Cell Research 167 (1986) 400-406
Preferential Modulation and Differentiation
of Embryonic Cell Proliferation by Embryonic Interferon
JAMES J. GREENE’, * and PAUL 0. P. TS’O’ ‘Department ‘Division
The Catholic Thcz Johns
Univer.cit~ of Atrreriw, Wa.\hington. DC 20064 trnd Hopkins University, Bultimore. MD 21205. USA
Embryonic-interferon (E-IFN), a novel species of interferon (IFN) produced solely by embryo cells, inhibited the proliferation of embryo cells in early gestation preferentially vis-a-vis fetal cells in late gestation but had little effect on the in vitro differentiation of embryonic pre-adipocytes to adipocytes. In contrast, mature-interferon (M-INF) did not exhibit this preferential inhibition of cell proliferation and did inhibit pre-adipocyte differentiation. These results suggest that the function of E-INF is different from that of M-IFN and that this function may involve modulation of the developmental process. 0 1~86 Academic
Interferons are biological response modifiers capable of exerting a profound influence on cells. While the antiviral action of interferons (IFNs) is their best known and most thoroughly characterized effect, a diversity of many other effects are now generally recognized to be exerted by IFNs. The ability of IFNs to affect such basic cellular processes as proliferation [ 1, 21, differentiation [3-81, cytoskeletal organization [9-l 11, and cell surface receptors [ 12-141 suggests that the role of IFNs encompasses a far wider range of functions than merely the antiviral effect. It is becoming increasingly apparent that IFNs may serve a role as a natural regulatory molecule. Such a role may include a function as a modulator of the developmental process. We have recently shown that Syrian hamster embryo cells are capable of producing an ‘embryonic’ form of IFN (E-IFN) . This unique E-IFN is produced by cells upon induction only during the early stages of embryogenesis and differs structurally from ‘mature’ IFN (M-IFN) produced by adult, fetal, and embryo hamster cells. To evaluate the potential of E-IFN to modulate the developmental process, we examined the effects of both E-IFN and M-IFN on the proliferation and differentiation of hamster cells. MATERIALS
Cells Cells used in the proliferation studies were derived from embryos removed by cesarean section from pregnant Syrian hamsters at 9 or I3 days of gestation. The hamsters were tested and found free of persistent viral infections which are known to disturb the interferon system. The excised embryos * To whom offprint
requests should be addressed. Copyright @ 1986 by Academic Press, Inc. All rights of reproduction I” any form reserved 0014-4827/86 $03.00
Embryonic interferon modulation of cell proliferation and dgyerentiation
Table 1. Characteristics of M-IFN and E-IFN”
Molecular weight Affinity to blue-dextran’ Species-specific antiviral activity Stability at pH 2 Stability at pH 11
21 000 + + -
26 000 + + + +
LI Summary of physical characteristics described in Greene et al. [IS]. b Binding at pH 6.45; both E-IFN and M-IFN bind at pH 7.4.
were dispersed into single-cell suspensions which were used to initiate cell cultures [ 151.Cells were cultured in Dulbecco’s modified minimal essential medium (MEM) which was supplemented with 20 % heat-inactivated fetal bovine serum for growth (growth medium) or 12% inactivated fetal bovine serum for maintenance (maintenance medium).
Interferons The E-IFN was prepared from early passage cultures of 9-day gestation (9-dg) embryo cells, while the M-IFN was prepared from cultures of either 9-dg or 13-day gestation (13-dg) fetal cells. Both cultures were induced by treatment with UV-inactivated Newcastle disease virus (NDV) . The 9dg embryo cells were capable of producing both M-IFN and E-IFN, while the 13-dg fetal cells produced only M-IFN. For comparison, M-IFN was also prepared from a continuous clonal line of hamster 21fl cells. The E-IFN was purified by chromatography, first on superfine Sephadex G-100 columns, then on blue-dextran Sepharose columns. The M-IFN was purified initially by chromatography on superfine Sephadex G-100 columns followed by tandem chromatography on metal chelate (Cu”-agarose) and hydrophobic ligand (phenyl-Sepharose) columns ]16]. Specific activity of the IFNs were on the order of 1x 10’ reference unitsimg protein. Table 1 summarizes the characteristics of E-IFN and M-IFN.
