Lateral Diffusion of Membrane Receptors
1=
r
69
Fibroblasts
: Normal
Fibroblasts
: Starved
i
......
9
//
ill _--,/,"
/ /
A
,
,
2
3
4
Diffusion
,
....
J
S 6 7 8 9 I0
"
F i b r o b l a s t s : St . . . . d . s . .t. .
i
-
,
-
t . . . ~
20
0
50
constcmt |lO-lOcm2S -l)
100
fraction (%}
Mobile
Fig. 3. Lateral mobility characteristics of the PDGF receptor in the human A G 1523 fibroblasts labelled with rhodamine-PDGF, as described by the distributions of values for the diffusion constant (10 -1~ cm 2 s -1) and fraction of mobile receptors (%).
F[ b r o b l a s t s : N o r m a l
Fibrob(osts :Starved
Ul
/.,_~/~ ~"-%e
6
CL..,q
Y
l
I
i
I I Ill
~,
I .
.."
2*
='l-'---
~
9
Fi b r o b l o s t s : S t a r v e d *
|
.d.
serum
E
z
q. . . . . . .
~
~.A
d *,
*,
L I I~
I
~___.____l~
Fi b r o b t aSt s : St o r v e d * PDGF (10,'~/ml)
.8 9
1
Diffusion
2
3 constant
4
5 fi 7 8 9 1 0 (10-10cm25"1)
20
~l
i 9
@
0
50
Mobile
100
f r c a c t i o n {~
Fig. 4. Lateral mobility characteristics of WGA-labelled glycoconjugates in the human A G 1523 fibroblasts, as described by the distribution of values for the diffusion constant (10-to cm 2 s-1) and fraction of mobile receptors (%).
70
Ljungquist, Wasteson and Magnusson
L
(a)
#1
L
PMNL
]
[ PoGFI,ooog/rn,II 9
E=
5 6 2.8.91
Z
2
3
6 5 6 78910
I,.
20
Diffusion constont ( 10-10cm2s-1 )
PMN L ]
~q (b)
EPDGF
I E z
(100ng/m,)
]
" 50
0 Mobile
100
fraction
(%)
Fig. 5. Lateral mobility characteristics of the P D G F receptor in human PMNL labelled with rhodamine-PDGF, as described by the distribution of values for (a) the diffusion constant (10 lO cm 2 s - l ) and (b) the fraction of mobile receptors (%).
Table 2.
Effect of different concentrations of P D G F on the mobility characteristics I of the P D G F receptor in human PMNL PDGF (ng/ml) 100 25 20
Diffusion coefficient, D (10 -10 em 2 S-1)
Mobile fraction, R (%)
No. of exps, n
3.7:1:0.4 5.3 :t: 0.5 4.25:0.6
59 =I=2 65 • 4 68:s
35 33 11
1 Values are given as mean + standard error of the mean (SEM).
.5 .6:7 .8.91 2 3 L, 5 6 7 8910 Diffusion constant (10-10crn2s-1)
J 20
Fig. 6. Lateral mobility of WGA-labelled glycoconjugates in human PMNL as described by the distribution of values for the diffusion constant (10 - l ~ cm z s - t ) ; data for the mobile fraction shown only in Table 1.
71
Lateral Diffusion of Membrane Receptors
]b'~.~"J~
[Fibroblasts POGFI
t
# I b r o b l ~ t m - l x ~ g F - t r l t ~ - 2 3~-b18-488-15-2-120- B ~ I ~ B
1@ ~ f
L - J ~
10
14
Fo - ~
12 lO
s
B
B ,tO
4
LL 2
o
i
i
i
i
i
i
i
i
i
g
1B
2-/
3~
45
54
83
72
BI
e0
T1/2=&Os Fig. 7. Typical FRAP curve obtained for rhodamine-PDGF-labelled receptors in the human A G 1523 fibroblasts.
