Determinants of Spontaneous Recovery and Persistence in MDCK Cells Infected with Lymphocytic Choriomeningitis Virus

Transcription

1 J. gen. Virol. (I979), 44, ti3-12i Printed in Great Britain II 3 Determinants of Spontaneous Recovery and Persistence in MDCK Cells Infected with Lymphocytic Choriomeningitis Virus By S, JACOBSON, F. J. DUTKO* AND C. J. PFAU Department of Biology, Rensselaer Polytechnic Institute, Troy, New York I218I, U.S.A. (Accepted I8 December i978) SUMMARY MDCK cells that normally would have been killed by standard lymphocytic choriomeningitis (LCM) virus were saved either by pre- or coqnfection with defective interfering (DI) virus. The ability of these spared cells to produce virusspecific antigen (as well as infectious virus) and resist being killed by standard virus challenge was followed for at least 35 days. During this period both types of cultures displayed unique cycling patterns for the above characteristics. The most striking difference was the longevity of the infections. Cultures exposed to DI particles prior to standard virus became persistently infected, while co-infection with both virus types led to spontaneous curing with no trace of the previous infection. The basis for these dissimilar outcomes was traced to a hitherto undetected non-defective LCM virus (called SP) in the DI virus stocks used to preinfect MDCK cells. SP virus was not present in standard virus stocks but arose in long-term persistently infected L cells that had been initially infected with standard virus. Cloned SP virus shared species-specific antigens with standard virus, was resistant to inhibition by DI virus and was capable of turning self-curing cultures into cultures persistently synthesizing both DI and SP virus. INTRODUCTION The search for the biological principles allowing life-long infection of mice with LCM virus has steadily gained momentum since the initial discoveries were made over 4 years ago. It has become increasingly clear that the ability of the virus to be maintained as a persistent infection involves an intricate and fascinating weave of virological and immunological phenomena (Pfau, 1978). Virus persistence is based on the fact that these mice either develop no humoral response against the virus or one lacking neutralizing and complementfixing activity. These mice also mount no cell-mediated immunity necessary to recognize LCM-infected tissue. Furthermore, they must possess a mechanism for blunting the cytolytic potential of the virus (Dutko & Pfau, I978 ). The latter appears to be accomplished by the rapid genesis of defective interfering (DI) particles (Welsh & Pfau, I97Z; Popescu & Lehmann-Grube, I977). DI virus multiplies rapidly in the neonatal mouse within the first week after injection with standard self-replicating virus. This heralds a profound change in the plaque morphology of standard virus. When virus stocks of predominantly clear plaque morphology are injected into the neonates, peak titres are reached rapidly only a few days before maximum DI virus * Present address: Department of Immunopathology, Scripps Clinic and Research Foundation, La Jolla, California 92037, U.S.A.

2 II4 S. JACOBSON, F. J. DUTKO AND C. J. PFAU activity (Popescu & Lehman-Grube, 1977)- In most organs of these carrier mice there is then a precipitous drop in both DI and clear plaque standard virus with the latter being replaced by turbid plaque types. The significance of the turbid plaque types was not established beyond noting that in vitro they produced less DI virus than clear plaque types (Popescu & Lehmann-Grube, 1977). In the course of following LCM virus synthesis in 'low interference' cells, i.e. cells that are non-permissive in the induction but not in the synthesis or expression of DI particles, a self replicating difficult-to-plaque LCM virus was discovered. Some of the properties of this newly cloned virus, as well as data showing its indispensable role in maintaining persistent infection are presented here. This tissue culture system is also discussed as a model to explain murine infections where selective pressure brought about by DI particles may lead to the rapid evolution of new virus types. METHODS Cells and virus. The origin and propagation of MDCK cells and the UBC strain of LCM virus have been described previously (Dutko & Pfau, I978). The standard LCM virus stock used in these studies was