- Email: [email protected]
Morphological transformation, autonomous proliferation and colony formation by chicken heart mesenchymal cells infected with avian sarcoma, erythroblastosis and myelocytomatosis viruses
Life Sciences, Vol. 35, pp. 1157-1171 Printed in the U.S.A.
Pergamon Press
MORPHOLOGICAL TRANSFORMATION, AUTONOMOUS PROLIFERATION AND COLONY FORMATION BY CHICKEN HEART MESENCHYMAL CELLS INFECTED WITH AVIAN SARCOMA, ERYTHROBLASTOSIS AND MYELOCYTOMATOSIS VIRUSES Samuel D. Balk, Harriet S. Gunther, and Andrea Morisi Department of Pathology and Cancer Research Institute, New England Deaconess Hospital, Boston, MA 02215 and Department of Pathology, Harvard Medical School, Boston, MA (Received in final form July 3, 1984)
Summary Normal chicken heart mesenchymal cells at low density in monolayer culture in plasma-containing medium have a polygonal shape and are proliferatively quiescent. The combination of epidermal growth factor and insulin at hyperphysiological concentration, an insulin-like growth factor surrogate, causes these cells to assume a fusiform shape and to increase 40-fold in number during four days of incubation. These mitogenic hormones do not, however, induce normal chicken heart mesenchymal cells to form colonies in agarose suspension culture. Chicken heart mesenchymal cells infected with the Schmidt-Ruppin or Prague-A strains of Rous sarcoma virus or with the Fujinami or Y73 avian sarcoma viruses assume spindle and round shapes, increase 50-100 fold in number during four days of monolayer culture in the absence of mltogenic hormones and form macroscopic colonies during 3-4 days of agarose suspension culture. The autonomous (mitogenic hormone-independent) proliferation, in monolayer culture, of cells infected with temperature-sensitive transformation mutants of Rous sarcoma virus (tsNY68, tsNY72, tsLA24, tsLA29) is temperature-sensitive. Chicken heart mesenchymal ce[[s infected with avian erythroblastosis virus assume spindle shapes and proliferate in monolayer culture at a rate comparable to that of sarcoma virus-infected cells but do not, however, form colonies in agarose suspension culture. Cells infected with the myelocytomatosis virus MC29 assume stellate shapes and increase 18-fold in number during four days of monolayer culture. Cells infected with the myelocytomatosis virus MH2 assume fusiform shapes and increase fourfold in number during four days of monolayer culture. Neither MC29 nor MH2 renders chicken heart mesenchymal cells capable of colony formation in agarose suspension culture. Infection with avian leukosis viruses (RAV-I, RAV-2, RPL-42) or with transformationdefective mutants of Rous sarcoma virus (tdNYl05,107,109) does not affect the morphology or proliferative behavior of chicken heart mesenchymal cells. Monolayer culture of chicken heart mesenchymal cells in plasma-containing medium appears, therefore, to define the ability of onc genes of acute transforming avian retroviruses to induce autonomous (mitogenic hormone-independent) cell proliferation,
0024-3205/84 $3.00 + .00 Copyright (c) 1984 Pergamon Press Ltd.
1158
Avian Retroviruses
and Growth Autonomy
Vol. 35, No. Ii, 1984
the essential characteristic of neoplasia. The differences in transformed morphology and rates of autonomous proliferation between cells infected with different acute transforming retroviruses probababiy reflects differences in the modes of action of the transforming proteins encoded by the onc genes of the respective viruses. The ability of avian retroviral onc genes to induce autonomous cell proliferation in monolayer culture can be distinguished from their ability to induce colony fonnation in agarose suspension culture. Infection with Rous sarcoma virus or with myelocytomatosis virus MC29 does not render chicken heart mesenchymal cells immortal in culture. Autonomous cell proliferation is the essence and defining characteristic of neoplasia (I): While normal cells proliferate only in the presence of mitogenic hormones (or the absence of negative-feedback inhibitors), neoplastic cells proliferate in the absence of mitogenic stimulation. The cells of benign neoplasms manifest only autonomous proliferative behavior, while those of malignant neoplasms invade and/or metastasize as well. Current evidence indicates that neoplastic behavior is the result of the presence in cells of active onc genes, i. e. genes whose protein products trigger, in the absence of exogenous mitogenic hormones, critical steps in the mitogenic hormone-receptor cascade (2). The sis gene (v-sis) of simian sarcoma virus, for example, appears to encode a platelet-derived growth factor-like protein that is produced by infected cells in an autocrine manner (3). Similarly, the erb-b gene of avian erythroblastosis virus (AEV) appears to encode a truncated, constitutively active fonn of the epidermal growth factor receptor, or a closely related protein (4). The onc genes of tumor viruses appear to have arisen from proto-onc genes, their normal cellular homologues, during evolution and are transduced into cells upon infection (2). The onc genes of naturally-occurring human tumors appear to be generated by somatic mutation; their presence has been demonstrated by the experimental technique of transfection (5). An understanding of the phenomenology and mechanisms of on___~cgene function will be critical to an understanding of neoplasia. We have developed a monolayer culture system in which chicken heart mesenchymal cells, at low densities, are proliferatively quiescent in medium containing chicken plasma at 10% (6-8). The proliferative quiescence of normal chicken heart mesenchymal cells appears to be physiological, being based neither on serum-starvation nor on density-dependent inhibition. Chicken heart mesenchymal cells proliferate very actively in response to the hormone combination of epidermal growth factor (EGF) plus insulin-like growth factor-I (IGF-I), IGF-II or insulin at hyperphysiological concentration, an IGF-surrogate (9,10). Similarly, we have demonstrated that chicken heart mesenchymal cells proliferate very actively upon infection with Rous sarcoma virus (RSV). The proliferative quiescence of low-density normal (uninfected) chicken heart mesenchymal cells in plasma-containing medium contrasts markedly with the great proliferative activity that normal (uninfected) chick embryo fibroblasts, a standard cell type used in avian tumor virus studies, show under the same conditions (6). Chicken heart mesenchymal ceils, for this reason, are probably better suited for studies of autonomous proliferation than are chick embryo fibroblasts. Avian retroviruses (RNA-containlng tumor virus) may be of the acutely or slowly transforming types (11-13). Acutely transforming retroviruses contain an onc gene and induce neoplasms with a short latent period. The avian sarcoma viruses are acutely transforming agents and may be replicationcompetent (~. ~., Schmidt-Ruppin RSV, Prague A-RSV) or defective (~. ~., Fujinami or Y73 avian sarcoma viruses). The avian acute leukemia viruses
Vol. 35, No. Ii, 1984
Avian Retroviruses
and Growth Autonomy
1159
(~. ~., avian erythroblastosis virus, myelocytomatosis viruses MC29 and MH2) are replication-defective and induce leukemias, carcinomas or sarcomas. The slowly transforming avian retroviruses or avian leukosis viruses, on the other hand, lack an onc gene, induce neoplasms with a long latent period and are all replication-competent. Avian leukosis viruses serve as "helpers" for replication of defective acutely-transformlng retroviruses. Tumorigenesis, ~. ~. lymphomagenesis, by slowly-transforming retroviruses appears to involve a "promoter-insertion" (II) mechanism whereby reverse transcripts of avian leukosis virus genomes, with their promoter and enhancer sequences, are integrated into host cell chromosomes in proximity to proto-onc genes. We report here the effects of avian sarcoma viruses, acute defective leukemia viruses and avian leukosis viruses on chicken heart mesenchymal cells as regards morphology, proliferative behavior in monolayer culture and colony formation in agarose suspension culture. Avian and mammalian cells are generally capable of sustaining only 25-50 population doublings upon serial subculture in monolayer (14). After these doublings have occurred the cultures generally enter a phase of "senescence" during which proliferation ceases and the cells ultimately die. "Immortal" or "established" cell lines sometimes, however, arise from cultures of rodent or human embryo fibroblasts. Immortalization or establishment has been defined as the ability of cells to sustain an unlimited number of doublings upon serial subculture (14,15). Such immortal lines arise spontaneously from senescent cultures or as the result of transduction or transfection into newly established cultures of genes like mxc , adenovirus EIA or polyoma large T (15). Immortality in culture resulting from the activity of an immortalization gene appears often to be associated with tumorigenicity of mammalian cells (15). We have, therefore, attempted to immortalize chicken heart mesenchymal cells by infection with MC29 virus, which contains the prototypic myc immortalization oncogene, as well as by infection with Rous sarcoma virus. Materials and Methods Our methods and synthetic medium for monolayer culture of chicken heart mesenchymal cells and our method for preparation of heparinized, heatdefibrinogenated rooster plasma have been published (6,8). Primary cultures were prepared using synthetic medium with heparinized, heat-defibrinogenated rooster plasma at 5%. Ventricles from the hearts of two 8-12 week old SPF, Cofal-negative cockerels (Spafas, Inc., Norwalk, CT) were enzymatically dissociated (trypsin, 0.1%; collagenase 0.05%). The suspension of cells and tissue debris obtained by enzymatic dissociation was centrifuged and the pellet resuspended in plasma-containing medium and used to seed twenty 60-mm Falcon tissue culture dishes. After a three-hour attachment period in the incubator the cultures were washed four times with a physiological electrolyte solution and changed to fresh, plasma-containing medium with EGF at I00 ng/ml and insulin at 10 ~g/ml. During the three-hour attachment period, chicken heart mesenchymal cells attach to the culture dish while myocytes, capillary segments and debris do not. After washing there remains, therefore, an essentially pure population of chicken heart mesenchymal cells. Few, if any, macrophages are seen in the primary cultures. On the day following their preparation, two dishes from each set of primary cultures were inoculated with the following avian retroviruses: Schmidt-Ruppin A-RSV or its temperature-sensitive transformation mutants tsNY68 or tsNY72 (Dr. H. Hanafusa) or transformation-defective mutants tdNYl05, tdNYl07 or tdNYl09 (Dr. R. Erikson); Prague A-RSV (Dr. R. Erikson) or its temperature-sensitive transformation mutants tsLA24 or tsLA29 (Dr. M.