Effects of IFN on Proliferation Proliferating cultures of each cell type were prepared by inoculating I x IO5cells into duplicate or triplicate 35mm dishes in growth medium. Cultures were incubated for 24 h to allow time for the cells to attach, after which cell counting established the initial cell number. The cultures were refed with growth medium either alone or with medium containing E-IFN or M-IFN, or mock IFN, after which they were incubated for a further 48 h before determining the final cell numbers.
Effects of IFN on Pre-adipocyte Differentiation The modulatory effects of IFN on pre-adipocyte differentiation were studied on E9-Cll. a clonal line of pre-adipocytes isolated from a 9-dg primary culture. Adipocytes were identified by morphology and staining of formalin-fixed cells with the lipid-sensitive stain Oil Red 0 (I .5 % w/v solution in 60 ‘% ethanol). Previous studies in our laboratory have verified that Oil Red 0 staining E9 cells are adipocytes, on the basis of triacylglycerol accumulation and electron microscopy [ 171.
RESULTS AND DISCUSSION Fig. 1 shows the effect of purified E-IFN and M-IFN on the proliferation of 9dg and 13-dg cells. To allow comparative analysis of the biological effects of these different IFNs, the IFN concentrations were normalized with respect to their antiviral activities when titered on the clonal 21fl cell line. Each of these IFNs E-rr, Cell Rcs 167 (1986)
402 Greene and Ts’o
I N&ERON (antivIral
so 70 too140
Fig. I. Antiproliferative effects of E-IFN and M-IFN on 9-dg and 13-dg cells. Proliferation was quantitated as d, scribed in Materials and Methods. Results are express& as the increase in cell number during the 48-h treatment as a percentage of the increase in control cultures during the same period. Similar results were obtained in two additional experiments. Effects on (A) l3-dg; (B) 9-dg cells. O-O, 13-dg; n---n, 9-dg M-IFN; O-0, 9-dg E-IFN.
was found to manifest a potent antiproliferative effect. The activities of M-IFN and E-IFN in inhibiting the proliferation of 13-dg fetal cells were essentially identical (fig. 1A). However, E-IFN was significantly more effective than was MIFN in inhibiting the proliferation of the earlier gestation 9-dg embryo cells (fig. 1B). The concentration of E-IFN required for inhibiting the proliferation of 9-dg cells was less than that required for 13-dg cells. At a concentration of 60 U/ml, EIFN completely abolished the proliferation of embryo cells without toxicity but allowed the proliferation of fetal cells to continue at 5045% of their untreated rates. For M-IFN, similar concentrations were required for inhibition of 9-dg and I3dg cells. This suggested that E-IFN (but not M-IFN) was capable of preferential inhibition of proliferation of earlier gestation embryo cells. The cellular origin of the M-IFN was shown not to be important in these studies, since the antiproliferative activities of M-IFNs derived from 9-dg, 13-dg (fig. 1A,B). and even from the clonal line 21fl cells (data not shown) were indistinguishable, regardless of the cell type on which the antiproliferative activities were assayed. Primary and low passage secondary cultures of Syrian hamster embryo cells contain a subpopulation of progenitors of adipocytes (pre-adipocytes) capable of undergoing in vitro differentiation to adipocytes . A clonal line, designated E9, isolated from a culture of 9-dg embryo cells exhibited a high propensity for undergoing conversion to adipocytes. Fig. 2 shows the morphology of E9 cells
Fig. 2. Phase-contrast photomicrograph of monolayer cultures of E9 cells. (A) lmmediately upon reaching confluence; (B) after incubation for an additional 12 days at confluency. the appearance of adipocytes are indicated by the appearance of the lipid droplets. x240.