harvested from L cells 27 h after an input m.o.i, of o" ~. This virus stock was relatively free of DI particles as judged by its high MDCK cell-killing potential (Dutko & Pfau, I978 ). Stocks of LCM DI particles were harvested 27 h after a medium change, from persistently infected L cell monolayers that had just reached conftuency. These persistently infected L cells had ceased to produce standard virus (the UBC strain) by the 9znd generation post-infection (Pfau, 1977). They had been frozen at the 129th generation p.i. and stored in liquid nitrogen since I97I. The cells were then carried in continuous culture for another 24o generations during the course of these studies. A freshly generated line of persistently infected L cells was also initiated. Cells were infected at an m.o.i, of o'l with cloned LCM-UBC virus and maintained as previously described (Pfau, I977). Plaque p.f.u, assays. The BHK2I/I3S agarose suspension assay was the one routinely used in our laboratory (Pulkkinen & Pfau, I97o). The MDCK assay was carried out as follows: confluent monolayers (4"5 x io n cells) in 6o mm plastic Petri dishes were inoculated with I ml of a virus dilution containing 50/zg DEAE-dextran (which increased plaque counts by I log~0 unit). After 1.5 h at 37 C the virus was removed by decantation. Each surface was then overlaid with a solution containing 2 ml of 1% agarose, 2 ml of 2 x Dulbecco's modification of Eagle's medium (G.B.I., Grand Island, N.Y., U.S.A. catalogue no. H-z0, 0"5 ml of heat-inactivated foetal calf serum and 0"5 ml of sodium bicarbonate (40 g/l). After incubation for 4 days at 37 C in a humidified atmosphere of 4 % CO2 in air the monolayers were counter-stained with neutral red. Plaques were counted 24 to 36 h later. MDCK cell-killing assay. The assay, described in detail by Dutko & Pfau (I978), is based on quantitative comparison of the colony-forming ability of normal cells with that of cells infected with various dilutions of virus. For the experiments described here the percentage cell-killing in persistently infected MDCK cells was derived from the colony-forming ability of these cells versus normal cells after exposure to the same standard (high cell-killing) challenge virus stock. Preparation ofantisera. Syrian hamsters weighing IOO g were injected subcutaneously with mixtures of equal vol. of LCM virus and Freund's complete adjuvent (o'3 ml total injection vol.). The MDCK plaque-purified virus for these injections contained I. t or 24.0 x io 6 p.f.u./ ml of the newly identified or standard virus, respectively. Two weeks after the third bimonthly injection the animals were exsanguinated and the serum stored at -2o C until used. lmmunofluoreseent staining procedure. The general procedure for cytoplasmic staining has been described (Staneck et al. I972). For the direct test FITC-labelled antibody against the

3 Persistence in LCM virus infected cells I I5 Fortner strain of LCM was used. This was obtained from Microbiological Associates (Walkersville, Md., U.S.A.), and was routinely diluted I/2o. For the indirect test conjugated LCM hyperimmune serum from hamsters was used along with FITC-labelled (IgG) goat anti-hamster gamma globulin (from Microbiological Associates). Both sera were diluted I/4 O before use. Indirect immunofluorescent staining of antigens on the surface of viable infected MDCK cells was performed with the same antibodies (diluted 1/4) as described by Buchmeier & Oldstone (i978). Radioimmunoassay. Virus was concentrated by polyethylene glycol and centrifuged into the interface between a zo and 5o % sucrose cushion (Dutko et al. r976). It was then diluted tenfold in phosphate buffered saline (minus Ca and Mg) and pelleted by centrifugation in a Spinco SW4I rotor for 3 h at 35ooo rev/min. Virus was resuspended in 0"5 ml PBS and labelled by reductive methylation with 14C-formaldehyde as described by Rice & Means (I97)) except that the ph of the solution was initially at 7-2 and sodium cyanoborohydride was used instead of sodium borohydride. After overnight incubation at 4 C approx. I ml of this solution was layered on top of an 18 to 40 % Renografin density gradient (Gschwender et al. I975) and centrifuged for 2 h at 35ooo rev/min in a Spinco SW5o.I rotor. Fractions from the gradient containing peak radioactivity were diluted in Eagle's medium containing o'5 % bovine serum albumin to a concentration