1160
Avian Retroviruses
and Growth Autonomy
Vol. 35, No. Ii, 1984
Bissell); Fujinami or Y73 avian sarcoma viruses (Dr. H. Hanafusa); avian erythroblastosis virus ES4 (Dr. C. Moscovici); MC29 and MH2 avian myelocytomatosis viruses (Dr. C. Moscovici); avian leukosis viruses RAV-I and RAV-2 (Dr. H. Hanafusa) and RPL-42 (Dr. L. Crittenden). Primary cultures inoculated with wild-type viruses were maintained at 42°C in a humidifiedair/5% CO~ environment (Napco 6300 Incubator), while cultures inoculated z with temperature-sensitive transformation mutants were incubated at 35 o C. After a six-hour adsorption period, with hourly agitation, the inocula were withdrawn and replaced with fresh, plasma-containing medium with EGF at i00 ng/ml and insulin at i0 ~g/ml. All cells in primary cultures were morphologically transformed within two days of inoculation with wild-type and temperature sensitive sarcoma viruses and with acute leukemia viruses. The levels of virus specific proteins in cultures inoculated with avian leukosis viruses or transformation-defective mutants of RSV were compared to the ~ v e l s in cultures infected with wild-type RSV by immune precipitation of S-methionine labelled cell extracts using serum from tumor-bearing rabbits (16). This test, performed by Dr. Ray Erikson, demonstrated that cultures inoculated with nontransforming viruses were infected. On the third day after their preparation, all primary cultures were changed to fresh, plasma-containing medium without EGF or insulin. On the fourth day after preparation, primary cultures of normal and retrovirus-infected chicken heart mesenchymal cells were subcultured for experiments into replicate secondary monolayer and agarose suspension cultures in 35-mm Falcon dishes. Normal chicken heart mesenchymal cells and those infected with myelocytomatosis viruses, lymphoid leukosis viruses and transformation-defective mutants of RSV were seeded into secondary monolayer cultures at 12 x l0 T cells per dish. Cells infected with wild-type and temperature-sensitive.mutant sarcoma viruses and with erythroblastosis virus 4 were seeded at 2 x i0 per dish. Experiments were begun on the day following subculture and were done with culture medium containing heparinized, heat-defibrinogenated rooster plasma at a concentration of 10%. Two milliliter of experimental medium was used per 35 mm monolayer culture dish and the medium was changed daily. Each experimental point on proliferation curves represents the mean of two culture dishes counted with a Coulter electronic cell counter. At most experimental points, neither dish varied by more than 5% from the mean. In no case did either dish vary from the mean by more than 10%. Photomicrography was done with a Leitz Diavert Microscope equipped with a Nikon AFX-II camera using Kodak Technical Pan film. Secondary agarose suspension cultures were prepared in 35-mm dishes and consisted of a one milliliter underlay of 0.9% Sea Plaque Low Melt Agarose (FMC Corporation, Rockland, ME), a one milliliter cell layer containing 0.34% agarose and a two milliliter fluid overlay. The underlay, cell layer and fluid overlay each contained plasma at 10%. Because the agarose suspension cultures contained a total volume of four milliliter of medium (one milliliter agarose underlay, one milliliter cell layer agarose, two milliliter fluid overlay), while the corresponding monolayer experimental cultures contained two milliliter of medium, the agarose suspension cultures were seeded with twice the number of cells as the corresponding monolayer cultures. The ratio of cells to medium (cell density) was, thereby, the same in suspension and monolayer cultures. The agarose underlay was prepared by combining nine parts of twiceconcentrated synthetic medium without calcium, nine parts of 2% agarose in 2.4 mM CaCI 2 and two parts of plasma. The cell layer was prepared by combining one part of 0.68% agarose in plasma-containing medium, prepared as above, with an equal part of cell suspension in plasma-containing medium.
Vol. 35, No. ii, 1984
Avian Retroviruses
and Growth Autonomy
1161
Both agarose layers had, thereby, the same physiological calcium concentration as does our synthetic medium (9). Where indicated, EGF and insulin at twice the desired final concentration were included in the two milliliter fluid overlay. Receptor-grade EGF was ~Irchased from Collaborative Research, Inc. (Lexington, MA) and crystalline bovine insulin was purchased fr~n Sigma. The fluid overlay of the agarose suspension cultures was changed daily. Experimental determinations of the lifespan, in monolayer culture, of hormone-stimulated normal chicken heart mesenchymal cells, MC29-infected cells and RSV-infected cells were done by carrying the respective cells through successive culture passages in 60-mm Falcon dishes. Plasmacontaining medium with EGF at i000 ng/ml and insulin at 10 ~g/ml was used to make normal and MC29-infected cells proliferate as fast as RSV-Infected cells. Population doublings per passage were calculated from the quotient of the number of cells yielded per dish and the number of cells with which the dish had been seeded. Because the efficiency of passage of normal and retrovirus-infected chicken heart mesenchymal cells approaches 100%, cell loss at subculture was neglected in the calculation of population doublings per passage. All experiments on proliferation in monolayer culture, colony fomnation in agarose suspension culture and cell llfespan in culture were repeated at least three times. Results As indicated by morphological transformation, primary cultures of chicken heart mesenchymal cells become quantitatively infected within two days of inoculation with avian retrovirus. These rigorously comparable populations of normal (uninfected) and tumor virus-infected cells were passaged into secondary cultures for experimental comparison of morphology, proliferative behavior in monolayer culture and colony formation in agarose suspension culture. Normal secondary chicken heart mesenchymal cells at low density in monolayer culture in medium containing heparinized, heat-defibrinogenated rooster plasma at 10% have a polygonal or fan shape and show distinct ruffling of the cytoplasm along one or more borders (Fig. I). These cells are prollferatively quiescent (Fig. 2) and do not form colonies in agarose suspension culture (Table I). In the presence of EGF and hyperphysiological insulin, an IGF surrogate, normal chicken heart mesenchymal cells assume a fusiform shape and increase 40-fold in number during four days of monolayer culture (Figs. 1,2). Hormone-stimulated normal chicken heart mesenchymal cells do not form colonies in agarose suspension culture (Table I). Chicken heart mesenchymal cells infected with the Schmidt-Ruppin and Prague A strains of RSV and those infected with the Fujinami avian sarcoma virus assume spindle and round shapes; those infected with the Y73 avian sarcoma virus assume spindle shapes (Fig. I). Cells infected with these viruses increase 50-100 fold in number during four days of monolayer culture in plasma-containing medium without added mitogenic hormones (Fig. 2) and form macroscopic colonies during 3-4 days of agarose suspension culture (Table I). Sarcoma virus-infected cells form colonies with an efficiency that approaches 100%, and these colonies range in diameter from 100-300 microns. Chicken heart mesenchymal cells were infected with the temperaturesensitive transformation mutants tsNY68 and tsNY72, derived from
1162
Avian Retroviruses
and Growth Autonomy
Vol. 35, No. ii, 1984
Normol Cells
Normal Cells, EGF + Insuh'n
SR- RSV
PrA-RSV
FSV
YTJ-ASV
AEV
MC29
~IH2
FIG. i Phase contrast photomicrographs of quiescent and mitogenic hormonestimulated chicken heart mesenchymal cells and of chicken heart mesenchymal cells infected with the following acutely transfomning avian retrovlruses: Schmidt-Ruppin Rous sarcoma virus (SR-RSV), Prague A-RSV (PrA-RSV), Fujinami avian sarcoma virus (FSV), Y73 avian sarcoma virus (Y73-ASV), avian erythroblastosis virus (AEV), myelocytomatosis virus MC29 and myelocytomatosis virus MH2. The cells were infected in primary culture and photographed after three days in secondary culture; X 120. The culture medium contained rooster plasma at 10%. EGF at i000 ng/ml and insulin at I0 ~g/ml, an IGF surrogate, were added to the medium for stimulation of normal cells.
Vol. 35, No. Ii, 1984
Avian Retroviruses and Growth Autonomy
1163
I000
j
SR- RSV PrA- RSV
i ¥73-Asv '--
Normal Cells, EGF + Insulin
FSV '~AEV
100 MH2
I
_-
.43
NormalCells
t01
0
I
I
2
4
INCUBATION, days FIG. 2 Autonomous (mitogenic hormone-independent) proliferation in monolayer culture of chicken heart mesenchymal cells infected with acutelytransforming avian retroviruses. The cells were infected in primary culture and subcultured for experiments into replicate secondary cultures. The culture medium contained rooster plasma at 10% and was changed daily. EGF at i000 ng/ml and insulin at i0 ~g/ml, an IGF surrogate, were added to the medium for stimulation of normal cells. Each point represents the mean of two dishes counted with a Coulter electronic cell counter.
1164
Avian Retroviruses
and Growth Autonomy
Vol. 35, No. ii, 1984
TABLE I Behavior in Agarose Suspension Culture of Chicken Heart Mesenchymal Cells Infected with Avian Sarcoma Viruses, Erythroblastosis Virus (AEV) and Myelocytomasis Viruses MC29 and MH2
Colony Formation Normal Cells Normal cells with EGF, i000 ng/ml and Insulin, I0 ~g/ml SR-RSV-infected
+
PrA-RSV-infected
+
Fujinami ASV-infected
+
Y73 ASV-infected
+
AEV-infected MC29-infected MH2-infected
Chicken heart mesenchymal cells were infected in primary culture and subcultured for experiments into replicate agarose suspension cultures in 35-mm dishes. The one ml agarose underlay (0.9%), one ml cell layer agarose (0.34%) and 2 ml fluid overlay all contained rooster plasma at 10%. Because the agarose suspension cultures contained twice the total volume of culture medium (4 ml/dish) as did the monolayer cultures (2 ml/dish), the suspension cultures were seeded with twice the number of cells as the corresponding experimental monolayer cultures. Agarose suspension cultures and monolayer cultures were, thereby, rigorously comparable in terms of plasma concentration, cell density (cells/vol. medium), etc. The fluid overlay was changed daily. Where indicated EGF and insulin were added to the fluid overlay at twice the desired final concentration. Sarcoma virus-infected chicken heart mesenchymal cells formed macroscopically visible colonies, with 100% efficiency, within 3-4 days. All suspension cultures were observed for a period of three weeks. EGF and insulin did not induce colony formation by erythroblastosis and myelocytomatosis virus-infected cells. Schmidt-Ruppin RSV, and the temperature-sensitive transformation mutants tsLA24 and tsLA29, derived from Prague A-RSV. While these cells show little proliferative activity at 42°C, the nonpermissive temperature for viral transformation, they increase 8-23 fold in number during four days of monolayer culture at 35°C, the permissive temperature for viral transformation (Fig. 3). At 350C, a subphysiological, albeit permissive temperature, cells infected with the mutant viruses proliferate at rates comparable to
Vol. 35, No. ii, 1984
Avian Retroviruses
and Growth Autonomy
1165
10 O0 420C /////~
/
,0
Normal Cells, EGF+ Insulin
35°C
PrA-RSV SR-RSV
///
100
/o
////
//
/J/
I
iI
Normal Cells, EGF + Insulin
7r-
/
/
//
L
p
/, j /
/
^ ~ //~
PrA - RSV IsNY68 tsLA29
1.
