Induction of Interferon in Chick Cells by Temperaturesensitive Mutants of Sindbis Virus

Transcription

1 J. gen. ViroL 0974), 25, o Printed in Great Britain 38I Induction of Interferon in Chick Cells by Temperaturesensitive Mutants of Sindbis Virus By G. J. ATKINS, M. D. JOHNSTON, LINDA M. WESTMACOTT AND D. C. BURKE Department of Biological Sciences, University of Warwick, Coventry, CV4 7AL, England (Accepted I4 August 1974) SUMMARY Interferon induction by I I temperature-sensitive (ts) mutants of the HR strain, and 38 ts mutants of the AR339 strain of Sindbis virus was investigated. All HR and AR339 mutants induced interferon at 3 C. Induction by the mutants at the restrictive temperature (39 C) was dependent on the time of incubation, and, in some instances, on the input multiplicity. At an input multiplicity of 5, and after an incubation time of 16 h at 39 C, induction by the AR339 mutants was dependent on virus RNA synthesis. All HR mutants showing detectable RNA synthesis at 39 C induced interferon, but of those showing undetectable RNA synthesis at 39 C, three induced interferon and three did not. At 42 C, RNA synthesis by the HR wild-type and all ts mutants was depressed, and only one mutant and the wild type induced interferon. Temperature shift experiments showed that interferon induction was dependent on an early virus function. It is postulated that interferon induction is dependent on a low threshold level of RNA synthesis occurring early in infection. INTRODUCTION One method of investigating the mechanism of interferon induction by viruses is to use conditional lethal mutants. Mutants defective in a virus function necessary for interferon induction would induce interferon under permissive conditions but not under restrictive conditions. This approach has been used in three studies of interferon induction by temperature-sensitive (ts) mutants of RNA viruses. Lockart et al. 0968) investigated four ts mutants of Sindbis virus, originally isolated and characterized by Burge & Pfefferkorn (I966a, b). These mutants were all defective in a different function, three being RNA + (able to make virus RNA at restrictive temperature) and one being RNA- (unable to make virus RNA at restrictive temperature). All four mutants induced interferon in chick cells at 29 C but not at 42 C, although the wild-type virus induced interferon at both temperatures. Since virus RNA synthesis was detectable in cells infected with the RNA + mutants at 42 C, Lockart et al. (I968) concluded that virus RNA synthesis was not sufficient to account for interferon induction. Lomniczi & Burke (197o) studied interferon induction in chick cells by 2I ts mutants of the closely related Semliki Forest virus (SFV). It was found that, at high input multiplicities (> 25 p.f.u./cell), all mutants but one were inducers at the restrictive temperature. At an input multiplicity of 5, RNA+ mutants were able to induce interferon at restrictive temperature, whereas RNA- mutants were not. Thus, it appeared that interferon induction at low input multi-

2 382 G.J. ATKINS AND OTHERS plicities was dependent on virus RNA synthesis. However, no attempt was made to measure leak or reversion to wild-type of the mutants during interferon induction, or to correlate the amount of virus RNA synthesis with interferon induction. Thus, apparently conflicting results were obtained in the two studies. Lai & Joklik (1973) investigated the ability of nine reovirus ts mutants to induce interferon in mouse L cells. All mutants were defective in their ability to induce interferon at the restrictive temperature. The only virus function which correlated with interferon induction was the formation of intact virus particles. In the present