giving Ioooo to 12ooo ct/min/o.2 ml. Virus was then mixed with hamster antiviral antibody and Staphylococcal protein A as described by Jahrling et al. (I978). RESULTS Initiation of persistently infected cultures Two types of persistently infected MDCK cultures were established. One type, called DS, was infected with an LCM-DI virus stock before challenge with lytic (standard) virus. The other type, MS, had no pre-treatment prior to inoculation with a lytic virus stock which contained a small amount of DI virus (Dutko & Pfau, I978). As predicted from previous observations (Dutko & Pfau, I978 ) cells in the DS culture showed no signs of infection, while most of the cells in the MS culture were destroyed within 48 to 72 h. The few remaining cells in the latter culture represented no more than Io % of the original population and continued thereafter to divide with no further visible cytopathology. Those cells re-populated the growth surface of the flask by the Ioth day (with two to three intervening media changes). Beyond this point both types of cultures reached the same degree of confluency at the end of each 4-day subcultivation interval. However, unlike normal MDCK monolayers which did not require a medium change between dispersals with trypsin, both types of cultures required fresh media daily. Virus synthesis in persistently infected cultures as determined by plaque assay and immunofluorescence Using our customary BHKzI/I3S agarose suspension assay (Putkkinen & Pfau, I97 o) the p.f.u, titres in both DS and MS cultures were found to peak within two days after infection (Fig. I). At this time the yield of standard virus in the MS culture was over one log10 unit higher than in the DS culture (pre-treatment of cells with DI virus has been shown to be more effective in inhibiting standard virus than simultaneous infection - Welsh et al: I972). Titres declined sharply after this, reaching less than ~o ~ p.f.u./ml on days 6 and Io p.i. for the DS and MS cultures respectively. In the DS culture this signalled the permanent shutdown of BHK titratable p.f.u. (less than Iol/ml). However, by direct immunofluorescence assay, these same cells contained virus antigen throughout the entire 35 days observation period (this fluctuated in from 2o to 9o % of the cells): On the other hand p.f.u, synthesis in

4 II6 S. JACOBSON~ F. J. DUTKO AND C. J. PFAU IO?'I:ZfIIII'IIIIIIIIIIIIIIll Illlllllll lo0 109~llr tltrltiti JIM tllirllllrvi;llll~ lo0 E~ Trypsin dispersal (a) I -~ Trypsin t (b)..i[ I dispersal and i \\,:'~ and subculture ~-~/ subculture / \\ g.'~,o6, 8o lo6 /I,t8o= // ~.,, ~.-.~,, ~ _-", :". < F,, "\., // ~ ~- ~t ', :: i] ~ I-, ~," o 1o' 1 v v', V' 't lo' H I /~, :*,., /'I 11 "~ ~ 1- */\ ~,,' "~ ~,~lo l ~'I I I I I t - ~'l'il j I ~- I I I ' I l I -' I I ' I-' I I I I'~'! Li 0 ~<lo]~" ' L' i '~ ' ' I I I :li'l'l I I I i\' I ' I ' i~l ' U'i 0 5~ l t' f f I' I' f t' '1' i' i' '1' I' t' t' f f Time post infection (days) Fig. I. Long-term infection in MDCK cells. One of two confluent MDCK monolayers (5 z@ cells/ 25 cm 2 plastic flask) was treated for [.5 h with I ml of 5o-fold concentrated LCM DI virus stock (a, DS culture). After removal of the inoculum by repeated washing, this monolayer, as well as the other (b, MS culture), was infected with I ml of standard LCM virus at an input m.o.i, of o.[. The initial inoculum for both monolayers, whether DI or standard virus, contained 5 #g DEAEdextran. After repeated washing, growth medium was replaced to the original vol. and incubation was continued at 37 C. ~--~, Standard p.f.u, titre; ~-.-@, SP p.lu. titre; m--i, ~ FAB; ---, ~ cell killing. the MS culture continued until the 23rd day p.i. but, in direct contrast to the DS culture, virus antigen also disappeared shortly thereafter. Susceptibility of persistently infected cultures to lytic infection The DS and MS cultures were infected with lyric virus at various intervals coincident with monitoring their virus antigen content by fluorescence microscopy. The percentage cellkilling in these cultures was then compared to that found in normal cells. In general it appeared that persistently infected cells containing virus antigen were resistant to the lethal effects of standard virus (Fig. 5). For example at I6 days p.i., cells in the MS culture were completely resistant to cell-killing and all contained virus antigen. In the DS culture, cells were