%
"4
Normal Cells IsLA24 tsNY68 tsLA29
10~ /
l-/- ~
. . . .
IIs
l=
/ IsNY72
11 0
I
I
2
4
I 0
2
4
INCUBATION, days FIG. 3 Temperature-sensitive proliferation in monolayer culture of chicken heart mesenchymal cells infected with temperature-sensitive mutants tsNY68 and tsNY72 of SR-RSV and tsLA24 and tsLA29 of PrA-RSV. Primary cultures were infected with t~nperature-sensitive mutants and wildtype viruses at 35°C, the temperature permissive for transformation. Experiments were done at 35°C and 42°C, the nonpermissive temperature. The culture medium contained rooster plasma at 10% and was changed daily. EGF at i000 ng/ml and insulin at I0 ~g/ml were added to the medium for stimulation of normal cells. Each point represents the mean of two dishes. cells infected with the parental wild-type viruses or uninfected cells stimulated with mitogenic hormones. Chicken heart mesenchymal cells infected with avian erythroblastosis virus (AEV) assume a spindle shape and form distinctive fascicles in monalayer culture (Fig. I). These cells proliferate at a rate comparable to that of sarcoma vlrus-infected cells (Fig. 2), but do not form colonies in agarose suspension culture (Table I). Cells infected with the myelocytomatosis virus MC29 assume a stellate shape and increase 18-fold in number during four days of monolayer culture (Figs. 1,2). Cells infected with the myelocytomatosis virus MH2 assume a fusiform shape and increase fourfold in number during four days of monolayer culture (Figs. 1,2). We have found that cells infected with MC29 or MH2
1166
Avian Retroviruses
and Growth Autonomy
Vol. 35, No. Ii, 1984
regularly manifes~ their optimal autonomous proliferative activity only when seeded at 12 x i0~/35-mm dish, while cells infected with sarcoma viruses or erythroblastosis virus manifest their optimal autonomous proliferative activity ~ e n seeded at the lower density of 2 x I0 /35-mm dish. Neither MC29 nor MH2 renders chicken heart mesenchymal cells capable of colony f o r m a t i o n in agarose suspension c u l t u r e (Table I ) . A v a r i e t y of growth factors, including platelet-derived growth factor, have been found to induce colony formation by SV40 transformed BALB/c-3T3 ceils in serum-free agarose suspension culture, while these same cells proliferated in monolayer
30 SR-RSV MC29 Norm01 Cells
/~/j/p/
20 _
IO --
I//m//v J:t/'/
J:f L 0
I
L
10
I_ 20
[
L 50
INCUBATION, days
FIG. 4 Cumulative population doublings in culture of MC29-infected, RSVinfected and hormone-stimulated nomnal chicken heart mesenchymal cells. The cells were infected in primary culture and serially subcultured in 60-mm dishes every three or four days in medium containing rooster plasma at 10%. Normal (uninfected) chicken heart mesenchymal cells were rendered proliferatlvely active and MC29-infected cells were caused caused to proliferate at a maximal rate by inclusion in the culture medium of EGF at i000 ng/ml and insulin at I0 ~g/ml, an IGF surrogate. Population doublings per passage were calculated from the quotient of the n~nber of cells yielded per culture dish and the number of cells with which the dish had been seeded. (The efficiency of subculture in the chicken heart mesenchymal cell system approaches 100%.) Culture media were changed daily.
Vol. 35, No. Ii, 1984
Avian Retroviruses
and Growth Autonomy
1167
culture in the absence of these growth factors (17). Epidermal growth factor and insulin do not, however, induce erythroblastosis or myelocytomatosis virus-infected cells to form colonies in our agarose suspension cultures. Infection with the avian leukosis viruses RAV-I, RAV-2 or RPL-42 or with transformation-defective mutants of RSV (tdNYl05,107,109) does not alter the morphology or proliferative behavior of chicken heart mesenchymal cells (data not shown). Immune precipitation tests for virus specific proteins, performed by Dr. Ray Erikson, demonstrated that cultures inoculated with these viruses were infected (see Materials and Methods). Tests for immortalization (establishment) of chicken heart mesenchymal cells by retrovirus infection yielded negative results: Hormone-stimulated normal cells, Rous sarcoma virus-infected cells and cells infected with the myelocytomatosis virus MC29 underwent senescence after approximately 25 population doublings in monolayer culture (Fig. 4). The same was true of cells doubly infected with RSV and MC29 (data not shown). Discussion Uninfected chicken heart mesenehymal cells have a polygonal or fan shape and show distinct ruffling of the cytoplasm along one or more borders (Fig. i). Chicken heart mesenchymal cells stimulated to proliferate with EGF and hyperphysiological insulin, an IGF surrogate, assume a fusiform shape. Chicken heart mesenchymal cells infected with Schmidt-Ruppin A-RSV, with Prague A-RSV or with FuJinami ASV assume spindle and round shapes like those described by Hanafusa (18) for Schmidt-Ruppin-infected chick embryo fibroblasts. Chicken heart mesenchymal cells infected with Y73 ASV assume predominantly spindle shapes. Chicken heart mesenchymal cells infected with avian erythroblastosis virus assume spindle shapes and form distinctive fascicles in monolayer culture. Similar changes have been described for chick embryo fibroblasts infected with this agent (19). Chicken heart mesenchymal cells infected with MC29 assume a stellate shape similar to that described for MC29-infected chick embryo fibroblasts (19), while infection with MH2 renders chicken heart mesenchymal cells fusiform (Fig. i). Normal chicken heart mesenchymal cells in monolayer culture, unlike chick embryo fibroblasts, are proliferatively quiescent at low densities in plasmacontaining medium (Fig. 2). Chicken heart mesenchymal cells proliferate only on addition of mitogenic hormones like EGF and insulin at hyperphysiological concentration, an IGF surrogate. Chicken heart mesenchymal cells infected with avian sarcoma viruses, with avian erythroblastosis virus or with avian myelocytomatosis viruses, on the other hand, proliferate autonomously, i. e. in the absence of mitogenic hormones. Cells infected with sarcoma viruses or with erythroblastosis virus increase 50-100 fold in number during four days of monolayer culture (generation time ~ 16 hrs.) while cells infected with the myelocytomatosis viruses MC29 or MH2 increase, respectively, 18-fold or fourfold during the same period (generation times ~ 24 hrs., 48 hrs.). The transforming proteins of the avian sarcoma viruses and of avian erythroblastosis virus associate with the inner aspect of cell membranes, while the transforming protein of MC29 associates with the nuclear matrix (2). The transforming proteins of the sarcoma viruses have tyrosine-specific kinase activity, like that of many protein kinases associated with hormone receptors (2). The protein product of erb-b, the principal onc gene of the erythroblastosis virus, may well represent a truncated, constitutively active
1168
Avian Retroviruses
and Growth Autonomy
Vol. 35, No. Ii, 1984
form of the EGF receptor (4). The sarcoma and erythroblastosis viruses appear, therefore, to transform cells by direct activation of or subrogation for functions normally activated by hormone-receptor interactions. While the mechanism of action of the transforming protein of MC29 is not yet known, its nuclear, rather than cell membrane site of action suggests &.transforming function that is considerably different from that of sarcoma and erythroblastosis virus transforming proteins. As already noted, avian sarcoma or avian erythroblastosis virus-infected chicken heart mesenchymal cells are transformed to a spindle and round morphology and increase 50-100 fold in number during four days of monolayer culture in plasma-containing medium. MC29-infected chicken heart mesenchymal cells, on the other hand, are transformed to a stellate morphology and increase 18-fold during the same period. These differences in transformed morphology and in rates of autonomous (hormone-independent) proliferation probably reflect differences in the modes of action of the transforming proteins coded for by the onc genes of the respective viruses. While the MC29 virus renders chicken heart mesenchymal cells stellate and causes them to increase 18-fold during four days of monolayer culture, infection of these cells with MH2 virus renders them fusiform and causes them to proliferate at a more modest (fourfold increase in four days), albeit significant, rate. The MH2 virus has two onc genes, the myc gene and the mht gene and regularly causes carcinomas as well as myelocytomatosis (20). The function of the protein product of the mht gene is not yet known, nor is it known whether the myc gene in MH2 has identical structure and activity as the myc gene in MC29. These differences in oncogene complement and/or activity may explain the differences in morphology and rate of autonomous proliferation of chicken heart mesenchymal cells infected with these two agents. A question has been raised as to whether transformation of hematopoietic cells with acute leukemia viruses, ~. ~. avian erythroblastosis virus, inhibits differentiation by causing autonomous proliferation or, vice versa, causes autonomous proliferation by inhibiting differentiation (21). Cultured chicken heart mesenchymal cells, unlike hematopoietic cells, are proliferatively quiescent and show no tendency to differentiate. Our observation that infection with erythroblastosis virus renders chicken heart mesenchymal cells proliferatively active suggests therefore that initiation of cell replication, rather than inhibition of differentiation, is responsible for oncogenesis by this agent. Temperature-sensitive mutants of RSV transform cells at 35-37°C, the temperature permissive for transformation, but not at 42°C, the nonpermissive temperature. Other laboratories have reported that chick embryo fibroblasts infected with tsNY68, a standard temperature-sensitive mutant of SRA-RSV, grow as tumors at 42°C on the chorioallantoic membrane of embryonated eggs (22) and that tsNY68 causes autonomous proliferation at 42°C of cultured chick embryo chondrocytes (23) and neuroretinal cells (24). These published reports that tsNY68 causes autonomous cell proliferation at a temperature (42°C) at which the mutant transforming protein p60 src has considerably reduced tyrosine-specific kinase activity form the basis for an extant hypothesis that a function of the src protein other than phosphorylation may be responsible for autonomous proliferation (24). Autonomous proliferation of chick embryo neuroretinal cells infected with the temperature-sensitive mutant tsNY72 has, on the other hand, been reported to be temperature sensitive (24). Unlike other groups, we have found that the proliferation of chicken heart mesenchymal cells infected with tsNY68 is temperature sensitive. Like
