Establishment of the antiviral state in alpha, beta-interferon-resistant Friend cells treated with gamma-interferon. Induction of 67-kilodalton protein kinase activity in absence of detectable 2-5A synthetase

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

response in IFN-treated cells has been shown to be often accompanied by the induction of at least two enzymatic pathways. The first one is a 2',5'-oligoadenylate synthetase capable of generating 2',5'-linked oligoadenylates from ATP, upon activation by double-stranded RNA (11,12). These oligomers in turn activate RNase L, a latent endoribonuclease that cleaves single-stranded RNAs at preferred sites (13)(14)(15)(16)(17)(18)(19)(20). The second enzymatic activity is a 67-kDa protein kinase that, if activated by double-stranded RNA, phosphorylates and thereby impairs the activity of the a peptide chain of eukaryotic initiation factor 2 (21-23). The activities of both protein kinase and 2-5A synthetase may account for the inhibition of overall protein synthesis that has been detected in lysates from IFN-treated cells.
In the past few years, we have been studying the effects of mouse IFN administration on Friend leukemia cells (FLC) and clones thereof which are resistant to the a,PIFN-induced antiviral state. FLC are mouse nucleated erythroid precursors able to differentiate in vitro when treated with various agents including dimethyl sulfoxide (24). a,PIFN is able to enhance or inhibit erythroid differentiation of wild-type FLC when administered a t low or high doses, respectively (7,26). The variant clones are resistant to the enhancement of erythroid differentiation caused by low doses of pure a,@IFN, but high doses thereof, which do not decrease Friend murine leukemia virus release and vesicular stomatitis virus (VSV) yields, inhibit erythroid differentiation of the resistant clones tested (27).
Both a,@IFN-resistant and -sensitive FLC clones show a specific saturable binding site for mouse a,@IFN with similar affinity constant. However, the variant clones appear to be unable to "induce" the 2-5A synthetase pathway even upon treatment with 10,000 units/ml a,@IFN (28).
In this report, we describe data showing that treatment with yIFN of the variant clones results in the establishment of a bona fide antiviral state showing that y-and a,@IFNs may have different interaction with the same cell system. Furthermore, the antiviral state induced by yIFN takes place in the absence of any detectable 2-5A synthetase activity. As for the 67-kDa protein kinase activity, the variant FLC clones do induce it upon exposure to 20-500 units/ml yIFN, whereas no such activity is detectable when the same cells are exposed to up to 13,000 units/ml a,@IFN.

Effects of yIFN in a,PIFN-resistant Friend Cells
Ci/mmol) were obtained from the Radiochemical Centre, Amersham, United Kingdom. Poly(r1) .poly(rC) and poly(r1) .poly(rC)-agarose were from P-L Biochemicals. Benzidine dihydrochloride was from Ciba. Poly(A) was from Sigma. Oligo(dT)12.18 from Collaborative Research Inc., Waltham, MA. All other chemicals and solvents were of reagent grade.
Cells-FLC were grown in RPMI 1640 medium supplemented with 5% fetal calf serum and antibiotics. L929 and C-243-3 mouse cells were grown in minimum essential medium supplemented with 5% fetal calf serum and newborn calf serum, respectively.
Viruses-VSV (Indiana strain) and encephalomyocarditis virus (EMCV) stocks, obtained by infecting L929 cell monolayers with low multiplicity of infection (0.1 pfu/cell), were titrated by plaque assay on the same cells. Titers ranged between and lo9-' pfu/ml.
Newcastle disease virus, strain F, stocks were obtained by infecting 10-11-day-old embryonated eggs in allantoic cavity. The virus was harvested after 3 days of incubation a t 35 "C and titered by hemagglutination assay. Titers ranged between 300 and 600 hemagglutinating units/ml. Reverse Transcriptase Assay-The assay was performed in a final volume of 100 p1 containing 20 mM Tris-HC1, pH 7.8, 60 mM NaC1, 1 mM MnCIZ, 5 mM dithiothreitol, 0.015% Nonidet P-40, 50 pg/ml poly(A), 5 pg/ml oligo(dT)lz.la, 2 p~ TTP. Samples of clarified supernatants, ranging from 5 to 30 pl, with additional medium containing 5% fetal calf serum to a total volume of 30 pl, were added to the reaction mixture. To rule out artifactual results due to the presence of inhibitors in the fluids, the enzymatic activity was derived from the linear portion of the curves. Values were expressed as picomoles of TTP incorporated into acid-insoluble materials/106 cells.