completely resistant to ceil-killing at day 7 and from day 32 onward. A minimum of 50 % of the cells contained virus antigen at these times. Detection of occult-self replicating virus in DI virus preparations At 35 days p.i. the MS culture appeared self-cured, i.e. the celts were completely susceptible to standard (lyric) virus infection. Furthermore, at this stage the culture contained no virus antigens (as determined by fluorescence microscopy) and did not require daily media changes. On the other hand, all the cells in the DS culture contained virus antigen and were completely resistant to standard virus. This was thought to be a reflection of either the sequence in which each culture received DI virus (prior to or simultaneously with standard virus infection) or an intrinsic difference in the DI viruses used (one type coming from persistently infected cultures and the other within standard virus stocks). The latter interpretation was favoured

5 Persistence in L C M virus infected cells I17 Fig. 2. Standard and SP plaque morphology in MDCK rnonolayers. Serial tenfold dilutions of various virus stocks were used to infect cells as described in the Methods. Neutral red was added to the agar overlaid monolayers 96 h p.i. At this time standard plaques (b) were easily observed without the counterstain. However, cloned SP virus plaques (a) only became visible between Io and 24 h after staining. Further exposure to neutral red did not increase the clarity of the plaques. Plaques in (d) were seen when virus was harvested from L cells IOl days p.i. with cloned standard virus (b). Plaques in (e) were from the same L cell culture 13 days later. Photographs were taken 24 h after addition of neutral red. since the supernatant fluid from the persistently infected DS culture was f o u n d to contain a self-replicating virus that could be detected only by plaque assay in M D C K cells (Fig. 2). Since the titre of this virus, called SP (stubborn plaquer) was relatively constant t h r o u g h o u t the entire observation period (Fig. I), we assumed that its genesis occurred prior to initiation o f the DS culture, i.e. it was in the infecting DI virus stock (and not the standard virus stock since this virus was c o m m o n to both DS and MS cultures while SP virus was only found in the former). The 5o-fold concentrated D I virus stock used to initiate the DS culture contained Io 4 SP-p.f.u./ml. Culture fluids taken directly from persistently infected L cells (the source o f the D I virus) were frequently found to contain as many as 4 x xo s SP-p.f.u./ml. Cloning of SP virus A n M D C K assay plate was inoculated with diluted tissue culture fluid taken from the DS culture 8 days post-infection. Centres o f six of the resulting I z plaques were excised, expelled into growth medium and briefly exposed to ultrasonic irradiation (Pulkkinen & Pfau, I970). Their contents varied between I.z and 4"5 x to 4 p.f.u. F o u r of these isolates were then used

6 S. IACOBSON, F. J. DUTKO AND C. J. PFAU Table i. Neutralization of standard and SP LCM virus by antibody raised against the primary strain Virus Antibody*,- P.f.u./ml after incubation at 37 C for (time in h): 0 IO 62t I I4~ STD - 8"3 x IO s 7"6 x ~o a "I IO s < lo J < IO < IO SP x io ~ 4"5 I x IO 2 4.8x i0 ~ 3.2 I05 i.ixlo 6 * I : 5 dilution of hamster anti-standard virus sera. t After ro h at 37 C, I ml of the antigen-antibody (Ag-Ab) mixture was used to inoculate MDCK cells in a 25 cm ~ flask. The inoculum was removed at the end of I h and incubation was continued after media replacement. At 62 h the MDCK cells received a complete media change, to infect MDCK monolayers at an m.o.i, of approx, o.oi. Seventy-two h later no cytopathic effects (c.p.e.) were evident, and the tissue culture fluids contained between 1.2 and 2"3 x Io 6 p.f.u./ml. With over Iooo plaques counted from the above virus stocks, as well as others grown since then, no clear or standard plaques have ever been observed. Concurrent plaque purification of standard virus taken from the MS culture was also carried out. All plaque isolates contained about the same number of p.f.u, as noted above but with no evidence of any SP contamination. These were able to kill MDCK cells in the expected manner (Dutko & Pfau, I978). Genesis of SP virus The above studies indicated that SP virus was present in long-term persistently-infected L cells. To