Vol. 35, No. ii, 1984
Avian Retroviruses
and Growth Autonomy
1169
other groups, we have found that the proliferation of chicken heart mesenchymal cells infected with tsNY72, tsLA24 or tsLA29 is temperature sensitive. The most straightforward explanation for the difference in temperature-sensitive proliferation between tsNY68-infected chicken heart mesenchymal cells and the tsNY68-infected cells studied by other groups may be in "tighter" behavior of this mutant in the present system. In any case, the temperature-sensitive proliferative behavior of chicken heart mesenchymal cells infected with tsNY68, as well as with tsNY72, tsLA24 and tsLA29 suggests that enzymatically active p60 src is necessary for autonomous (hormone-independent) proliferative behavior. The slowly oncogenic avian retroviruses (avian leukosis viruses), lacking one genes as they do, are thought to cause autonomous proliferation of infected cells by "promoter insertion" into host cell chromosomes (11-13): Host cell proto-onc genes appear to be activated by vicinal integration of avian leukosis proviruses with their regulatory LTR sequences. Such critical vicinal integration appears to require the passage of months in the animal host. Consistent with this view, we have found that infection with avian leukosis viruses (RAV-I, RAV-2, PRL-42) or with src-gene deletion mutants of RSV (transformation-defective mutants) does not alter the morphology or proliferative behavior of chicken heart mesenchymal cells in monolayer culture. These findings also serve as negative controls for any potential effects of the avian leukosis virus "helpers" that accompany the defective sarcoma (FSV, Y73-ASV) and defective acute leukemia viruses (AEV, MC29, MH2) used in our studies. While chicken heart mesenchymal cells infected with avian sarcoma viruses, erythroblastosis virus or the myelocytomatosis viruses MC29 and MH2 proliferate autonomously in monolayer culture, only the sarcoma virusinfected cells are capable of colony formation in agarose suspension culture. As noted in Materials and Methods, cell density, plasma concentration, calcium concentration and all other parameters are the same in our monolayer and agarose suspension cultures, making the two rigorously comparable. Erythroblastosis virus-infected cells, like sarcoma virus-infected cells, proliferate very rapidly in monolayer culture (50-100 fold increase in four days), but do not form colonies in agarose, indicating that rapid proliferation in mouolayer culture is not sine qua non for colony formation in suspension culture. Other investigators have reported that AEV and MC29-infected chick embryo fibroblasts do form colonies in an anchorageindependent manner, i. e. in suspension culture (19). This difference between the chicken heart mesenchymal cell system and the chick embryo fibroblast system may he a consequence of different properties of the respective cell types themselves or of differences in the composition of the culture media employed (plasma vs. serum, synthetic media, etc.). In any case, the chicken heart mesenchymal cell system can distinguish between the ability of avian retroviral onc genes to induce autonomous (mitogenic hormone-independent) cell proliferation in monolayer culture and their ability to induce colony formation in agarose suspension culture. Colony formation in suspension culture has not been found to be a general and specific index of the ability of cells to form tumors in vivo (25,26) or of the ability of ceils to grow as metastases (27-29). Because mitogenic hormone-independent (autonomous) proliferation is the essence and defining characteristic of neoplasia, the proliferation of retrovirus-infected chicken heart mesenchymal cells in plasma-containing medium in monolayer culture may represent a reliable and biologically meaningful in vitro index of the tumorigenicity of the onc genes of those retroviruses. It is interesting to note, in this connection, that autonomous proliferation of chicken heart
i170
Avian Retroviruses
mesenchymal cells in vitro onc gene, e. ~., src (RSV) embryo fib~oblasts appears genes, !. ~. ras plus myc
and Growth Autonomy
Vol. 35, No. ii, 1984
appears to require the transduction of a single or myc (MC29), while tumorigenicity in vivo of rat to require transfection of the cells with two onc (5,30).
Mammalian cells transfected with myc or with myc-related oncogenes have been found to be capable of sustaining an unlimited number of population doublings in monolayer culture, i. e. to be "immortalized" or "established" (15). We have found, on the other hand, that chicken heart mesenchymal cells infected with MC29 undergo senescence in monolayer culture, as do chicken heart mesenchymal cells infected with Rous sarcoma virus (Fig. 4). Other workers have, likewise, found that chick embryo fibroblasts infected with RSV undergo senescence (M. Weber, personal communication), as do chick embryo fibroblasts infected with MC29 (19,31), but that MC29-infected quail embryo fibroblasts are "immortalized" (31). "Nonimmortalized," RSV-infected chick embryo fibroblasts have been reported to form tumors upon transplantation into athymic nude mice, indicating that a capacity of cells to sustain an unlimited number of doublings may not be necessary for tumorigenicity (32). The relevance of the cell culture criterion of "immortalization" to oncogenicity of avian retroviruses is therefore uncertain. Acknowledsments This work was supported by Grant #1536 from the Council for Tobacco Research and by a grant from the Thoracic Foundation. The authors thank Dr. R. Erikson for his invaluable advice and other investigators for viral inocula, Dr. M. A. Legg, Mr. R. D. Pence and Mr. Joseph Flaherty for their support and encouragement and Mrs. Ruth Dziadul for expert preparation of the manuscript. This is reprint No. 752 from the Cancer Research Institute of the New England Deaconess HospitaL. References i. 2. 3. 4. 5. 6. 7. 8. 9. i0. 11. 12. 13. 14. 15. 16. 17. 18. 19.
A BRAUN, The Story of Cancer, Addison-Wesley, Reading, MA (1977). J. M. BISHOP, Ann. Rev. Biochem. 52 301-354 (1983). T. DUELL, J. HUANG, S. HUANG, P. STROOBANT and M. WATERFIELD, Science 221 1348-1350 (1983). J. DOWNWARD, Y. YARDEN, E. MAYES, G. SCRACE, N. TOTTY, P. STOCKWELL, A. ULLRICH, J. SCHLESSINGER and M. WATERFIELD, Nature 307 521-527 (1984). H. LAND, L. PARADA and R. WEINBERG, Science 222 771-777 (1983). S. BALK, Proc. Natl. Acad. Sci. USA 77 6606-6610 (1980). S. BALK, S. LEVINE, L. YOUNG, M. LA FLEUR and N. RAYMOND, Proc. Nat. Acad. Sci. USA 78 5656-5660 (1981). S. BALK, H. GUNTHER and A. MORISI, Life Sciences 34 803-808 (1984). S. BALK, R. SHIU, M. LA FLEUR and L. YOUNG, Proc. Natl. Acad. Sci. USA 79 1154-1157 (1982). S. BALK, A. MORISI, H. GUNTHER, M. SVOBODA, J. J. VAN WYK, S. P. NISSLEY, and C. G. SCANES, Life Sciences, in press. M. HAYMAN, J. Gen. Virol. 52 1-14 (1981). P. ENRIETTO and J. WYKE, Adv. Can. Res. 39 269-314 (1983). Y. FUNG, W. LEWIS, L. CRITTENDEN and If. KUNG, Cell 33 357-368 (1983). R. HOLLIDAY, Nature 306 742 (1983). J. CAIRNS and J. LOGAN, Nature 304 582-583 (1983). J. BRUGGE and R. ERIKSON, Nature 269 346-348 (1977). D. MC CLURE, Cell 32 999-1006 (1983). H. HANAFUSA, Proc. Natl. Acad. Sci. USA 63 318-322 (1969). B. ROYER-POKORA, H. BEUG, M. 6q.AVIEZ, H. WINKHARDT, R. FRIIS and T. CRAF, Cell 13 751-760 (1978).
Vol. 35, No. ii, 1984
Avian Retroviruses and Growth Autonomy
1171
20. N. KAN, C. FLORDELLIS, C. GARON, P. DUESBERG and T. PAPAS, Proc. Natl. Acad. Sci. USA 80 6566-6570 (1983). 21. T. GRAF and H. BEUG, Cell 34 7-9 (1983). 22. G. POSTE and M. FLOOD, Cell 17 789-800 (1979). 23. A. TANAKA, C. PARKER and A KAJI, J. Virol 35 531-541 (1980). 24. F. POIRIER, G. CALOTHY, R. KARESS, E. ERIKSON and H. HANAFUSA, J. Virol. 42 780-789 (1982). 25. C. STILES, W. DESMOND, L. CHUMAN, G. SATO and M. SAIER, Cancer Res. 36 3300-3305 (1976). 26. L. SMETS, Biochim. Biophys. Acta 605 93-111 (1980). 27. R. LOTAN and A. RAZ, Cancer Res. 43 2088-2093 (1983). 28. H. RUBIN, C. ROMERDAHL and B. CHU, J. Natl. Cancer Inst. 70 1087-1096 (1983). 29. M. DODSON, F. GELDER, J. SLOTA and C. LANGE, Int. J. Cancer 32 211-217 (1983). 30. H. LAND, L. PARADA and R. WEINBERG, Nature 304 596-602 (1983). 31. K. BISTER, M. HAYMAN and P. VOGT, Virology 8 2 431-448 (1977). 32. P. KAHN, K. NAKAMURA, S. SHIN, R. SMITH and M. WEBER, J. Virol. 42 602611 (1982).