study we have investigated interferon induction by eleven ts mutants of the HR strain of Sindbis virus (Burge & Pfefferkorn, I966a), and by 39 ts mutants of the AR339 strain of Sindbis virus (Atkins, Samuels & Kennedy, 1974). METHODS General techniques. All materials, method of measuring total RNA synthesis, method of plaque assay and temperature control, were described in the previous paper (Atkins et al. 1974). The AR339 mutants are described in the previous paper (Atkins et al. I974). The 39 mutants used in this study were chosen for their low or undetectable reversion frequency at 39 C. The HR mutants are described by Burge & Pfefferkorn (I966a). All mutants were cloned before use by picking single plaques and first passage stocks of these clones used in all experiments. Interferon assays. A modified version of the inhibition of nucleic acid synthesis (INAS) method described by McWilliam et al. (I97I) was used. Chick cells were seeded in 2 ml growth medium at a concentration of lo s cells/ml in 2 cm diam. glass vials (Pye-Unicam Ltd). They were incubated overnight in a humidified atmosphere of 5 % CO2[air at 37 C. Dilutions of the interferon sample (from lo -1 to lo -3.5 in steps of IO -o.5 ) were made in maintenance medium to duplicate monolayers. The monolayers were drained and I ml of each of the interferon dilutions added to duplicate monolayers. The samples were incubated overnight, and drained before the addition of 5o p.f.u./cell of SFV in I ml of maintenance medium plus I #g/ml actinomycin D as challenge virus. The monolayers were incubated for 3 h at 37 C before being drained, and I ml of maintenance medium containing I #g actinomycin D and o'5 #Ci [~H]-uridine added to each sample. After a further 3 h incubation at 37 C the monolayers were washed once with ice-cold phosphatebuffered saline (PBS), twice with ice-cold 5 % trichloroacetic acid, and once with ice-cold ethanol. They were then dissolved in 2oo/*I tissue solubilizer (I/3 dilution of Soluene (Packard Instruments Ltd) in toluene) and ioo #1 from each duplicate vial added to one scintillation vial. After the addition of 8 ml of toluene-based scintillation fluid, acidified with o.i % glacial acetic acid, the samples were counted in a Packard liquid scintillation counter. The counts were plotted against log10 (interferon dilution), the interferon titre being taken as - loglo (interferon dilution) at which the incorporation of label into virus RNA was reduced by 5 o ~o (in units of log10 INASso ). Uninfected and infected control cultures without added interferon, and an internal interferon standard, were included in all assays. As the first dilution in the assay was IO -1, interferon titres below one log10 (INASso) were regarded as insignificant. The titre of the reference research standard A for chick interferon (62/4), containing IOO research standard units of chick interferon (obtained from the National Institute for Biological Standards and Control, Holly Hill, Hampstead, London NW3 6RB) was found to be 2"72 logio (INASso) by this method. Interferon induction. Chick cell monolayers, set up the previous day in 3o ml screwcapped bottles and at a concentration of 3 lo6 cells in 3 ml of growth medium, were left