rule out its adventitious nature a new L cell culture was inoculated with cloned standard virus that appeared by a number of criteria to be devoid of SP virus. These criteria were that (0 no visible SP plaque types were ever observed in standard virus plaque assay plates; (2) no SP plaque types were apparent after selective neutralization of standard virus by homologous antibody (Table I); (3) no replication of either virus type was apparent after inoculating MDCK ceils with neutralized standard virus; and (4) no SP virus arose in over a year of continuous cultivation of MDCK ceils inoculated with standard virus alone (the MS culture, Fig. 0. L cell tissue culture fluid taken prior to infection with standard virus was also devoid of SP virus as shown by inoculation on to MDCK cells. As expected (Staneck et al. 1972; Pfau, i977) cycling of standard p.f.u, titres within these cell cultures began within a few days after infection and continued until I I4 days p.i. (Table 2). A mixture of two plaque types clearly detectable by MDCK assay was first seen IOI days p.i. (Fig. 2, Table 2). One was a clear standard type, except that the plaques were smaller. The other was a turbid SP type, except that the plaques were larger. When several of these plaque types were excised with Pasteur pipettes, sonically disrupted and reassayed their morphologies were identical to the SP and standard plaque prototypes. The newly isolated turbid plaque type was considered to be an SP virus not only by its plaque morphology but also by its lack of cytolytic ability in MDCK cells and its one-way antigenic cross reactivity with standard virus (see below). At 114 days and thereafter no plaque forming virus was detectable by BHK assay but indeed SP virus was maintained at a rather constant titre as shown by MDCK assay (Table 2). Properties of SP virus SP virus possessed the general physical-chemical properties of the arenavirus family (Pfau et al. I974). The virus was ether sensitive, contained the expected two virus specific classes of RNA (E. Vander Kruik & C. J. Pfau, unpublished observations) and had a u.v.

7 - - 20"8 Persistence in LCM virus infected cells II9 Table 2. Transition in plating efficiency and plaque morphology of LCM virus h7 persistently infected L cells P.f.u./ml* Time p.i. BHK assay MDCK assay 85 1"8 x lo 6 STD 1.2 x io 5 STD IOI I" 3 IO 6 STD 8'5 I05 STD+SP 114 < Io STD 3'0 x IO s SP 199 < I0 S'ID 1.2 x io 6 SP 215 < Io STD I.ox~o 6SP * STD = standard virus plaque morphology, SP = SP virus plaque morphology. Table 3- Separation of immune complexes from mixtures of 14C-labelled LCM virus and antiserum by use of Staphylococcal protein A * Antigen Antibody ~ Ag % Ag-Ab against? bound bound~ Std-LCM Std-LCM 7o"I 5o"I SP Tissue culture fluid 63"0 27"3 43 '8 7'3 -- SP virus SP 46"3 I9"4 Std-LCM ' 1 Tissue culture fluid 3['4 4' '9 -- * 0.2 ml of x4c-labelled virus (see Methods) was mixed with an equal volume of antiserum and incubated at 37 C with intermittent shaking for I h; 0"2 ml of Staph. aureus protein A was then added and shaken intermittently and vigorously for 3o rain at room temperature. The mixture was centrifuged for I5 rain at 3ooog. Radioactivity measurements were made on both the supernatants and resuspended pellets. Std = standard.? I "50 dilution. ~ immune complex bound was the ~ Ag bound in the pellet minus non-specific binding in the absence of Ab addition. o'o3~ triton-x ioo added after incubation of antigen and antibody to prevent high (64~) non-specific binding of Ag to protein A. target size identical to standard virus (straight line, semilogarithmic, survival curves were obtained over a 2 log~0 unit range with a D37 value for both viruses of I57 ergs/mm~). Serum from hamsters immunized with SP virus reacted by indirect fluorescence microscopy with both SP and standard virus infected cells. SP virus-infected acetone fixed cells showed the usual cytoplasmic punctate fluorescence while this and membrane fluorescence was seen in the standard virus infected cells. Viable SP virus-infected cells exposed to homologous antiserum showed either diffuse, speckled, or cap-like membrane fluorescence. Standard virus infected cells treated in this way exhibited diffuse and intense membrane staining. Using hamster antiserum against gtandard virus, positive reactions with both types of prepared cells were observed with homologous