DISCUSSION The present investigation is aimed at comparing the lateral mobility of membrane receptors labelled with lectin WGA or PDGF in human fibroblasts and PMNL. The receptor for PDGF has recently been purified from pig uterus (R6nnstrand et al., 1987). It has been shown to be a homogenous component with a molecular weight of about 170 kD. PDGF is able to induce phosphorylation of this component on tyrosine residues. This indicates that the kinase domain is an integral part of the receptor molecule (R6nnstrand et al., 1987). We studied first whether we could detect PDGF receptors in human PMNL using rhodaminated PDGF, which was indeed the case (Fig. 2). Secondly, we tried to assess the mobility characteristics with the FRAP-technique (Figs. 5, 6, 8, Tables 1 and 2). Third, the properties of the PDGF receptors were compared with membrane glycoconjugates labelled with the lectin WGA, both in PMNL and fibroblasts (Table 1, Figs. 1, 3, 4 and 7). Biophysical characterisation, e.g. with respect to the lateral mobility, has not yet been done on the PDGF receptor on granulocytes. Therefore, the characteristics of the PDGF receptor in PMNL was compared with a well-known target cell for PDGF, the fibroblast cell line AG 1523. PMNL cells stained with WGA showed lower D and R compared to PDGF-labelled cells.
Ljungquist, Wasteson and Magnusson
72
IP~NL PNNL-RId0D-F'DGFI~I~-r~I/ml-4C-~tn
PDGF I
l~-~: ~
ICO-ID-2$^-$^-,~Ome-l.r~-2l
B71217
110
ge
77
~J c-
s$
~J U
ijr --
Fo ---N4 0
F~
39
11
0
i
g
i
IB
i
2"/
i
i
i
i
i
i
30
4'~
54
~
72
B1
00
Time(s)
T1/2= S, 8= Fig. 8. TypicalFRAP curve obtained for rhodamine-PDGF-labelledreceptorsin human PMNL. Starved cells had a higher diffusion coefficient (D) and percent recovery (R) than cells maintained in serum supplemented-medium. Moreover when serum was added to the starved cells they Significantly increased the diffusion coefficient and the mobile fraction (R). We also observed under the microscope that the PDGF-associated fluorescence was higher in "starved" fibroblast than in "normal" fibroblasts. Also, the PMNLs displayed weak binding of P D G F (Fig. 2). Incidentally, the present results obtained for P D G F receptor compare well with the biophysical studies that have been done on the receptor for EGF. When Balb 3T3 clone c/3 fibroblasts were grown in a similar way as in our experiment, it was found that D was 3 . 2 + 0 . 7 x ]0-1~ and R was 15 + 7 % for the EGFreceptor (Yarden et al., 1981). In order to further elucidate the dynamics of the P D G F receptor, studies are in progress to use fluorescent monoclonal antibodies to the PD G F receptor rather than using the active ligand itself for this purpose. In analogy with the present study the FRAP technique will be used for assessment of the lateral mobility, and image intensified digital video microscopy (Gustafsson et al., 1988) for characterization of receptor localisation. In conclusion, we found that starved fibroblasts with serum added for a short time, gave the highest values of D and R for both W G A and PDGF. In the PMNL, higher figures for D and R were observed for the PD G F receptor, than for the glycoconjugates labelled with WGA. PD G F by itself, did not appear to change the mobility characteristics of the receptors either in the fibroblasts or in the PMNL.
Lateral Diffusion of Membrane Receptors
73
ACKNOWLEDGEMENTS D r A l v a r o M a c i e i r a C o e l h o is g r a t e f u l l y t h a n k e d f o r v a l u a b l e d i s c u s s i o n a n d c o m m e n t s o n t h e m a n u s c r i p t , M r s A n i t a L 6 n n for e x p e r i m e n t a l a s s i s t a n c e a n d M s A n i t a L a r s s o n f o r s e c r e t a r i a l assistance. T h i s r e s e a r c h was s u p p o r t e d b y grants f r o m t h e S w e d i s h M e d i c a l R e s e a r c h C o u n c i l ( P r o j e c t N o s . 4486 a n d 6251), t h e S w e d i s h B o a r d for T e c h n i c a l D e v e l o p m e n t ( P r o j e c t N o . 8 7 - 0 0 2 4 8 ) , M a g n . B e r g v a l l s Stiftelse a n d S t i f t e l s e n L a r s H i e r t a s M i n n e .