3 Mutant AR339 wild-type F36 (RNA+) ~: RI9t (RNA-) ~: HR wild-type ts 5 (RNA+):]: ts 7 (RNA-)~: Interferon induction by ts mutants 383 Table I. Dependence of interferon induction on input multiplicity Induction Input- Titre at Titre at Interferon temperature multiplicity 30 C 39 C titre ( (2) (p.f.u./cell) (p.f.u./ml) (p.f.u./ml) (Iog~o INASs0)* 39 0"5 5"0 io' 4"0 x ios 2"65 5 2"3 IOS 7"7 X IO 5 2"2 5 7'8 X IO s 5'5 los 2" "5 i-3 x IOs 1.2 x ios 2" X 107 2" 5 5 4"I X 107 3"5 X I07 2" "5 2"8 I04 ut 2"35 5 4"7 X to 5 u 2'95 5 3'2 X IO' U 2"7 30 0"5 5"5 x lo 7 u 2"7 5 5'o x lo 7 u 2"0 5 4"I IO "t u 2"2 39 0'5 7"I Io 4 u u 5 9"3 lo4 u u 50 4"0 x Ios u "5 4'3 IO' u "4 1o' u z' x IO ~ u o'5 5'6 lo x to 8 3"o 5 I'6 IOa 2"2 X IO ~ 3"3 50 1"3 x IO 8 1"9 x Io ~ o'5 1"3 Io 5 4"0:< IO 3 1"7 5 2"7 I0 ~ 4"0 x IO ~ '2 IO s I'OX 103 2" x Io ~ 3'0 x 103 < I 5 3"0X IO 3 5"0X IO 3 < I 50 1"8 104 I'0 IO s < I * Interferon was harvested I6 h after infection at 39 C, or 24 h after infection at 30 C. t Undetectable. RI9I and F36 are ts mutants of the AR339 strain, ts 5 and ts 7 of the HR strain. at 4 C for 15 rain. One ml of pre-cooled virus suspension was added to each monolayer, and the virus allowed to adsorb at 4 C for I h. The inoculum was removed, and 5 ml of pre-warmed maintenance medium added. The bottles were incubated in a water-bath set to 39 (+ 0"05) C or in a 30 C incubator. After the appropriate time, the fluidswereharvested and the virus inactivated by heating to 65 C for 2 h (heat-inactivated virus was not an interferon inducer). All inductions were carried out in duplicate, and the samples pooled before assay. Temperature shift-up experiments. Monolayers were infected as described above and incubated at 30 C. At ½ h intervals after infection, the monolayers were transferred to a 39 C waterbath for the appropriate time. RNA synthesis was measured in the presence of actinomycin D, and after a pulse of I/zCi/ml of [3H]-uridine in 5 ml medium (consisting of Earle's medium +2 % dialysed calf serum), 2½ to 4½ h after infection. Interferon was harvested I6 h after infection. Fluid for plaque assay was collected i6 h after infection, the monolayer having been washed twice with warm PBS 4 h after infection and fresh warm maintenance medium added.

4 I 384 G.J. ATKINS AND OTHERS I I I AR339 wild-type I I. H107 (RNA ) I F294 ' (R'N A - )~ ' < Z _0 -E HR wild-type ts 2 ( R N A ~ w ts 7 (RNA-) O A Time after infection (h) Fig. I. Time course for the induction of interferon by the HR and AR339 wild-types, and one RNA + and one RNA- mutant for each strain. Q Q, at 39 at 30 C. RESULTS Effect of input multiplicity Table I shows the effect of the input multiplicity on interferon induction by the two wildtypes, and an RNA- and RNA + mutant for each strain. The interferon was harvested at 16 h after infection at 39 C and at 24 h after infection at 30 C (see below). For the AR339 strain, interferon induction by the wild-type and the RNA + mutant at either temperature or by the RNA- mutant at 30 C, was unaffected by the input multiplicity. For both AR339 mutants, however, at an input multiplicity of 5o, leak at 39 C was greatly increased over that at input multiplicities of 5 and o'5. The interferon induced by the RNA- mutant at 39 C and at an input multiplicity of 50 can therefore probably be ascribed to increased leak. No reversion was detected for the two AR339 mutants at any input multiplicity. This multiplicity effect was not found for the HR mutants, although reversion for these mutants was high (Table I). A constant input multiplicity of 5 was used in all subsequent experiments. Effect of incubation time Fig. I shows the effect of incubation time on interferon induction for the HR and AR339 wild-types, and for one RNA + and one RNA- mutant for each strain. For both wild-types