virus only. This one way cross reactivity was confirmed by radioimmunoassay (Table 3)- The trivial explanation for this one way cross reactivity, i.e. SP virus was contaminated with standard virus and thus antibodies would be made against both, was ruled out on the following basis: the growth of standard virus and expression of its cytolytic activity in MDCK cells pre-infected with SP virus was found to proceed normally. MDCK cells infected with cloned SP virus and followed for ~oo days (see below) never showed any c.p.e, or expression of standard antigen as determined by fluorescence microscopy and plaque assay. Since temperature sensitive (ts) mutants, as well as DI virus, appear to play a critical role in certain persistent infections (Youngner, 1977) SP virus was examined for this characteristic. Using the MDCK plaque assay the efficiency of plating (e.o.p.) of standard virus at 39"5 C was I.o (when compared to counts at 37 C) while the e.o.p, of SP virus was o'33. The e.o.p. a~

8 I20 S. JACOBSON~ F. J. DUTKO AND C. J. PFAU Table 4. Long-term effect of SP virus on resistance of MDCK cells to killing by standard virus Day of SP virus addition after initial infection No virus added 8 1o Days after, ~ " " ~- " ' '- initial % ~ cells Std % ~ cells SP ~ ~ cells SP std virus fluorescing resistant p.f.u./ml fluorescing resistant p.f.u./ml fluorescing resistant p.f.u./ml infection cellst to killing cellst to killing IO -6 cellst to killing Io -6 I6 o -- o o o -- 1"o o ' " O "5 73 o o o I ' "7 87 o I4 o 95 1oo o'4 9o 85 o'3 * An MS culture was initiated as described in the legend of Fig. i and at the time of the first trypsin dispersal of the descendants of the surviving cells (8 days post-infection), or z days later, cloned SP virus was added to separate subcultures at m.o.i, of o.ooi. t FITC-labelled antiserum to standard virus. of standard virus at 32 C was 0.7' 5. While SP virus formed no visible plaques at 32 C it grew at this temperature as shown by plaque assays carried out at 37 C. Its growth curve at 32 bc paralleled about one log10 unit below that of standard virus, the same differential as observed at 37 C. Requirement of SP virus for long-term persistence of standard virus infection The MDCK culture (DS) pie-treated with DI virus stocks containing SP virus became persistently infected with the standard virus that had been used in the secondary infection. Since the MDCK culture (MS) infected with standard virus alone cured itself, experiments were undertaken to rescue this infection by deliberate addition of SP virus to cells surviving the initial standard virus inoculum. These cultures, along with control standard virus infected cells and cells infected with SP virus alone were followed for IOO days with regard to expression of standard virus antigens and resistance to cell-killing. The addition of SP virus to standard virus-infected cells produced an easily recognizable effect in that within z days the cells became t.5 to 2 times their normal size. They began to grow again after several more days, but throughout the entire observation period they continued to appear abnormally large for the first two days after each trypsin dispersal. Furthermore, daily media changes became even more crucial for ceil survival. As seen in Table 4, the control culture infected with standard virus alone had cured itself by the i6th day p.i. At no time during the observation period was standard virus or antigen detected within these cells or in the cells infected with only SP virus. However, the genetic information for standard virus antigen expression was preserved with late addition of SP virus and eventually antigen increased within the cells in concert with their ability to resist killing by standard virus challenge. This was shown by the fact that 77 to 79 days after infection with SP virus most, if not all, cells contained standard antigens as determined by fluorescence microscopy (Table 4). These standard virus antigens were assumed to be due to the synthesis of DI virus since the surrounding tissue culture fluids now yielded particles with their sedimentation and interfering (with standard virus synthesis) properties. DISCUSSION The above studies have led to the isolation of a non-cytopathic, self-replicating, LCM virus that appears critical in changing standard (lytic) virus infection of MDCK cells from a transient to a persistent state. This type of virus which, in our hands, could only be plaque