REFERENCES Axelrod, D., Koppel D. E., Schlessinger, J., Elson, E. and Webb W. W. (1976) Biophys. J. 16:1055-1069. Bockus, B. J. and Stiles, C. D. (1984) Exp. Cell Res. 153:186-197. Bowen-Pope, D. F. and Ross, R. (1984) Clin. Endocrinol. Metabol. 1:191-201. Br A. (1968) Scand. J. Clin. Lab. Invest. Suppl. 97:77-89. Callahan, M., Cochran, B. H. and Stiles, C. D. (1985) In: Growth Factors in Biology and Medicine, Ciba Foundation Symposium 116: 87-97. Dahlgren, C., Rundqvist, I., Stendahl , O. and Magnusson, K. E..(1980) Cell Biophys. 2:253-267, Deuel, T. F., Senior, R. M., Huang, J. S. and Griffins, G_ L. (1982) J. Clin. Invest. 69:1046-1049. Ek, B. and Heldin, C. H. (1982) J. Biol. Chem. 257:10486-10492. Grotendorst, G. R., Sepp~i, H. E. J., Kleinman, H. K. and Martin, G. R. (1981) Proc. Natl. Acad. Sci. USA 78" 3669-3672. Gustafsson, M., Jacobson, K., Sundqvist, T. and Magnusson, K. E. (1988) Bioscience 88 Abstr: 120: Vol 1, Maim6, Sweden, April 17-20. Heldin, C. H., Wasteson,/~. and Westermark, B. (1985) Mol. Cell Endocrinol. 39:169-187. Jacobson, K., Derzko, Z., Wu, E. S., Hou, Y. and Poste, G. J. (1976) Supramolec. Struct. 5: 565-576. Johansson, B., Sundqvist, T. and Magnusson, K. E. (1987) Cell Biophys. 10:233-244. Magnusson, K. E., Sundqvist, T. and Johansson, B. (1987) Eur. Biophys. J. 15:141-45. Peters, R. (1981) Cell Biol. Int. Rep. 5:733-760. Peters, R., Peters, J., Tews, K. H. and B~ihr, W. (1974) Biophys. Biochim. Acta 367:282-294. Pledger, W. J., Stiles, C. D. and Antoniades, H. N. (1977) Proc. Natl. Acad. Sci. USA 74: 4481-4485. Rees, A. R., Gregoriou, M., Johnson, P. and Garland, P. (1984) E M B O J. 3:1843-1847. R6nnstrand, L., Beckmann, M. P., Faulders, B., (3stman, A., Ek, B. and Heldin, C. H. (1987) J. Biol. Chem. 7: 2929-2932. Schlessinger, J., Koppel, D. E., Axelrod, D., Jacobson, K., Webb, W. W. and Elson, E. L. (1976a) Proc. Natl. Acad. Sci. USA 73:2409-2413. Schlessinger, J., Elson, E. L., Webb, W. W. and Metzger, H. (1976b) Nature 264:550-552. Schlessinger, J., Axelrod, D., Koppel, D. E., Webb, W. W. and Elson, E. L. (1977) Science 195: 307-309. Schreiber, A. B., Liberman, T. A., Lax, I., Yarden, Y. and Schlessinger, J. (1983) J. Biol. Chem. 258: 846-853. Stiles, C. D. (1983) Ceil 33:653-655. Stiles, C. D., Capone, G. T. and Scher, C. D. (1979) Proc. Natl. Acad. Sci. USA 76:1279-1283. Westermark, B. and Heldin, C. H. (1985) J. Cell Physiol. 124:43-48. Westermark, B., Heidin, C. H. and Ek, B. (1983) In: Growth and maturation factors (Guroff, G., ed), Wiley, New York, pp. 73-114. Williams, L. T. (1983)J. Clin. Invest. 72:1759-1763. Yarden, Y. and Schlessinger, J. (1985) In: Growth Factors in Biology and Medicine, Ciba Foundation Symposium 116: 23-45. Yarden, Y., Gabbay, M. and Schlessinger, J. (1981) Biophys. Biochim. Acta 674:188-203. Zidovetzki, R., Yarden, Y., Schlessinger, J. and Tovin, T. M. (1981) Proc. Natl. Acad. Sci. USA 78: 6981-6985.