immediately upon reaching confluency (fig. 2A) and after maintaining confluency for 10 days (fig. 2B). While no adipocytes were initially apparent, a large number of E9 cells were converted to adipocytes when kept in a non-proliferative state (i.e., contact-inhibited) for 7-12 days. Thus, the E9 cells were used to measure the effects of M-IFN and E-IFN on differentiation as reflected in the modulation of E9 pre-adipocytes to adipocyte conversion. The conversion of pre-adipocytes in the presence and absence of IFN was quantitated by both mass culture and clonal assays. In the mass culture assay, E9 cells were seeded at high density into culture dishes and allowed to reach confluency before treating with IFNs. The clonal assay involved seeding E9 cells at low density and allowing for the development of clones before treating with IFN. The results in table 2 indicate that in mass culture, and at the concentration of 10 units/ml, E-IFN only minimally inhibited differentiation (less than 7%), while M-IFN inhibited differentiation by 26-44 %. Clonal assays also showed that E-IFN was less inhibitory than M-IFN in arresting the conversion of preadipocytes clones to adipocytes (table 3). The differences in the inhibitory activities between M-IFN and E-IFN were dependent upon the criterion used in scoring adipocyte clones. In these experiments, clones were scored as ‘positive’ (an E.rp Cd Re\ 167 (IY86)\
404 Greene and Ts’o
Table 2. Inhibitory effects of two types of interferon on pre-adipocyte differentiation in mass culturea
No. of adipocytesi 10 mm’
Control E-IFN M-IFN (from 9-dg cells) M-IFN (from 13-dg cells)
384 358 282 214
6.8 26.9 44.3
u Confluent monolayers of E9-Cl1 cells were established by inoculating 1 x IO5 of E9 cells at passage 6-10 into growth medium in 25cm’ flasks. After reaching confluency (48 h), cultures were refed with either maintenance medium alone or medium containing 10 U/ml of E-IFN or M-IFN. One IFN unit equalled one laboratory reference unit of antiviral activity when assayed on the 21fl clonal line of hamster tibroblast cells [151. After 7 days, the cultures were scored for the presence of adipocytes. Cell counts at the time of assay indicated no significant differences in the total number of cells in the control and treated cultures. Adipocytes were identified as described in Materials and Methods. Results are expressed as the number of adipocytesilo mm2 and is the average of 10 counts of duplicate assays. The maximum variation between the duplicate assays is 28%.
adipocyte clone) when at least 20 % (low criterion) or at least 30 % (high criterion) of the cells in the clone had undergone conversion to adipocytes. Typically, MIFN-treated ‘positive’ clones contained approx. 20 % adipocytes, while E-IFNtreated ‘positive’ clones contained 30-40 % adipocytes. By adjusting the criterion for a clone scored as positive to a higher adipocyte percentage (i.e. from 20 to 30%), the differences between M-IFN and E-IFN activities could be greatly magnified. These results appeared to derive from the direct action of IFNs rather than
Table 3. Inhibitory effects of two types of interferon on pre-adipocyte differentiation in a clonal assaya 20 % conversion
30 % conversion
Control E-IFN M-IFN (from 9-dg cells) M-IFN (from 13-dg cells)
24133 17142 5138 513I
72.7 40.5 13.2 16.1
23133 17142 1138 0131
69.7 40.5 2.6 0
0 Clones of E9-Cl1 cells were established by inoculating 2~ lo3 cells into 25 cm’ flasks. After allowing 6 days for the development of clones, the cultures were refed with either medium alone or medium containing 10 U/ml E-IFN or M-IFN. After 7 days, the clones were examined for the presence of adipocytes as described in table 2. A clone was scored as positive when greater than 20% or greater than 30% of the cells in the clone were adipocytes. Results are an average of duplicate assays. Variation in the number of positive clones between duplicates assays was f 11%. b Ratio corresponds to no. of clones scored as positive per no. of total clones scored. Exp Cell Res 167 (1986)
of cell prol(fercrtion