9 Persistence in LCM virus infected cells I2I assayed in MDCK cells may have been found by others, but its salient propelties as described here have not been reported. A most likely candidate is an LPV (low pathogenic variant) LCM described by Hotchin & Sikora (I973). Unfortunately, all stocks have been lost through deep-freeze failures (J. Hotchin, I977, personal communication). Other candidate SP-like viruses may be among those recently described by Popescu & Lehmann-Grube (1976). The presence of SP virus in our LCM-UBC DI virus stocks raises the possibility of re-evaluating previous reports on the properties of DI virus. Although none of the DI virus preparations we have are free of SP virus, the properties of DI virus can be judged by comparing the biological activity of these preparations with cloned SP virus (which contains little, if any, DI virus, see below). From experiments of this type, not presented here, it is clear that SP virus neither has the ability to protect cells against standard (lytic) virus infection (Dutko & Pfau, 1978) nor interfere with the latter's synthesis (Welsh et al ). The presence and possible role of an SP-like virus in other LCM-DI virus systems (Welsh et al I977; Welsh & Oldstone, 1977) warrants attention. Until the discovery of the MDCK cell-killing assay (Dutko & Pfau, I978) it was not possible to quantify the cytolytic potential of LCM virus nor demonstrate the ability of DI virus to abrogate this phenomenon. One of the initially perplexing observations from this study was that a maximum of 8o to 85 % of the cells could be killed no matter what concentration of virus was used during the infection process. This was attributed to the background DI virus content of our standard virus stocks and not to genetic resistance of some cells since 20 randomly picked clones from normal MDCK cells failed to yield any with altered resistance. This latter interpretation seems even more remote now since our studies show that in the MS culture the few cells which survive standard virus infection quickly become completely susceptible again (Fig. I). Between initial infection and complete self curing in the MS infection there was a single resurgence in standard virus synthesis. This new production of standard virus closely coincided with maximum antigen expression in the cells. Paradoxically it was also the time at which the cells were totally resistant to being killed by the standard virus they were producing (virus isolated from cells at this time had full cytolytic ability in normal MDCK cells). Since antibodies made against standard virus did not cross-react with the cytoplasmic antigens in SP virus-infected cells we assume that maximum standard virus antigen expression (in the absence of detectable standard virus) in the DS culture, which coincided with maximum resistance to cell-killing, was due to DI virus synthesis. Thus, we suggest that in the MS culture there was concordant synthesis of standard and DI virus, i.e. the lytic potential of standard virus was blocked by ongoing replication of DI particles. If we assume that the resistance to cell-killing in both the MS and DS cultures is due to the synthesis of DI virus and that this is easily enriched at the expense of standard virus while at the same time needing the presence of the complete genome for its own replication, a cycling in the synthesis of both types of virus might be expected (Huang & Baltimore, 1977). It is also conceivable that without the genesis or presence of a self-replicating virus relatively resistant to the DI enrichment advantage an infection could eventually burn itself out. Thus, we believe that the cause of the marked difference in the outcome of the DS and MS infection may be due to SP virus supplying the complete genome necessary for continued DI virus replication. Data so far gathered are consistent with this hypothesis in that: (I) An MS culture can be changed to a DS-type culture by addition of cloned SP virus (Table I); (2) SP virus synthesis must be relatively resistant to inhibition by DI virus because it was originally found in cells synthesizing DI virus. Titres of SP virus were also relatively constant in the DS culture although there was a slight but noticeable cycling which was inversely coincident with cycling of standard antigen expression.