Pergamon Press
MORPHOLOGICAL TRANSFORMATION, AUTONOMOUS PROLIFERATION AND COLONY FORMATION BY CHICKEN HEART MESENCHYMAL CELLS INFECTED WITH AVIAN SARCOMA, ERYTHROBLASTOSIS AND MYELOCYTOMATOSIS VIRUSES Samuel D. Balk, Harriet S. Gunther, and Andrea Morisi Department of Pathology and Cancer Research Institute, New England Deaconess Hospital, Boston, MA 02215 and Department of Pathology, Harvard Medical School, Boston, MA (Received in final form July 3, 1984)
Summary Normal chicken heart mesenchymal cells at low density in monolayer culture in plasma-containing medium have a polygonal shape and are proliferatively quiescent. The combination of epidermal growth factor and insulin at hyperphysiological concentration, an insulin-like growth factor surrogate, causes these cells to assume a fusiform shape and to increase 40-fold in number during four days of incubation. These mitogenic hormones do not, however, induce normal chicken heart mesenchymal cells to form colonies in agarose suspension culture. Chicken heart mesenchymal cells infected with the Schmidt-Ruppin or Prague-A strains of Rous sarcoma virus or with the Fujinami or Y73 avian sarcoma viruses assume spindle and round shapes, increase 50-100 fold in number during four days of monolayer culture in the absence of mltogenic hormones and form macroscopic colonies during 3-4 days of agarose suspension culture. The autonomous (mitogenic hormone-independent) proliferation, in monolayer culture, of cells infected with temperature-sensitive transformation mutants of Rous sarcoma virus (tsNY68, tsNY72, tsLA24, tsLA29) is temperature-sensitive. Chicken heart mesenchymal ce[[s infected with avian erythroblastosis virus assume spindle shapes and proliferate in monolayer culture at a rate comparable to that of sarcoma virus-infected cells but do not, however, form colonies in agarose suspension culture. Cells infected with the myelocytomatosis virus MC29 assume stellate shapes and increase 18-fold in number during four days of monolayer culture. Cells infected with the myelocytomatosis virus MH2 assume fusiform shapes and increase fourfold in number during four days of monolayer culture. Neither MC29 nor MH2 renders chicken heart mesenchymal cells capable of colony formation in agarose suspension culture. Infection with avian leukosis viruses (RAV-I, RAV-2, RPL-42) or with transformationdefective mutants of Rous sarcoma virus (tdNYl05,107,109) does not affect the morphology or proliferative behavior of chicken heart mesenchymal cells. Monolayer culture of chicken heart mesenchymal cells in plasma-containing medium appears, therefore, to define the ability of onc genes of acute transforming avian retroviruses to induce autonomous (mitogenic hormone-independent) cell proliferation,
0024-3205/84 $3.00 + .00 Copyright (c) 1984 Pergamon Press Ltd.
1158
Avian Retroviruses
and Growth Autonomy
Vol. 35, No. Ii, 1984
the essential characteristic of neoplasia. The differences in transformed morphology and rates of autonomous proliferation between cells infected with different acute transforming retroviruses probababiy reflects differences in the modes of action of the transforming proteins encoded by the onc genes of the respective viruses. The ability of avian retroviral onc genes to induce autonomous cell proliferation in monolayer culture can be distinguished from their ability to induce colony fonnation in agarose suspension culture. Infection with Rous sarcoma virus or with myelocytomatosis virus MC29 does not render chicken heart mesenchymal cells immortal in culture. Autonomous cell proliferation is the essence and defining characteristic of neoplasia (I): While normal cells proliferate only in the presence of mitogenic hormones (or the absence of negative-feedback inhibitors), neoplastic cells proliferate in the absence of mitogenic stimulation. The cells of benign neoplasms manifest only autonomous proliferative behavior, while those of malignant neoplasms invade and/or metastasize as well. Current evidence indicates that neoplastic behavior is the result of the presence in cells of active onc genes, i. e. genes whose protein products trigger, in the absence of exogenous mitogenic hormones, critical steps in the mitogenic hormone-receptor cascade (2). The sis gene (v-sis) of simian sarcoma virus, for example, appears to encode a platelet-derived growth factor-like protein that is produced by infected cells in an autocrine manner (3). Similarly, the erb-b gene of avian erythroblastosis virus (AEV) appears to encode a truncated, constitutively active fonn of the epidermal growth factor receptor, or a closely related protein (4). The onc genes of tumor viruses appear to have arisen from proto-onc genes, their normal cellular homologues, during evolution and are transduced into cells upon infection (2). The onc genes of naturally-occurring human tumors appear to be generated by somatic mutation; their presence has been demonstrated by the experimental technique of transfection (5). An understanding of the phenomenology and mechanisms of on___~cgene function will be critical to an understanding of neoplasia. We have developed a monolayer culture system in which chicken heart mesenchymal cells, at low densities, are proliferatively quiescent in medium containing chicken plasma at 10% (6-8). The proliferative quiescence of normal chicken heart mesenchymal cells appears to be physiological, being based neither on serum-starvation nor on density-dependent inhibition. Chicken heart mesenchymal cells proliferate very actively in response to the hormone combination of epidermal growth factor (EGF) plus insulin-like growth factor-I (IGF-I), IGF-II or insulin at hyperphysiological concentration, an IGF-surrogate (9,10). Similarly, we have demonstrated that chicken heart mesenchymal cells proliferate very actively upon infection with Rous sarcoma virus (RSV). The proliferative quiescence of low-density normal (uninfected) chicken heart mesenchymal cells in plasma-containing medium contrasts markedly with the great proliferative activity that normal (uninfected) chick embryo fibroblasts, a standard cell type used in avian tumor virus studies, show under the same conditions (6). Chicken heart mesenchymal ceils, for this reason, are probably better suited for studies of autonomous proliferation than are chick embryo fibroblasts. Avian retroviruses (RNA-containlng tumor virus) may be of the acutely or slowly transforming types (11-13). Acutely transforming retroviruses contain an onc gene and induce neoplasms with a short latent period. The avian sarcoma viruses are acutely transforming agents and may be replicationcompetent (~. ~., Schmidt-Ruppin RSV, Prague A-RSV) or defective (~. ~., Fujinami or Y73 avian sarcoma viruses). The avian acute leukemia viruses
Vol. 35, No. Ii, 1984
Avian Retroviruses
and Growth Autonomy
1159
(~. ~., avian erythroblastosis virus, myelocytomatosis viruses MC29 and MH2) are replication-defective and induce leukemias, carcinomas or sarcomas. The slowly transforming avian retroviruses or avian leukosis viruses, on the other hand, lack an onc gene, induce neoplasms with a long latent period and are all replication-competent. Avian leukosis viruses serve as "helpers" for replication of defective acutely-transformlng retroviruses. Tumorigenesis, ~. ~. lymphomagenesis, by slowly-transforming retroviruses appears to involve a "promoter-insertion" (II) mechanism whereby reverse transcripts of avian leukosis virus genomes, with their promoter and enhancer sequences, are integrated into host cell chromosomes in proximity to proto-onc genes. We report here the effects of avian sarcoma viruses, acute defective leukemia viruses and avian leukosis viruses on chicken heart mesenchymal cells as regards morphology, proliferative behavior in monolayer culture and colony formation in agarose suspension culture. Avian and mammalian cells are generally capable of sustaining only 25-50 population doublings upon serial subculture in monolayer (14). After these doublings have occurred the cultures generally enter a phase of "senescence" during which proliferation ceases and the cells ultimately die. "Immortal" or "established" cell lines sometimes, however, arise from cultures of rodent or human embryo fibroblasts. Immortalization or establishment has been defined as the ability of cells to sustain an unlimited number of doublings upon serial subculture (14,15). Such immortal lines arise spontaneously from senescent cultures or as the result of transduction or transfection into newly established cultures of genes like mxc , adenovirus EIA or polyoma large T (15). Immortality in culture resulting from the activity of an immortalization gene appears often to be associated with tumorigenicity of mammalian cells (15). We have, therefore, attempted to immortalize chicken heart mesenchymal cells by infection with MC29 virus, which contains the prototypic myc immortalization oncogene, as well as by infection with Rous sarcoma virus. Materials and Methods Our methods and synthetic medium for monolayer culture of chicken heart mesenchymal cells and our method for preparation of heparinized, heatdefibrinogenated rooster plasma have been published (6,8). Primary cultures were prepared using synthetic medium with heparinized, heat-defibrinogenated rooster plasma at 5%. Ventricles from the hearts of two 8-12 week old SPF, Cofal-negative cockerels (Spafas, Inc., Norwalk, CT) were enzymatically dissociated (trypsin, 0.1%; collagenase 0.05%). The suspension of cells and tissue debris obtained by enzymatic dissociation was centrifuged and the pellet resuspended in plasma-containing medium and used to seed twenty 60-mm Falcon tissue culture dishes. After a three-hour attachment period in the incubator the cultures were washed four times with a physiological electrolyte solution and changed to fresh, plasma-containing medium with EGF at I00 ng/ml and insulin at 10 ~g/ml. During the three-hour attachment period, chicken heart mesenchymal cells attach to the culture dish while myocytes, capillary segments and debris do not. After washing there remains, therefore, an essentially pure population of chicken heart mesenchymal cells. Few, if any, macrophages are seen in the primary cultures. On the day following their preparation, two dishes from each set of primary cultures were inoculated with the following avian retroviruses: Schmidt-Ruppin A-RSV or its temperature-sensitive transformation mutants tsNY68 or tsNY72 (Dr. H. Hanafusa) or transformation-defective mutants tdNYl05, tdNYl07 or tdNYl09 (Dr. R. Erikson); Prague A-RSV (Dr. R. Erikson) or its temperature-sensitive transformation mutants tsLA24 or tsLA29 (Dr. M.