5 Interferon induction by ts mutants 385 Table 2. Induction of interferon by mutants of the HR strain Induction Yield assayed Yield assayed RNA synthesis Interferon temperature at 39 C at 30 C (% of wild- titre Mutant* ( C) (p.f.u./ml) (p.f.u./ml) type at 39 C) (loglo INASs0)t ts I 42 4.o lo s < I 39 5"0 Io 5 4"6 IO e < I 2"I IO s 2'1 ts IO2 < I lo 5 5.ox io 5 < I "4 IO 8 2"25 ts7 42 4"6x io 2 < t 39 2"9x I05 2"i X I0 ~ < I < I 30 9'0 x i o $ 2. i ts II 42 3.i lo 2 I x io s 2.6 lo 6 < I 1" '0 IO a 2.0 ts "o < I 39 1"2 IO a 8-0xi02 < I < 1 3o 2.0 x ios 2'4 ts 2I x io s < I 39 I'2 x x io 5 I'O < I 30 5" '45 ts 'o 1o s I'2 < I 39 7"O lo 3 4"ox io 5 53"5 2" ox lo s 2'I ts IO I'O < I IO a 2"0 IO x 10 a 2"6 ts2 42 7"0X I0 s 2'8 < I 39 7"0 x to 3 1"5 x IO* " "0 x IO 7 2"2 ts "9 IO s 5"I < I 39 3'0 IO 5 1"2 X IO 5 62"3 3' x I0 s 2"85 ts ox lo s < i < i 39 3"3 Io x to i.o 30 5"0 IO s 2"4 Wild- 42 I'O IO s 5'2 1"6 type X 109 1"3 x 109 I00 2"3 30 3"9 X I0 s I'9 * Mutants arranged in order of level of RNA synthesis at 39 C. Interferon was harvested at 16 h after infection at 39 C, or at 24 h after infection at 30 C and 42 C. and both RNA + mutants, interferon induction had occurred by 16 h after infection at 39 C, although some delay in induction was observed for the HR RNA + mutant (ts 2). For the RNA- mutants at 39 C, induction had occurred at 18 h after infection with the AR339 mutant (F294), but not by 24 h after infection with the HR mutant (ts 7). At 3o C, interferon induction had occurred by 21 h, or earlier, after infection for all mutants. In subsequent experiments, interferon was harvested i6 h after infection at 39 C and 24 h after infection at 30 C. 26 VIR 25

6 386 G. J. ATKINS AND OTHERS Table 3. Induction of interferon by mutants of the AR339 strain RNA Induction Yield assayed Yield assayed synthesis Interferon temperature at 39 C 3o C (% of wild- titre Mutant* ( C) (p.f.u./ml) (p.f.u./ml) type at 39 C) (loglo INAS~o)~" A20I 39 1'9 >( x lo 4 < I < I lo lo 7 1'5 HI75 39 u:~ 6"7 lo 3 < I < I 30 u 7"0 IO HII9 39 u 3.8 io 4 < i < I 30 u 4'6 lo r I'55 AI5I 39 u 1.6 IO 3 < i < I IO x IOn 1-3 F u 3"6x IO z < I < I 30 3"0 X 102 2"2 X 107 2"45 RI9I 39 u I.I IO 4 < I < I 30 u 7" "7 FIo4 39 u 6"4 ]o 3 < i < i 30 tl 1.6 ios 2-1o N2 39 u 4'2xI& < I < "0 x io ~ 3"8 lo 6 2"35 N ox IO I 5.4 x IO n < 1 < I 30 u 3"2 lo 6 1"95 H75 39 u 1"8 x lo 4 1"4 I.O 30 u 2.6 lo 6 I.O E u 1'7 x < I 30 u 5.2 x IO 6 2'55 E48 39 u I'I lo 5 2'0 < I 30 u 7.2 x i& 1.9 F u 5"6 x lo o 5 30 u I-I x IO 7 2'35 N x lo 2 1'I IO 4 2" u I'4 x IO 7 2'4 H x lo 4 4'9 lo " x 107 1"55 A82 39 u 2"6 I03 3"3 I'2 30 u 7"0 IOt; 2"5 N "0 x Io z 3'4 lo4 3"8 2'0 30 u 3"2 x lo 7 2"7 AI91 39 u I'7 x lo 4 4"3 1"65 30 u 7"5 x lo s I'25 A u I.I x l& 4"6 1"o5 30 u 8"4 I0t; 2"85 F u 1.6x IO a 5.0 < I 3 U I'I x 10 t; 2"2 N7 39 u 2"O x I02 5"4 I'85 30 u 1.4 x IOs 3' I AI lo 2 5"9 x IO :~ 6"2 < I 30 u 2"6x IO 7 2"I NI7 39 u 4.2 x IO a 8"5 I'8 30 u I'4 x IOt; 2'45 H5I 39 u 2. 4 x lo g IO.O u 2"7 I0 ~ 2'6

7 Interferon induction by ts mutants 387 Table 3 (cont.) RNA Induction Yield assayed Yield assayed synthesis Interferon temperature at 39 C 30 C (% of wild- titre Mutant* ( C) (p.f.u./ml) (p.f,u./ml) type at 39 C) (log10 INASs0)t H76 39 H~ I'4 x 108 I2.1 I'3 30 u 3"2 10 n I'5 AI58 39 u I'5 x IO 4 13'3 2'75 30 u I " HI8 39 u 6"6 Io s I6" U I'I X 107 2" 7 N32 39 u I'o x IO ~ I8'4 2'65 3o u 2"8 IO n 2"65 HI5o 39 u 4"6 IO 4 25"o "o Io 2 I' 'I H39 39 u 1-4 IO ~ 28"3 2"25 3 I1 I "9 107 I'2 HIo7 39 u 2-0 Io ~ 4o'8 1'9 30 u 3"0 x io 7 I'75 FI27 39 u 1'6 IO 3 4o'9 2"65 30 u I '6 x 106 2"7 H u 3'o IO ~ 4I'o I"5 30 u 4"5 IO6 1.8 H "ox lo 8 5'9 I0a u 5'4 x IO 7 3"0 HI78 39 u 7"2 IO 4 72"0 2'0 3o u 9"I 106 2"o5 HI34 39 u 1"4 lo 4 81" o u 6-I IO 7 1"6 HI91 39 u 3"2 ] "4 3o u 3'3 IO7 2"45 F u 1.2 IO ~ I38 2'I5 30 u 6.6 X IO ~ 2"0 F36 39 I'0 IO a 2"8 X I0 ~ 156 I'9 3 I "5 X IO 4 2'2 X IO 7 2'I 5 * Mutants arranged in order of level of RNA synthesis at 39 C. t Interferon was harvested at I6 h after infection at 39 C, or at 24 h after infection at 3o C. :~ Undetectable. RNA synthesis by HR mutants The values obtained for the level of RNA synthesis by