VIRAL INTERFERENCE

SOME CONSIDERATIONS OF BASIC MECHANISMS AND THEIR POTENTIALRELATIONSHIP TO HOST RESISTANCE'

ROBERT R. WAGNERDepartment of Microbiology, The Johns Hopkins University School of Medicine, Baltimore, Maryland

CONTENTS

I. Introduction ........................................... 151II. Bacteriophage "Interference"............................................................. 152

A. Resistance to Lysis from Without ...................................................... 152B. Immunity of Lysogenic Bacteria........................................................ 152C. Mutual Exclusion and Depressor Effect.................................................. 153

III. Interference between Animal Viruses ............................... 155A. Site of Viral Interference............................................................. 155B. Role of Virus Nucleic Acids in Interference .............................................. 156C. Interferons ............................................................. 157D. Resistance to Superinfection of Persistently Infected Cell Cultures ..................... 162

IV. Summary and Theory ....................................... 162V. References ............................................ 163

I. INTRODUCTION

The profusion and diversity of the experimentaldata have led many students of the subject toconsider that viral interference may be not onephenomenon, but many. This view stems in largemeasure from the inability to predict with con-fidence which virus will interfere with another inany particular type of cell. The result has beenan imposing list of empirically determined pairsof interfering viruses which stands as eloquenttestimonial to our ignorance of the mechanismsinvolved. The literature on viral interference hasbeen collated and annotated so comprehensivelyby Schlesinger (1959) that it would serve no use-ful purpose to review his review. Inevitably, how-ever, the present discussion will cover some ofthe same well-spaded ground.

This analysis of viral interference was ap-proached with two specific objectives in mind:to examine the evidence for a unifying principlethat may underlie the seemingly diverse reactionsbetween viruses and cells that lead to interfer-ence; and to explore the possibility that thesereactions may be implicated in host defenses dis-tinct from specific immunity. These may be pre-mature and quixotic hopes. A theory of themechanism of viral interference must eventually

1 The experimental studies reported herein weresupported by grants from the U. S. Public HealthService and the National Science Foundation.

be based on chemical evidence, and needless tosay, such evidence is scant. Nevertheless, thediscovery of interferon (Isaacs and Lindenmann,1957), a cell product of determinable chemicalnature, holds out promise for future understand-ing of at least some of the cellular reactions in-volved in viral interference. The interpretationspresented herein have been greatly influenced bythis finding and are put forth with full awarenessthat they may be controversial.An exact definition of viral interference is not

possible at the present time. It has generally beenassumed that the phenomenon represents com-petition between two viruses for the same hostcell, but, as will be indicated later, the validityof this concept may be open to question. Never-theless, meaningful interpretation of data stillrequires that the use of the term be restricted toevents that take place at a cellular level. In thiscontext, therefore, interference signifies acquiredcellular resistance to viral infection. It wouldprobably be wise to impose the additional quali-fication that inhibition of virus multiplication bethe essential criterion of interference. Althoughinterference may result in enhanced capacity ofa host to survive infection or in decreased im-munologic responsiveness, these are secondarymanifestations of a cellular environment inimicalto the infecting virus. By convention, the agentthat induces the state of cellular resistance to in-fection is referred to as the interfering virus and

151

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

ROBERT R. WAGNER

the one that is suppressed as the superinfectingvirus. Under certain conditions, multiplication ofboth viruses may be inhibited. Although thesedesignations imply dissimilarity between the viruspairs, it has long been recognized that related oreven biologically indistinguishable viruses caninterfere with each other. Contrary to custom, noattempt will be made to differentiate autointer-ference from heterologous interference betweenunrelated viruses. In the writer's opinion, clas-sifications of interfering systems based on anti-genic or other biological relationships amongviruses may be unduly restrictive and may im-pose unwarranted complications in an analyticalapproach to the subject.

II. BACTERIOPHAGE "INTERFERENCE"

The term, viral interference, has not been pop-ular among bacterial virologists (Adams, 1959).To these quantitative biologists, interference be-tween animal viruses has borne the unhappyconnotation of disconnected, and perhaps unre-lated, events that transpire in dissimilar cells ofa multicellular host. This view, of course, is trueof almost all studies of mixed viral infections ofintact animals. However, the advent of improvedcell culture techniques has provided a means bywhich interference between two viruses that in-fect a single animal cell can be studied and com-pared with mixed infections of a bacterial cell.There are indications that the comparison willreveal more than a superficial resemblance. Itseemed appropriate, therefore, to attempt a briefanalysis of resistance of bacteria to superinfectionwith bacterial viruses in the hope of gaining someinsight into possible mechanisms of interferencebetween animal viruses.