1160
Avian Retroviruses
and Growth Autonomy
Vol. 35, No. Ii, 1984
Bissell); Fujinami or Y73 avian sarcoma viruses (Dr. H. Hanafusa); avian erythroblastosis virus ES4 (Dr. C. Moscovici); MC29 and MH2 avian myelocytomatosis viruses (Dr. C. Moscovici); avian leukosis viruses RAV-I and RAV-2 (Dr. H. Hanafusa) and RPL-42 (Dr. L. Crittenden). Primary cultures inoculated with wild-type viruses were maintained at 42°C in a humidifiedair/5% CO~ environment (Napco 6300 Incubator), while cultures inoculated z with temperature-sensitive transformation mutants were incubated at 35 o C. After a six-hour adsorption period, with hourly agitation, the inocula were withdrawn and replaced with fresh, plasma-containing medium with EGF at i00 ng/ml and insulin at i0 ~g/ml. All cells in primary cultures were morphologically transformed within two days of inoculation with wild-type and temperature sensitive sarcoma viruses and with acute leukemia viruses. The levels of virus specific proteins in cultures inoculated with avian leukosis viruses or transformation-defective mutants of RSV were compared to the ~ v e l s in cultures infected with wild-type RSV by immune precipitation of S-methionine labelled cell extracts using serum from tumor-bearing rabbits (16). This test, performed by Dr. Ray Erikson, demonstrated that cultures inoculated with nontransforming viruses were infected. On the third day after their preparation, all primary cultures were changed to fresh, plasma-containing medium without EGF or insulin. On the fourth day after preparation, primary cultures of normal and retrovirus-infected chicken heart mesenchymal cells were subcultured for experiments into replicate secondary monolayer and agarose suspension cultures in 35-mm Falcon dishes. Normal chicken heart mesenchymal cells and those infected with myelocytomatosis viruses, lymphoid leukosis viruses and transformation-defective mutants of RSV were seeded into secondary monolayer cultures at 12 x l0 T cells per dish. Cells infected with wild-type and temperature-sensitive.mutant sarcoma viruses and with erythroblastosis virus 4 were seeded at 2 x i0 per dish. Experiments were begun on the day following subculture and were done with culture medium containing heparinized, heat-defibrinogenated rooster plasma at a concentration of 10%. Two milliliter of experimental medium was used per 35 mm monolayer culture dish and the medium was changed daily. Each experimental point on proliferation curves represents the mean of two culture dishes counted with a Coulter electronic cell counter. At most experimental points, neither dish varied by more than 5% from the mean. In no case did either dish vary from the mean by more than 10%. Photomicrography was done with a Leitz Diavert Microscope equipped with a Nikon AFX-II camera using Kodak Technical Pan film. Secondary agarose suspension cultures were prepared in 35-mm dishes and consisted of a one milliliter underlay of 0.9% Sea Plaque Low Melt Agarose (FMC Corporation, Rockland, ME), a one milliliter cell layer containing 0.34% agarose and a two milliliter fluid overlay. The underlay, cell layer and fluid overlay each contained plasma at 10%. Because the agarose suspension cultures contained a total volume of four milliliter of medium (one milliliter agarose underlay, one milliliter cell layer agarose, two milliliter fluid overlay), while the corresponding monolayer experimental cultures contained two milliliter of medium, the agarose suspension cultures were seeded with twice the number of cells as the corresponding monolayer cultures. The ratio of cells to medium (cell density) was, thereby, the same in suspension and monolayer cultures. The agarose underlay was prepared by combining nine parts of twiceconcentrated synthetic medium without calcium, nine parts of 2% agarose in 2.4 mM CaCI 2 and two parts of plasma. The cell layer was prepared by combining one part of 0.68% agarose in plasma-containing medium, prepared as above, with an equal part of cell suspension in plasma-containing medium.
Vol. 35, No. ii, 1984
Avian Retroviruses
and Growth Autonomy
1161
Both agarose layers had, thereby, the same physiological calcium concentration as does our synthetic medium (9). Where indicated, EGF and insulin at twice the desired final concentration were included in the two milliliter fluid overlay. Receptor-grade EGF was ~Irchased from Collaborative Research, Inc. (Lexington, MA) and crystalline bovine insulin was purchased fr~n Sigma. The fluid overlay of the agarose suspension cultures was changed daily. Experimental determinations of the lifespan, in monolayer culture, of hormone-stimulated normal chicken heart mesenchymal cells, MC29-infected cells and RSV-infected cells were done by carrying the respective cells through successive culture passages in 60-mm Falcon dishes. Plasmacontaining medium with EGF at i000 ng/ml and insulin at 10 ~g/ml was used to make normal and MC29-infected cells proliferate as fast as RSV-Infected cells. Population doublings per passage were calculated from the quotient of the number of cells yielded per dish and the number of cells with which the dish had been seeded. Because the efficiency of passage of normal and retrovirus-infected chicken heart mesenchymal cells approaches 100%, cell loss at subculture was neglected in the calculation of population doublings per passage. All experiments on proliferation in monolayer culture, colony fomnation in agarose suspension culture and cell llfespan in culture were repeated at least three times. Results As indicated by morphological transformation, primary cultures of chicken heart mesenchymal cells become quantitatively infected within two days of inoculation with avian retrovirus. These rigorously comparable populations of normal (uninfected) and tumor virus-infected cells were passaged into secondary cultures for experimental comparison of morphology, proliferative behavior in monolayer culture and colony formation in agarose suspension culture. Normal secondary chicken heart mesenchymal cells at low density in monolayer culture in medium containing heparinized, heat-defibrinogenated rooster plasma at 10% have a polygonal or fan shape and show distinct ruffling of the cytoplasm along one or more borders (Fig. I). These cells are prollferatively quiescent (Fig. 2) and do not form colonies in agarose suspension culture (Table I). In the presence of EGF and hyperphysiological insulin, an IGF surrogate, normal chicken heart mesenchymal cells assume a fusiform shape and increase 40-fold in number during four days of monolayer culture (Figs. 1,2). Hormone-stimulated normal chicken heart mesenchymal cells do not form colonies in agarose suspension culture (Table I). Chicken heart mesenchymal cells infected with the Schmidt-Ruppin and Prague A strains of RSV and those infected with the Fujinami avian sarcoma virus assume spindle and round shapes; those infected with the Y73 avian sarcoma virus assume spindle shapes (Fig. I). Cells infected with these viruses increase 50-100 fold in number during four days of monolayer culture in plasma-containing medium without added mitogenic hormones (Fig. 2) and form macroscopic colonies during 3-4 days of agarose suspension culture (Table I). Sarcoma virus-infected cells form colonies with an efficiency that approaches 100%, and these colonies range in diameter from 100-300 microns. Chicken heart mesenchymal cells were infected with the temperaturesensitive transformation mutants tsNY68 and tsNY72, derived from
1162
Avian Retroviruses
and Growth Autonomy
Vol. 35, No. ii, 1984
Normol Cells
Normal Cells, EGF + Insuh'n
SR- RSV
PrA-RSV
FSV
YTJ-ASV
AEV
MC29
~IH2
FIG. i Phase contrast photomicrographs of quiescent and mitogenic hormonestimulated chicken heart mesenchymal cells and of chicken heart mesenchymal cells infected with the following acutely transfomning avian retrovlruses: Schmidt-Ruppin Rous sarcoma virus (SR-RSV), Prague A-RSV (PrA-RSV), Fujinami avian sarcoma virus (FSV), Y73 avian sarcoma virus (Y73-ASV), avian erythroblastosis virus (AEV), myelocytomatosis virus MC29 and myelocytomatosis virus MH2. The cells were infected in primary culture and photographed after three days in secondary culture; X 120. The culture medium contained rooster plasma at 10%. EGF at i000 ng/ml and insulin at I0 ~g/ml, an IGF surrogate, were added to the medium for stimulation of normal cells.
Vol. 35, No. Ii, 1984
Avian Retroviruses and Growth Autonomy
1163
I000
j
SR- RSV PrA- RSV
i ¥73-Asv '--
Normal Cells, EGF + Insulin
FSV '~AEV
100 MH2
I
_-
.43
NormalCells
t01
0
I
I
2
4
INCUBATION, days FIG. 2 Autonomous (mitogenic hormone-independent) proliferation in monolayer culture of chicken heart mesenchymal cells infected with acutelytransforming avian retroviruses. The cells were infected in primary culture and subcultured for experiments into replicate secondary cultures. The culture medium contained rooster plasma at 10% and was changed daily. EGF at i000 ng/ml and insulin at i0 ~g/ml, an IGF surrogate, were added to the medium for stimulation of normal cells. Each point represents the mean of two dishes counted with a Coulter electronic cell counter.
1164
Avian Retroviruses
and Growth Autonomy
Vol. 35, No. ii, 1984
TABLE I Behavior in Agarose Suspension Culture of Chicken Heart Mesenchymal Cells Infected with Avian Sarcoma Viruses, Erythroblastosis Virus (AEV) and Myelocytomasis Viruses MC29 and MH2
Colony Formation Normal Cells Normal cells with EGF, i000 ng/ml and Insulin, I0 ~g/ml SR-RSV-infected
+
PrA-RSV-infected
+
Fujinami ASV-infected
+
Y73 ASV-infected
+
AEV-infected MC29-infected MH2-infected
Chicken heart mesenchymal cells were infected in primary culture and subcultured for experiments into replicate agarose suspension cultures in 35-mm dishes. The one ml agarose underlay (0.9%), one ml cell layer agarose (0.34%) and 2 ml fluid overlay all contained rooster plasma at 10%. Because the agarose suspension cultures contained twice the total volume of culture medium (4 ml/dish) as did the monolayer cultures (2 ml/dish), the suspension cultures were seeded with twice the number of cells as the corresponding experimental monolayer cultures. Agarose suspension cultures and monolayer cultures were, thereby, rigorously comparable in terms of plasma concentration, cell density (cells/vol. medium), etc. The fluid overlay was changed daily. Where indicated EGF and insulin were added to the fluid overlay at twice the desired final concentration. Sarcoma virus-infected chicken heart mesenchymal cells formed macroscopically visible colonies, with 100% efficiency, within 3-4 days. All suspension cultures were observed for a period of three weeks. EGF and insulin did not induce colony formation by erythroblastosis and myelocytomatosis virus-infected cells. Schmidt-Ruppin RSV, and the temperature-sensitive transformation mutants tsLA24 and tsLA29, derived from Prague A-RSV. While these cells show little proliferative activity at 42°C, the nonpermissive temperature for viral transformation, they increase 8-23 fold in number during four days of monolayer culture at 35°C, the permissive temperature for viral transformation (Fig. 3). At 350C, a subphysiological, albeit permissive temperature, cells infected with the mutant viruses proliferate at rates comparable to
Vol. 35, No. ii, 1984
Avian Retroviruses
and Growth Autonomy
1165
10 O0 420C /////~
/
,0
Normal Cells, EGF+ Insulin
35°C
PrA-RSV SR-RSV
///
100
/o
////
//
/J/
I
iI
Normal Cells, EGF + Insulin
7r-
/
/
//
L
p
/, j /
/
^ ~ //~
PrA - RSV IsNY68 tsLA29
1.