the HR mutants at 39 C were in general agreement with those obtained by Burge & Pfefferkorn 0966a) (Table 2). At 42 C the levels of RNA synthesis for the wild-type and the RNA + mutants were much lower than at 39 C. Interferon induction by HR mutants Interferon induction at 39 C and 3o C was measured using six RNA- mutants and five RNA + mutants of the HR strain (Table 2). Fluids incubated at 3o C were plaque assayed for virus at 3o C only, before heat-inactivation of the virus and interferon assay. Those incubated at 39 C were plaque assayed at both 3o C and 39 C to determine the amount of leak and reversion. The yield of virus able to form plaques at 39 C is a measure of the number of wild-type revertants. The yield due to leak is estimated by subtracting the yield assayed at 39 C from that assayed at 3o C. At 39 C, five RNA+ mutants (ts 2o, xo, 2, 5 and 23) and the wild-type induced interferon 26-2

8 388 G. J. ATKINS AND OTHERS ~, 6 5 o o 4 Wild-type.' 30 C.C.,~d %..- N74 ' H194 J i, 30 C--~c? 30 C --t~q ~ I I I I F294 :N74 ~ H194 0,, I I I 1, ~ ~ild-type F294 ~. N 7 ~ H C-a~, 2Z= 2. 5 ~ 30oc,,~ l'0 I I - - I I! I i I Time at 30 C (h) Fig. z. Temperature shift experiments with AR339 ts mutants. For all experiments, the temperature was shifted from 30 C to 39 C at the times after infection indicated on the horizontal axis. Infectious virus was collected 4 to 6 h after infection and plaque assayed at 30 C, RNA synthesis was measured z½ to 4} h after infection, and interferon harvested 16 h after infection. RNA synthesis is expressed as a percentage of the level at 30 C in the same experiment. I6 h after infection, while three of the RNA- mutants (ts i, 6 and I I) induced interferon, but three others (ts 7, 15 and 2 I) induced no significant yield of interferon. The yield of virus at 39 C was largely due to reversion of the mutants to wild-type. At 30 C, all mutants and the wild-type induced interferon 24 h after infection. At 42 C, the wild-type induced interferon but the yield of infectious virus was much lower than at 39 C. Only one mutant (ts r I) induced interferon at 42 C. Interferon induction by AR339 mutants Interferon induction at 39 C and 30 C was measured using 39 ts mutants of the AR339 strain (Table 3). Nine mutants showed no detectable RNA synthesis, while the rest showed levels of RNA synthesis at 39 C ranging from 1% to greater than ioo % of the wild-type level (Atkins et al. I974). Interferon samples were plaque assayed at 3 o C and 39 C before the virus was heat-inactivated. All nine RNA- mutants induced no interferon at 39 C I6 h after infection. Mutants showing high levels of RNA synthesis at 39 C (> Io % of wildtype), and the wild-type, induced interferon at 39 C. Of the 14 mutants showing I to Io % of the wild-type level of RNA synthesis at 39 C, IO induced interferon at 39 C whereas 4 did not. Reversion to wild-type was undetectable for most mutants, the virus yield at 39 C being due to leak in most cases. No correlation could be found between production of infectious particles and interferon induction. Temperature shift-up experiments These were carried out for three AR339 mutants and the wild-type (Fig. z). For the wildtype, shifts from 30 C to 39 C during the first 2½ h after infection had no significant effect