Interferons-Mouse n,BIFN was essentially prepared by infecting mouse C-243-3 cells with Newcastle disease virus as described (29). The specific activities of n,BIFN preparations averaged lo5 units/mg of protein. Amounts of IFN are given throughout this paper in mouse units, i.e. the amount of IFN reducing by 50% plaque production by VSV. This unit equals 4 research reference units of the National Institutes of Health Search Standard preparation code G-002-904-51 1 whose titer was 12,000 international units/ml when reconstituted. Details about this standard preparation are reported in Research Reference Reagents Note 15 (56).
Mouse yIFN preparations were generous gifts of different laborat,ories. yIFN obtained from Drs. H. Heremans and A. Billiau (Rega Institut, Leuven, Belgium) was produced in cultures of mouse spleen cells stimulated with Staphylococcus aureus lysate and partially purified by silicic acid (30). The preparations obtained from Drs. F. Marcucci and H. Kirchner (Institut fur Vimsforschung, Heidelberg, West Germany) were produced in cultures of mouse spleen cells stimulated with concanavalin A and partially purified by (NH&SO4 precipitation (31). The preparation from Dr. S. Landolfo (Istituto di Microbiologia, Universita di Torino, Italy) was produced in cultures of a T-cell tumor clone (L12-R4) stimulated with phorbol myristate acetate and partially purified by (NH&SO4 precipitation (32). Specific activities ranged between lo4 and lo5 units/mg of protein. In absence of an international standard, amounts of yIFN are derived from the same assay described for ru,BIFN, i.e. are given in laboratory units. Recombinant murine yIFN from Escherichia coli was a generous gift of Boehringer Institut (specific activity 7.2 X 106 units/mg of protein).
Cell Cloning in Soft Agar Medium-Cells serially (10-fold) diluted in 5 ml of RPMI 1640 supplemented with 23% fetal calf serum, then added with melted Bacto-Agar (2%, 1 ml) to obtain a final 0.33% Bacto-Agar concentration, were plated in a 60-mm Petri dish. After medium solidification, cultures were incubated at 37 "C in a humidified 5% CO, atmosphere. Clones were counted 6-8 days later.
Assay of 2',5'-Oligoadenylate Synthetase Actiuity-30 p1 of cell extract were incubated with 20 p1 of poly(rI).poly(rC)-agarose for 1 h a t room temperature. The enzyme bound to the poly(r1) .poly(rC) was washed three times with 200 pl of 20 mM Hepes, pH 7.6,8.5 mM KC1, 50 mM MgOAc, 50 mM KC1, 7 mM EtSH, 20% (v/v) glycerol and spun at 10,000 X g for 20 s. The enzyme was then incubated with 20 p1 of the same buffer containing 3 mM ATP (neutralized with NaOH) overnight at 30 "C with gentle shaking. The yield of 2-5A was assayed by high pressure liquid chromatography analysis. The high pressure liquid chromatography analysis was carried out on a Lichrosorb RP-18 column (Merck) run in 50 mM ammonium phosphate, pH 7.0, at 1 ml/min with a 25-min linear gradient of 5050 metha-no1:water to reach a final value of 25% (33).

Induction of the Antiviral State in cu,pIFN-resistant FLC Clones Treated with Murine yIFN-As
shown in the doseresponse curves of Fig. 1, a 20-h treatment of wild-type FLC (745A clone) with a crude murine yIFN preparation results in a 3-log reduction of VSV yield (Fig. 1, lower panel). The same treatment is able to reduce VSV yield in two a,@IFNresistant FLC clones, but a t a lesser extent than in wild-type cells. Release of FLV is also affected by the same treatment ( Fig. 1, upper panel). Table I shows the extent of reduction of VSV yield and reverse transcriptase activity in wild-type and five a,@IFN-resistant FLC clones treated for 20 h with 200 units/ml murine y-IFN preparations.