%
"4
Normal Cells IsLA24 tsNY68 tsLA29
10~ /
l-/- ~
. . . .
IIs
l=
/ IsNY72
11 0
I
I
2
4
I 0
2
4
INCUBATION, days FIG. 3 Temperature-sensitive proliferation in monolayer culture of chicken heart mesenchymal cells infected with temperature-sensitive mutants tsNY68 and tsNY72 of SR-RSV and tsLA24 and tsLA29 of PrA-RSV. Primary cultures were infected with t~nperature-sensitive mutants and wildtype viruses at 35°C, the temperature permissive for transformation. Experiments were done at 35°C and 42°C, the nonpermissive temperature. The culture medium contained rooster plasma at 10% and was changed daily. EGF at i000 ng/ml and insulin at I0 ~g/ml were added to the medium for stimulation of normal cells. Each point represents the mean of two dishes. cells infected with the parental wild-type viruses or uninfected cells stimulated with mitogenic hormones. Chicken heart mesenchymal cells infected with avian erythroblastosis virus (AEV) assume a spindle shape and form distinctive fascicles in monalayer culture (Fig. I). These cells proliferate at a rate comparable to that of sarcoma vlrus-infected cells (Fig. 2), but do not form colonies in agarose suspension culture (Table I). Cells infected with the myelocytomatosis virus MC29 assume a stellate shape and increase 18-fold in number during four days of monolayer culture (Figs. 1,2). Cells infected with the myelocytomatosis virus MH2 assume a fusiform shape and increase fourfold in number during four days of monolayer culture (Figs. 1,2). We have found that cells infected with MC29 or MH2
1166
Avian Retroviruses
and Growth Autonomy
Vol. 35, No. Ii, 1984
regularly manifes~ their optimal autonomous proliferative activity only when seeded at 12 x i0~/35-mm dish, while cells infected with sarcoma viruses or erythroblastosis virus manifest their optimal autonomous proliferative activity ~ e n seeded at the lower density of 2 x I0 /35-mm dish. Neither MC29 nor MH2 renders chicken heart mesenchymal cells capable of colony f o r m a t i o n in agarose suspension c u l t u r e (Table I ) . A v a r i e t y of growth factors, including platelet-derived growth factor, have been found to induce colony formation by SV40 transformed BALB/c-3T3 ceils in serum-free agarose suspension culture, while these same cells proliferated in monolayer
30 SR-RSV MC29 Norm01 Cells
/~/j/p/
20 _
IO --
I//m//v J:t/'/
J:f L 0
I
L
10
I_ 20
[
L 50
INCUBATION, days
FIG. 4 Cumulative population doublings in culture of MC29-infected, RSVinfected and hormone-stimulated nomnal chicken heart mesenchymal cells. The cells were infected in primary culture and serially subcultured in 60-mm dishes every three or four days in medium containing rooster plasma at 10%. Normal (uninfected) chicken heart mesenchymal cells were rendered proliferatlvely active and MC29-infected cells were caused caused to proliferate at a maximal rate by inclusion in the culture medium of EGF at i000 ng/ml and insulin at I0 ~g/ml, an IGF surrogate. Population doublings per passage were calculated from the quotient of the n~nber of cells yielded per culture dish and the number of cells with which the dish had been seeded. (The efficiency of subculture in the chicken heart mesenchymal cell system approaches 100%.) Culture media were changed daily.
Vol. 35, No. Ii, 1984
Avian Retroviruses
and Growth Autonomy
1167
culture in the absence of these growth factors (17). Epidermal growth factor and insulin do not, however, induce erythroblastosis or myelocytomatosis virus-infected cells to form colonies in our agarose suspension cultures. Infection with the avian leukosis viruses RAV-I, RAV-2 or RPL-42 or with transformation-defective mutants of RSV (tdNYl05,107,109) does not alter the morphology or proliferative behavior of chicken heart mesenchymal cells (data not shown). Immune precipitation tests for virus specific proteins, performed by Dr. Ray Erikson, demonstrated that cultures inoculated with these viruses were infected (see Materials and Methods). Tests for immortalization (establishment) of chicken heart mesenchymal cells by retrovirus infection yielded negative results: Hormone-stimulated normal cells, Rous sarcoma virus-infected cells and cells infected with the myelocytomatosis virus MC29 underwent senescence after approximately 25 population doublings in monolayer culture (Fig. 4). The same was true of cells doubly infected with RSV and MC29 (data not shown). Discussion Uninfected chicken heart mesenehymal cells have a polygonal or fan shape and show distinct ruffling of the cytoplasm along one or more borders (Fig. i). Chicken heart mesenchymal cells stimulated to proliferate with EGF and hyperphysiological insulin, an IGF surrogate, assume a fusiform shape. Chicken heart mesenchymal cells infected with Schmidt-Ruppin A-RSV, with Prague A-RSV or with FuJinami ASV assume spindle and round shapes like those described by Hanafusa (18) for Schmidt-Ruppin-infected chick embryo fibroblasts. Chicken heart mesenchymal cells infected with Y73 ASV assume predominantly spindle shapes. Chicken heart mesenchymal cells infected with avian erythroblastosis virus assume spindle shapes and form distinctive fascicles in monolayer culture. Similar changes have been described for chick embryo fibroblasts infected with this agent (19). Chicken heart mesenchymal cells infected with MC29 assume a stellate shape similar to that described for MC29-infected chick embryo fibroblasts (19), while infection with MH2 renders chicken heart mesenchymal cells fusiform (Fig. i). Normal chicken heart mesenchymal cells in monolayer culture, unlike chick embryo fibroblasts, are proliferatively quiescent at low densities in plasmacontaining medium (Fig. 2). Chicken heart mesenchymal cells proliferate only on addition of mitogenic hormones like EGF and insulin at hyperphysiological concentration, an IGF surrogate. Chicken heart mesenchymal cells infected with avian sarcoma viruses, with avian erythroblastosis virus or with avian myelocytomatosis viruses, on the other hand, proliferate autonomously, i. e. in the absence of mitogenic hormones. Cells infected with sarcoma viruses or with erythroblastosis virus increase 50-100 fold in number during four days of monolayer culture (generation time ~ 16 hrs.) while cells infected with the myelocytomatosis viruses MC29 or MH2 increase, respectively, 18-fold or fourfold during the same period (generation times ~ 24 hrs., 48 hrs.). The transforming proteins of the avian sarcoma viruses and of avian erythroblastosis virus associate with the inner aspect of cell membranes, while the transforming protein of MC29 associates with the nuclear matrix (2). The transforming proteins of the sarcoma viruses have tyrosine-specific kinase activity, like that of many protein kinases associated with hormone receptors (2). The protein product of erb-b, the principal onc gene of the erythroblastosis virus, may well represent a truncated, constitutively active
1168
Avian Retroviruses
and Growth Autonomy
Vol. 35, No. Ii, 1984
form of the EGF receptor (4). The sarcoma and erythroblastosis viruses appear, therefore, to transform cells by direct activation of or subrogation for functions normally activated by hormone-receptor interactions. While the mechanism of action of the transforming protein of MC29 is not yet known, its nuclear, rather than cell membrane site of action suggests &.transforming function that is considerably different from that of sarcoma and erythroblastosis virus transforming proteins. As already noted, avian sarcoma or avian erythroblastosis virus-infected chicken heart mesenchymal cells are transformed to a spindle and round morphology and increase 50-100 fold in number during four days of monolayer culture in plasma-containing medium. MC29-infected chicken heart mesenchymal cells, on the other hand, are transformed to a stellate morphology and increase 18-fold during the same period. These differences in transformed morphology and in rates of autonomous (hormone-independent) proliferation probably reflect differences in the modes of action of the transforming proteins coded for by the onc genes of the respective viruses. While the MC29 virus renders chicken heart mesenchymal cells stellate and causes them to increase 18-fold during four days of monolayer culture, infection of these cells with MH2 virus renders them fusiform and causes them to proliferate at a more modest (fourfold increase in four days), albeit significant, rate. The MH2 virus has two onc genes, the myc gene and the mht gene and regularly causes carcinomas as well as myelocytomatosis (20). The function of the protein product of the mht gene is not yet known, nor is it known whether the myc gene in MH2 has identical structure and activity as the myc gene in MC29. These differences in oncogene complement and/or activity may explain the differences in morphology and rate of autonomous proliferation of chicken heart mesenchymal cells infected with these two agents. A question has been raised as to whether transformation of hematopoietic cells with acute leukemia viruses, ~. ~. avian erythroblastosis virus, inhibits differentiation by causing autonomous proliferation or, vice versa, causes autonomous proliferation by inhibiting differentiation (21). Cultured chicken heart mesenchymal cells, unlike hematopoietic cells, are proliferatively quiescent and show no tendency to differentiate. Our observation that infection with erythroblastosis virus renders chicken heart mesenchymal cells proliferatively active suggests therefore that initiation of cell replication, rather than inhibition of differentiation, is responsible for oncogenesis by this agent. Temperature-sensitive mutants of RSV transform cells at 35-37°C, the temperature permissive for transformation, but not at 42°C, the nonpermissive temperature. Other laboratories have reported that chick embryo fibroblasts infected with tsNY68, a standard temperature-sensitive mutant of SRA-RSV, grow as tumors at 42°C on the chorioallantoic membrane of embryonated eggs (22) and that tsNY68 causes autonomous proliferation at 42°C of cultured chick embryo chondrocytes (23) and neuroretinal cells (24). These published reports that tsNY68 causes autonomous cell proliferation at a temperature (42°C) at which the mutant transforming protein p60 src has considerably reduced tyrosine-specific kinase activity form the basis for an extant hypothesis that a function of the src protein other than phosphorylation may be responsible for autonomous proliferation (24). Autonomous proliferation of chick embryo neuroretinal cells infected with the temperature-sensitive mutant tsNY72 has, on the other hand, been reported to be temperature sensitive (24). Unlike other groups, we have found that the proliferation of chicken heart mesenchymal cells infected with tsNY68 is temperature sensitive. Like