9 Interferon induction by ts mutants 389 on the formation of infectious particles, RNA synthesis, or interferon induction. For the RNA- mutant (F294), an increase in the number of infectious particles (produced 4 to 6 h after infection) was obtained if the temperature was shifted at any time subsequent to ~ h after infection. However, no large increase in RNA synthesis (measured 2½ to 4½ h after infection) was obtained if the temperature was shifted at any time during the first 2½ h after infection. A significant yield of interferon was obtained from this mutant if the shift was carried out at any time subsequent to I h after infection. For N74, a mutant showing 3"8 % of the wildtype level of RNA synthesis at 39 C, results similar to F294 were obtained for infectious particles and RNA synthesis. However, this mutant does induce some interferon at 39 C, and only a limited increase in interferon yield was obtained if infected cultures were incubated at 3o C before being shifted to 39 C. For HI94, a mutant showing 57.z % of the wild-type level of RNA synthesis at 39 C, results similar to those given by the wild-type were obtained for RNA synthesis and interferon induction. No increase in production of infectious particles was observed, however, unless the shift to 39 C was carried out later than 3 h after infection. DISCUSSION Before the mechanism of interferon induction by ts mutants can be related to the mechanism occurring in cells infected with the wild-type virus, two factors have to be considered. The first is the reversion of the mutant to wild-type, and the second is the leakiness of the mutation. Both of these processes could lead to induction of interferon at restrictive temperature by a similar mechanism to that occurring in wild-type infected cells. The results obtained in this study underline these difficulties. Interferon induction by mutants at high input multiplicities could be due either to an increased leakiness of the mutation, as shown in this study, or to an increase in the number of cells infected by wild-type revertants. This could well explain the results obtained by Lomniczi & Burke 0970) who observed an increase in interferon yield on increasing the input multiplicity of cells infected with ts mutants of SFV at the restrictive temperature. Similarly, the incubation time at the restrictive temperature would affect the induction of interferon by the processes of leak and reversion. In the present study, we have attempted to select a time and an input multiplicity at which interferon induction had occurred in cells infected by the wild-type virus, but at which leak and the multiplication of revertants had not led to the induction of interferon in mutantinfected cells. The results obtained for the HR mutants show that no clear distinction could be drawn between RNA + and RNA- mutants with regard to interferon induction. While all RNA + mutants induced interferon, three of the RNA- mutants induced interferon whereas three did not. However, since reversion for all the mutants was high, we cannot exclude the possibility that interferon induction was due to reversion in some cases. The results obtained with the AR339 mutants are more definitive. These mutants show lower reversion than the HR mutants, although leak does occur. Interferon induction at restrictive temperature for these mutants is generally dependent on RNA synthesis. However, for those mutants showing low RNA synthesis at 39 C (< to % of the wild-type level), the situation is complex; some of these induce interferon, others do not. This may be an indication of a low threshold level of RNA synthesis necessary for interferon induction in the virusinfected cells. The results obtained by Lockart et al. 0968) can possibly be attributed to the use of 42 C as the restrictive temperature. Our results show that, at 42 C, both the formation of infectious particles and RNA synthesis by the wild-type virus was substantially reduced. The absence of interferon induction by most of the ts mutants can probably be explained on this