Specificity of the Antiviral State Induced by yIFN in FLC Variants-The results described above were obtained with crude murine yIFN preparations. Control experiments have obviously been performed to demonstrate that the observed reduction of virus yield is indeed the consequence of the induction of a bona fide antiviral state. Steps taken were as follows.
1) Data obtained with the murine yIFN preparations from Drs. Marcucci and Kirchner (and presented in Fig. 1) were perfectly superimposable (data not shown) with those obtained when batches of murine yIFN prepared by various procedures in different cell lines (see "Experimental Procedures'') were also used.
2) VSV yields and reverse transcriptase activity were reduced in culture supernatants also following a 20-h treatment of sensitive and resistant cells with a preparation of recombinant mouse yIFN (7.2 x lo6 units/mg of protein) (Fig. 2).
3     Table I1 (right) show that both VSV and EMCV yields were markedly reduced in cells treated with yIFN. 5) A similar reduction of VSV yield was also observed when (a) cells were infected with 10 pfu/cell or (b) besides testing culture supernatants, virus production was also evaluated in freeze-thawed aliquots of the same culture conditions (data not shown). The latter experiment rules out the possibility that the partial reduction of VSV yield observed in the resistant cell clones was due to accumulation of infectious virus particles inside the cells.
These data are taken to show that the observed yIFNinduced reduction of virus replication in a,PIFN-resistant FLC is due to the establishment of a bona fide antiviral state.
IFN-induced 2-5A Synthetase Activity-To investigate whether the sensitivity to yIFN of aJIFN-resistant FLC variants was the result of the induction of the 2-5A synthetase pathway (uninduced by a,PIFN treatment) (28), 2-5A synthetase activity was analyzed in sensitive and resistant FLC treated with yIFN. The enzyme was assayed in presence of poly(r1) .poly(rC)-agarose (see "Experimental Procedures"). The agarose-bound enzyme is stable and free of any phosphatase activity which may be present in cell extracts (37), thus allowing synthesis and accumulation of 2-5A. Similarly to the data previously shown for a,PIFN, analysis of 2-5A synthetase activity reveals that treatment of lo6 cells/ml with 500 units/ ml yIFN for 20 h induces a 5-fold increase of oligomer concentration in the reaction mixture from FLC. No inducibility of this enzymatic activity is detectable in the a,PIFNresistant clones. The 2-5A synthetase-specific activity at different doses of yIFN is presented in Fig. 3 Fig. 1. 0, A, 745A; 0 zyme, the double-stranded RNA-dependent 67-kDa protein kinase, was induced in the variant cells exposed to yIFN. To validate detection of kinase activity, this was determined both in solution (Fig. 4) and by immobilization on poly(r1). poly(rC)-agarose (Fig. 5). When assayed in solution, the characteristic double-stranded RNA-dependent increase of 67-kDa band phosphorylation is clearly visualized in cell extracts of wild-type and variant FLC treated with 500 units/ml rIFN. In the former cell extract, the kinase was activated also when the cells had been exposed to 200 units/ml a,PIFN. The kinase was also immobilized on poly(r1) .poly(&)-agarose as described by Hovanessian and Kerr (37) and evaluated for its capacity to phosphorylate the 67-kDa endogenous substrate.  Here again, the results indicate that the enzyme activity is similarly induced in lysates from both wild-type and variant FLC treated with rIFN, whereas no inducibility thereof is detectable in FLC variants treated with 13,000 units/ml a,PIFN. The pattern of phosphorylation (solid phase assay) of extracts from 745A, 3C18, 1C11, 7, 8, and 9 treated with CY,@-or r I F N is presented in Fig. 5.