other inhibitory factors released from the cells. Failure of mock preparations of ‘IFN’, prepared from uninduced cultures and subjected to the same purification as authentic IFN, to mimic the results described above indicates that the inhibitory factor must be specifically induced by treatment with NDV, a characteristic of IFNs. It is conceivable that these effects are caused by non-IFN factors induced by NDV, though this is unlikely in view of the high degree of purification to which the interferons had been subjected. Moreover, the substances described here would have to exhibit the properties of species-specific antiviral activity and retain their biological activities after incubation at pH 2.0, properties largely regarded to be unique to IFNs. There remains, however, the ve:ry remote possibility that some contaminant could have been co-purified along with the IFNs. At first glance, the existence of a unique form of E-IFN with ostensibly the same antiviral properties as M-IFN appears to be redundant. This redundancy in IFNs, however, may be related to the consideration that differentiation of mammalian cells is most likely to be preceded by the arrest of proliferatilon. This is indeed the case for hematopoietic cells [ 181, epithelial cells of the sk.in [19, 201 and pre-adipocytes [2I]. If this requirement is common to all tissue/cell types, then it may be a necessity for a mechanism to transitorily inhibit cell proliferation with tissue/cell specificity during embryonic development. Such inhibition may be mediated by growth antagonist factors. A candidate for a growth antagonist which participates in the developmental process must fulfill certain criteria: (1) it must specifically inhibit the proliferation of cells to be differentiated-and not cells already committed to a given lineage; (2) the inhibition of cell proliferation in response to the factor must be reversible; (3) it must not inhibit the process of differentiation, but only the process of proliferation. E-IFN generally fulfills each of these criteria. Differential activity between E-IFN and M-IFN in inhibiting cell proliferation was evident when the actions of the two IFNs were compared on 9dg embryo and 13-dg fetal cells. On the embryo cells, E-IFN was more effective than M-IFN in inhibiting proliferation. This distinction was lost on fetal cells on which E-IFN and M-IFN were equally effective. Most significantly, E-IFN was more effective on 9-dg cells than on 13-dg cells. These observations, taken together, are consistent with the concept of E-IFN possessing target specificity for less differentiated cells. The antiproliferative actions of E-IFN ;as well as those of M-IFN were seen to be reversible, a characteristic feature of IFNs (data not shown). IFNs are considered to be inhibitory for many differentiation systems; for example, mouse IFN is inhibitory for the differentiation of mouse 3T3Ll 161and BALBlc3T3  pre-adipocytes, a situation similar to that of the inhibition exerted by hamster M-FIN on E9 pre-adipocytes. In contrast, E-IFN only minimally inhibits differentiation. The ability of E-IFN to selectively inhibit the proliferation of embryo cells while allowing differentiation to proceed suggests that IFN (or IFN-like substances such as E-IFN) may serve as a modulator of the developmental process. Exp Cell l
406 Greene and Ts’o REFERENCES I. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
Paucker, K, Cantell, K & Henle, W, Virology 17 (1962) 324. Brouty-Boye, D, Lymphokine reports 1 (1980) 99. Gresser, I, Proc natl acad sci US 47 (1961) 1817. Rossi, G B. Matarese, G P. Grapelli, C, Belardelle, F & Benedetto. A, Nature 267 (1977) 50. Rossi, G B, Dolei, A, Cioe, L, Benedetto, A, Matarese, G P & Belardelle, F, Proc natl acad sci US 74 (1977) 2036. Keay, S & Grossberg, S E, Proc natl acad sci US 77 (1980) 4099. Tomida, M. Yamamoto, Y & Hozumi, M. Cancer res 40 (1980) 2919. Vogel, S N. Weedon, L L. Moore, R N & Rosenstreich, D L, J immunol 128 (1982) 380. Pfeffer, L M, Wang, E & Tamm, I, J cell biol 85 (1981) 9. Chang, C, Lymphokine reports 4 (1981) 409. Wang, E, Pfeffer. L M & Tamm, I, Proc natl acad sci US 78 (1981) 6281. Fridman, W H. Gresser, 1, Bandu, M T, Aguet, A & Neauport-Sautes, C, J immunol 124 (1980) 2436. Lindahl. P. Gresser, 1, Leary, P & Tovey, M, Proc natl acad sci US 73 (1976) 1284. Heron, I, Hokland, M & Berg, K, Proc natl acad sci US 75 (1978) 6215. Greene, J J, Dyer, R H, Yang, L C & Ts’o. P 0 P. J interferon res 4 (1984) 517. Bollin, E & Sulkowski, E. J gen virol 52 (1981) 227. Ueo. H, Nakano, S, Bruce, S & Ts’o, P 0 P. In preparation. Lajtha, L G, Lord, B I, Dexter, T M, Wright, E G & Allen. T A. Cell differentiation and neoplasia (ed G F Saunders) p. 179. Raven Press, New York (1978). Iverson, 0 H, Bjerknes, R & Devik. F, Cell tissue kinetics I (1968) 3.51. Ham. R G & Veomett, M J, Mechanisms of development, p. 317. Mosby, St. Louis, MO. (1980). Scott, R E, Florine, D L. Wille, J J & Yun, K, Proc natl acad sci US 79 (1982) 845. Cioe. L, O’Brien, T G & Diamond. L, Cell biol int rep 4 (1980) 255.
Received February 12, 1986 Revised version received June 16, 1986
Exp Cell Res 167 (1986)