10 I22 S. J'ACOBSON, F. J. DUTKO AND C. J. PFAU This hypothesis may also help to explain results obtained from the LCM carrier mouse. It has recently been found that DI particles and standard (clear plaque type) virus multiply rapidly and in parallel when injected into neonatal mice (Popescu & Lehmann-Grube, 1977). Within I month the concentrations of both types of virus decline to a maximum of 1% of their former levels. During this time the input standard virus is largely replaced by turbid plaque types. We feel that DI virus may be vital in the initial stages of establishing the carrier state in mice by blunting standard virus cell-killing ability in all but specific subpopulations of T lymphocytes (Pfau, I978). Once this turning point has passed there may be a rapid selection of DI virus-resistant mutants by mechanisms that are not yet clear. In many ways this hypothesis is similar to those expressed by others (e.g. Younger, 1977). We thank Dr Peter B. Jahrling for a generous gift of Staphylococcal protein A and Drs D. E. Wilson and J. D. Gangemi for their interest and many helpful suggestions. REFERENCES BtJCHM~IER, U.J. & OLDSTONE, M. B. A. 0978)- Virus-induced immune complex disease: identification of specific viral antigens and antibodies deposited in complexes during chronic lymphocytic choriomeningitis virus infection. Journal of lmmunology xzo, I297-13o4. OUTKO, F. J. & VrA~, C. J. 0978). Arenavirus defective interfering particles mask the cell-killing potential of standard virus. Journal of General Virology 38, OUTKO, r. J., WRI6HT, E. A. & PFAU, C. J. 0976). The RNAs of defective interfering Pichinde virus. Journal of General Virology 3 I, OSCHWENDER, U. H., RUTTER, G. 8, POPESCU, M. (I975). Use of iodinated organic compounds for the density gradient centrifugation of viruses. Archives of Virology 49, HOTCHIN, J. & SIKORA, E. (I973). Low-pathogenicity variant of lymphocytic choriomeningitis virus. Infection and Immunity 7, I~UANG, A.S. & BALTIMORE, O. 0977). Defective interfering animal viruses. In Comprehensive Virology, vol. lo, pp I6. Edited by H. Fraenkel-Conrat and R. R. Wagner, New York: Plenum Press. JAHRLING, P. B., HESSE, R. A. & METZGER, J. F. 0978). Radioimmunoassay for quantitation of antibodies to alphaviruses with Staphylococcal protein A. Journal of Clinical Microbiology 8, PFAU, C. J., BERGOLD, G. H., CASALS, J., JOHNSON, K. M., MURPHY, F. A., PEDERSEN, I. R., RAWLS~ W. E., ROWE, W. P., WEBB, P. A. & WEISSENBACHER, M. C. 0974). Arenaviruses. Intervirology 4, PrAU, C. J. 0977). The role of defective interfering (DI) virus in arenavirus infections. Medicina, Buenos Aires 37, PrAU, C.J. 0978)- The immunological basis of persistent infection and disease in lymphocytic choriomeningitis virus-infected mice. In Comparative and Developmental Aspects of Immunity and Disease, pp o 7. Edited by M. E. Gershwin and E. L. Cooper. Oxford: Pergamon Press. I'OPESCU, M. & LEHMANN-ORUBE, r. 0976). Diversity of lymphocytic choriomeningitis virus: variation due to replication of the virus in the mouse. Journal of General Virology 3o, I t 3-I zz. POPESCU, M. ~, LEHMANN-GRUBE, V. (I977)- Defective interfering particles in mice infected with lymphocytic choriomeningitis virus. Virology 77, VULKKINEN, A. J. ~ PFAU, C. J. (197O). Plaque size heterogeneity: a genetic trait of lymphocytic choriomeningitis virus. Applied Microbiology zo, 123-I 28. RICE, R. ft. & MEANS, G. E. (197 I). Radioactive labeling of proteins in vitro. Journal of Biological Chemistry 246, STANECK, L. D., WELSH, R. M., TROWBRIDGE, R. S., WRIGHT, E. A. & PFAU, C. J. 0972). Arenaviruses: cellular response to long term in vitro infection with Parana and LCM viruses. Infection and Immunity 6, WELSH, R. M. & PFAU, C.J. 0972). Determinants of lymphocytic choriomeningitis interference. Journal of General Virology x4, I WELSH, R. M., O'CONNELL, C. M. & PFAU, C. J. (1972). Properties of defective lymphocytic choriomeningitis virus Journal of General Virology x 7, WELSH, R. M. & OLDSTONE, M. a. A. 0977). Inhibition of immunologic injury of cultured cells infected with lymphocytic choriomeningitis virus: role of defective interfering virus in regulating viral antigenic expression. Journal of Experimental Medicine x45, 1449~ WELSH, R, M., LaMPERT, V. W. & OLOSTONE, M. B. A. 0977). Prevention of virus-induced cerebellar disease by defective-interfering lymphocytic choriomeningitis virus. Journal of Infectious Diseases x36, YOUNGNER, J. S. (~ 977). Role of temperature-sensitive mutants in persistent infection. In Microbiology , PP Edited by D. Schessinger. Washington D.C. : American Society for Microbiology. (Received I8 August I978)