DISCUSSION
Recently, cell variants (38,39) have been described which are partially or totally sensitive to yIFN treatment, despite their a,PIFN-resistant phenotype. For example, a,PIFN-resistant murine leukemia L1210 cells, which lack high affinity binding receptor sites for a,PIFN (40), are sensitive to yIFN. Bovine serum alhumin (68,000). ovalbumin (45,000), and chvmotrypsinogen (2.5,000) were used as molecular weight standards. In addition, cell lines vary widely in their sensitivity to IFNs and in the constitutive level or inducihility of these enzymes. In humans, hoth enzymes are induced in HeLa cells hy I F S ; neither is detectahle in MRC5 cells, which are as sensitive to n,@IFN as HeLa cells (52); and neither recomhinant crA-or recomhinant yIFN induces the kinase activity in GM2767A fihrohlasts, whereas only the former is ahle to induce the kinase in amnion U cells (45). In the murine system, hoth enzymes are induced in L929 cells, whereas only the protein kinase is induced in K/RALR and only the 2-:iA synthetase in NIH/3T3 cells (53). One way to approach this prohlem is to isolate and characterize cell variants lacking the enzymatic activities involved in the mechanism of IFN action. In the present study. we showed that treatment of t h e n,MIFN-resistant FLC variants with yIFN preparations induces a hona fide antiviral state. The extent of reduction of I-ytic viruses yield induced hy yIFN in the n,/3IFN-resistant FLC variants (1-2 logs) is lower than that induced in wild-type FIX (3-4 lops), hut it is spcific.
Likewise, yIFN treatment is also ahle to reduce the chronic release of FLV in sensitive and resistant F I X even if t h e extent of reduction is lower in resistant cells. It is noteworthy to keep in mind that the quantitative aspects of this effect in resistant wrsus susceptihle FLC are hardly comparahle. hecause the reverse transcriptase assay used to test FI,V release is sensitive to detect only l-log differences hetween IFNtreated and untreated cells, as the remaining tail of the curves tends asymptotically to zero.
Evidence has heen provided that the tr,dTFS antiviral effect may he effective on retroviruses hut not on lytic viruses and vice versa (36, 54). As retroviruses estahlish a svmhiontic relationship with the host cell, it is conceivahle that IFN inhibition of murine leukemia virus release is mediated hv pathways different from those involved in the case of lytic RNA viruses. In fact, Sen and Hertz (36) as well as Salzherg et ai. ( 5 5 ) have descrihed a suhclone of murine leukemia virustransformed NIH/3T3 cells which has retained susceptihility to n,/jIFN anti-murine leukemia virus activitv hut is poorly responsive to its anti-EMCV and -\XV activity (5.5). Little, if any, induction of 2-5A synthetase activity over the hasal level was ohserved in these resistant cells treated with n,HTFN, suggesting that the 2-5A synthetase pathway is not involved in IFN-induced inhibition of chronic retrovirus release. Sen and Hertz (36) suggest that hoth 2-5A synthetase and protein kinase pathways do not play a direct role in the antiviral activity of IFN on retroviruses. In the FLC system, treatment with doses of a,PIFN up to 13,000 units/ml is not able to induce any 67-kDa protein kinase activity in the a$-resistant FLC clones. Thus, both these enzymatic pathways appear not to be activated in these variants treated with a,PIFN. In contrast, treatment with graded doses of yIFN (up to 500 units/ml) is unable to induce 2-5A synthetase activity, but is still able to induce the 67-kDa protein kinase activity.
These data are compatible with the following: ( a ) the 67-kDa protein kinase activity may be responsible for the observed antiviral state; and (b) activation of the 2-5A synthetase is not strictly required for this effect. As the reduction of virus yield induced by yIFN treatment is lower than that observed in wild-type FLC, the observed activation of the 67-kDa protein kinase may not be sufficient to completely inhibit virus replication. This is in agreement with the accepted hypothesis that the two different enzymatic pathways work independently and/or additively in the same direction. In addition, the uninducibility of 67-kDa protein kinase activity observed in the resistant FLC treated with a,PIFN suggests that the activation of this enzymatic pathway could be mediated by different intermediate steps subsequent to IFNs binding to their proper receptors.
Finally, the possibility that other IFN-induced proteins (25), in addition to those analyzed here, are involved in the mechanism(s) underlying the observed effects cannot be ruled out.