10 39 G. J. ATKINS AND OTHERS basis; RNA synthesis by the RNA + mutants, already reduced compared to the wild-type at 39 C, would not reach a sufficiently high level at 4z C to induce interferon. The one mutant (ts I I) which induced interferon at 42 C shows the highest reversion frequency of all those tested. Our results are difficult to compare with those of Lai & Joklik (I973), who studied interferon induction by ts mutants of reovirus, because of the unrelatedness of the two viruses used. However, we do not observe any correlation between the formation of infectious particles and interferon induction, as was observed by these authors. The temperature shift experiments showed that the kinetics of interferon induction for an RNA- mutant followed approximately those of the production of infectious particles although this was not so for an RNA + mutant having a late block in the production of infectious particles. This is further evidence for the dependence of interferon induction on RNA synthesis, and may be a further indication of a threshold level of RNA synthesis, necessary for interferon induction, occurring early in infection. Direct measurement of such low levels of RNA synthesis is complicated by the fact that incorporation of label into RNA decreases as soon as the temperature is shifted to 39 C. This is in contrast to the result obtained by Lomniczi & Burke (197o), and may indicate either the presence of a temperaturesensitive RNA polymerase, or that the formation of the enzyme is temperature-sensitive and its turnover is rapid. Further evidence for the existence of a threshold level of RNA synthesis necessary for interferon induction could be obtained by following the kinetics of interferon induction by mutants showing low levels of RNA synthesis at 39 C. Mutants whose level of RNA synthesis at 39 C was below the threshold level would be expected to show a delay in interferon induction. Such experiments would also show whether leak or reversion was responsible for interferon induction in some mutants. Thus, although the present study shows that RNA synthesis is probably necessary for interferon induction by Sindbis virus, we do not yet know whether total RNA, one or more RNA species, or a protein coded by the RNA, is the primary interferon inducer. The possibility that the inducer is virus-specified double-stranded RNA is presently under investigation. We would like to thank Mrs C. Lancashire and Miss F. Karachiwalla for skilled technical assistance, and Dr S. I. T. Kennedy for his interest. The work was supported by grants from the Science Research Council, Medical Research Council and the Cancer Research Campaign. REFERENCES ATKINS, G. S., SAMUELS, J. & KENNEl)Y, S. L T. (t 974). Isolation and preliminary characterization of temperaturesensitive mutants of Sindbis virus strain AR339. Journal of General Virology 25, o. BURGE, B. W. & PFEFFERKORN, E. R. (I966a). Isolation and characterization of conditional-lethal mutants of Sindbis virus. Virology 3o, 2o BUgGE, B.W. & PFEFFERKORN, E.R. (I966b). Complementation between temperature-sensitive mutants of Sindbis virus. Virology 30, LAX, M. X. & SOKLIK, W. K. (I973). The induction of interferon by temperature-sensitive mutants of reoviruses, UV-irradiated reovirus, and subviral reovirus particles. Virology 5x, I LOCKART, R. Z. SUN., BAYLISS, N. L., TOY, S. T. & YIN, F. H. (I968). Viral events necessary for the induction of interferon in chick embryo cells. Journal of Virology 2, LOMNICZr, B. & BURKE, D.C. (I970). Interferon production by temperature-sensitive mutants of Semliki Forest virus. Journal of General Virology 8, McWlLLIAM, M., FINKELSTEIN, M. S., ALLEN, P. T. & GIRON, O. J. 097I). Assay of chick interferon by the inhibition of viral ribonucleic acid synthesis. Applied Microbiology 21, I. (Received 20 May I974)