Vol. 35, No. ii, 1984
Avian Retroviruses
and Growth Autonomy
1169
other groups, we have found that the proliferation of chicken heart mesenchymal cells infected with tsNY72, tsLA24 or tsLA29 is temperature sensitive. The most straightforward explanation for the difference in temperature-sensitive proliferation between tsNY68-infected chicken heart mesenchymal cells and the tsNY68-infected cells studied by other groups may be in "tighter" behavior of this mutant in the present system. In any case, the temperature-sensitive proliferative behavior of chicken heart mesenchymal cells infected with tsNY68, as well as with tsNY72, tsLA24 and tsLA29 suggests that enzymatically active p60 src is necessary for autonomous (hormone-independent) proliferative behavior. The slowly oncogenic avian retroviruses (avian leukosis viruses), lacking one genes as they do, are thought to cause autonomous proliferation of infected cells by "promoter insertion" into host cell chromosomes (11-13): Host cell proto-onc genes appear to be activated by vicinal integration of avian leukosis proviruses with their regulatory LTR sequences. Such critical vicinal integration appears to require the passage of months in the animal host. Consistent with this view, we have found that infection with avian leukosis viruses (RAV-I, RAV-2, PRL-42) or with src-gene deletion mutants of RSV (transformation-defective mutants) does not alter the morphology or proliferative behavior of chicken heart mesenchymal cells in monolayer culture. These findings also serve as negative controls for any potential effects of the avian leukosis virus "helpers" that accompany the defective sarcoma (FSV, Y73-ASV) and defective acute leukemia viruses (AEV, MC29, MH2) used in our studies. While chicken heart mesenchymal cells infected with avian sarcoma viruses, erythroblastosis virus or the myelocytomatosis viruses MC29 and MH2 proliferate autonomously in monolayer culture, only the sarcoma virusinfected cells are capable of colony formation in agarose suspension culture. As noted in Materials and Methods, cell density, plasma concentration, calcium concentration and all other parameters are the same in our monolayer and agarose suspension cultures, making the two rigorously comparable. Erythroblastosis virus-infected cells, like sarcoma virus-infected cells, proliferate very rapidly in monolayer culture (50-100 fold increase in four days), but do not form colonies in agarose, indicating that rapid proliferation in mouolayer culture is not sine qua non for colony formation in suspension culture. Other investigators have reported that AEV and MC29-infected chick embryo fibroblasts do form colonies in an anchorageindependent manner, i. e. in suspension culture (19). This difference between the chicken heart mesenchymal cell system and the chick embryo fibroblast system may he a consequence of different properties of the respective cell types themselves or of differences in the composition of the culture media employed (plasma vs. serum, synthetic media, etc.). In any case, the chicken heart mesenchymal cell system can distinguish between the ability of avian retroviral onc genes to induce autonomous (mitogenic hormone-independent) cell proliferation in monolayer culture and their ability to induce colony formation in agarose suspension culture. Colony formation in suspension culture has not been found to be a general and specific index of the ability of cells to form tumors in vivo (25,26) or of the ability of ceils to grow as metastases (27-29). Because mitogenic hormone-independent (autonomous) proliferation is the essence and defining characteristic of neoplasia, the proliferation of retrovirus-infected chicken heart mesenchymal cells in plasma-containing medium in monolayer culture may represent a reliable and biologically meaningful in vitro index of the tumorigenicity of the onc genes of those retroviruses. It is interesting to note, in this connection, that autonomous proliferation of chicken heart
i170
Avian Retroviruses
mesenchymal cells in vitro onc gene, e. ~., src (RSV) embryo fib~oblasts appears genes, !. ~. ras plus myc
and Growth Autonomy
Vol. 35, No. ii, 1984
appears to require the transduction of a single or myc (MC29), while tumorigenicity in vivo of rat to require transfection of the cells with two onc (5,30).
Mammalian cells transfected with myc or with myc-related oncogenes have been found to be capable of sustaining an unlimited number of population doublings in monolayer culture, i. e. to be "immortalized" or "established" (15). We have found, on the other hand, that chicken heart mesenchymal cells infected with MC29 undergo senescence in monolayer culture, as do chicken heart mesenchymal cells infected with Rous sarcoma virus (Fig. 4). Other workers have, likewise, found that chick embryo fibroblasts infected with RSV undergo senescence (M. Weber, personal communication), as do chick embryo fibroblasts infected with MC29 (19,31), but that MC29-infected quail embryo fibroblasts are "immortalized" (31). "Nonimmortalized," RSV-infected chick embryo fibroblasts have been reported to form tumors upon transplantation into athymic nude mice, indicating that a capacity of cells to sustain an unlimited number of doublings may not be necessary for tumorigenicity (32). The relevance of the cell culture criterion of "immortalization" to oncogenicity of avian retroviruses is therefore uncertain. Acknowledsments This work was supported by Grant #1536 from the Council for Tobacco Research and by a grant from the Thoracic Foundation. The authors thank Dr. R. Erikson for his invaluable advice and other investigators for viral inocula, Dr. M. A. Legg, Mr. R. D. Pence and Mr. Joseph Flaherty for their support and encouragement and Mrs. Ruth Dziadul for expert preparation of the manuscript. This is reprint No. 752 from the Cancer Research Institute of the New England Deaconess HospitaL. References i. 2. 3. 4. 5. 6. 7. 8. 9. i0. 11. 12. 13. 14. 15. 16. 17. 18. 19.
A BRAUN, The Story of Cancer, Addison-Wesley, Reading, MA (1977). J. M. BISHOP, Ann. Rev. Biochem. 52 301-354 (1983). T. DUELL, J. HUANG, S. HUANG, P. STROOBANT and M. WATERFIELD, Science 221 1348-1350 (1983). J. DOWNWARD, Y. YARDEN, E. MAYES, G. SCRACE, N. TOTTY, P. STOCKWELL, A. ULLRICH, J. SCHLESSINGER and M. WATERFIELD, Nature 307 521-527 (1984). H. LAND, L. PARADA and R. WEINBERG, Science 222 771-777 (1983). S. BALK, Proc. Natl. Acad. Sci. USA 77 6606-6610 (1980). S. BALK, S. LEVINE, L. YOUNG, M. LA FLEUR and N. RAYMOND, Proc. Nat. Acad. Sci. USA 78 5656-5660 (1981). S. BALK, H. GUNTHER and A. MORISI, Life Sciences 34 803-808 (1984). S. BALK, R. SHIU, M. LA FLEUR and L. YOUNG, Proc. Natl. Acad. Sci. USA 79 1154-1157 (1982). S. BALK, A. MORISI, H. GUNTHER, M. SVOBODA, J. J. VAN WYK, S. P. NISSLEY, and C. G. SCANES, Life Sciences, in press. M. HAYMAN, J. Gen. Virol. 52 1-14 (1981). P. ENRIETTO and J. WYKE, Adv. Can. Res. 39 269-314 (1983). Y. FUNG, W. LEWIS, L. CRITTENDEN and If. KUNG, Cell 33 357-368 (1983). R. HOLLIDAY, Nature 306 742 (1983). J. CAIRNS and J. LOGAN, Nature 304 582-583 (1983). J. BRUGGE and R. ERIKSON, Nature 269 346-348 (1977). D. MC CLURE, Cell 32 999-1006 (1983). H. HANAFUSA, Proc. Natl. Acad. Sci. USA 63 318-322 (1969). B. ROYER-POKORA, H. BEUG, M. 6q.AVIEZ, H. WINKHARDT, R. FRIIS and T. CRAF, Cell 13 751-760 (1978).
Vol. 35, No. ii, 1984
Avian Retroviruses and Growth Autonomy
1171
20. N. KAN, C. FLORDELLIS, C. GARON, P. DUESBERG and T. PAPAS, Proc. Natl. Acad. Sci. USA 80 6566-6570 (1983). 21. T. GRAF and H. BEUG, Cell 34 7-9 (1983). 22. G. POSTE and M. FLOOD, Cell 17 789-800 (1979). 23. A. TANAKA, C. PARKER and A KAJI, J. Virol 35 531-541 (1980). 24. F. POIRIER, G. CALOTHY, R. KARESS, E. ERIKSON and H. HANAFUSA, J. Virol. 42 780-789 (1982). 25. C. STILES, W. DESMOND, L. CHUMAN, G. SATO and M. SAIER, Cancer Res. 36 3300-3305 (1976). 26. L. SMETS, Biochim. Biophys. Acta 605 93-111 (1980). 27. R. LOTAN and A. RAZ, Cancer Res. 43 2088-2093 (1983). 28. H. RUBIN, C. ROMERDAHL and B. CHU, J. Natl. Cancer Inst. 70 1087-1096 (1983). 29. M. DODSON, F. GELDER, J. SLOTA and C. LANGE, Int. J. Cancer 32 211-217 (1983). 30. H. LAND, L. PARADA and R. WEINBERG, Nature 304 596-602 (1983). 31. K. BISTER, M. HAYMAN and P. VOGT, Virology 8 2 431-448 (1977). 32. P. KAHN, K. NAKAMURA, S. SHIN, R. SMITH and M. WEBER, J. Virol. 42 602611 (1982).