There are probably three ways by which a bac-terial virus can induce a state of resistance tosuperinfection with the same or different viruses.These three phenomena have been saddled withfour rather unfortunate terms: resistance to lysisfrom without (Visconti, 1953); immunity of lyso-genic bacteria (Lwoff, 1953); mutual exclusion,and the depressor effect (Delbrilck, 1945). Thelatter two appear to be closely related and willbe discussed together. Their similarity to inter-ference between animal viruses has prompted theuse of the convenient term interference in thediscussion of mutual exclusion.

A. Resistance to Lysis from Without

Exposure of a bacterial cell to a high multiplic-ity of phage results in premature lysis of the cellwithout production of new phage progeny. Thishas been called lysis from without to distinguishthe reaction from active viral infection (Delbriick,1940). It resembles a toxic response in that virusinactivated by ultraviolet irradiation can producethe same lytic effect (Watson, 1950). Protectionagainst lysis from without does not occur in bac-teria previously infected with unrelated strains ofbacteriophage (Doermann, 1948). However,Visconti (1953) found that bacteria infected witha low multiplicity of T2 phage rapidly becameresistant to lysis from without caused by therelated T2r strain of phage. The mechanism ofthis protective effect is not clear. It may possiblybe a function of altered permeability of the hostcell wall, which has been noted to occur afterpenetration by one or a few phage particles (Puckand Lee, 1955). If so, the challenge virus, orrather its deoxyribonucleic acid (DNA), may beprevented from entering the resistant cell in quan-tities sufficient to induce rapid lysis, or the lyticenzyme of the superinfecting phage may beblocked. This protective effect is reminiscent ofacquired resistance to the toxic action of massivedoses of influenza virus in animals pretreated withsmall inocula of biologically related viruses (Wag-ner, 1952). Neither of these protective effectssheds much light on the mechanism of viral in-terference and, in fact, both are probably unre-lated to it.

B. Immunity of Lysogenic Bacteria

This phenomenon has been exhaustively stud-ied by precise quantitative methods. Despite thepresence of prophage in the lysogenic bacterium,certain unrelated virulent phages can adsorb on,penetrate, and multiply within the cell. The off-spring of the superinfecting virus released fromthe lysogenic cell are indistinguishable from virusformed in nonlysogenic bacteria. If, however, thesuperinfecting phage is closely related geneticallyto the prophage carried by the lysogenic bac-terium, it will adsorb and penetrate, but nomultiplication will occur (Lwoff and Gutmann,1950; Lwoff, 1953). Nor can a related temperatephage doubly lysogenize an "immune" bacter-ium, except under unusual circumstances (Ber-tani, 1956). One other exception to this gen-eralization is a class of virulent phage mutants

152 [VOL. 24

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

VIRAL INTERFERENCE

capable of infecting bacteria lysogenized with re-lated prophage (Bertani, 1953). A genetic basisfor immunity to superinfection of certain lyso-genic bacteria has been demonstrated by the ele-gant studies of Jacob and Wollman (1957). Theyshowed that no cross immunity existed between14 different temperate phages, each capable oflysogenizing Escherichia coli K12. Studies of bac-terial recombinants by crossing experiments re-vealed that each prophage occupied a differentgenetic locus. It appears, therefore, that the DNAof a superinfecting temperate phage cannot beincorporated into the same site on the bacterialchromosome preempted by a related temperateprophage.However, genetic competition may not be the

only factor operative in immunity of lysogenicbacteria. Bertani (1956) found that a lysogenicbacterium resists superinfection with phage P2 ifit carries prophage P2 at a secondary nonpre-ferred site on its chromosome. This led to thesuggestion that "immunity is physiologically con-trolled (by some product of the prophage?) andnot the result of competition for the standard siteof prophage attachment." In addition to "homol-ogous immunity," Lederberg (1957) has shownconclusively that E. coli or Shigella dysenteriaelysogenic for phages P2 or P1 also fail to supportmultiplication of certain heterologous bacterio-phages. Growth of superinfecting phage is sup-pressed at some stage following attachment. It isdifficult to invoke a genetic basis for this type ofresistance to infection which, in many respects,resembles interference (mutual exclusion) be-tween unrelated viruses. As one tentative hy-pothesis, Lederberg suggests that a specific in-hibitor, possibly deoxyribonuclease (DNase), maybe formed by the host cell in response to thesuperinfecting phage.

C. Mutual Exclusimon and Depressor EffectApproximately 8 to 10 T2 phage particles can

infect a single bacterium and participate in intra-cellular growth (Dulbecco, 1949a). The bacterialcell will produce the same number and type ofoffspring regardless of whether they are derivedfrom one or several parent phage particles. How-ever, the yield of phage may be inhibited if morethan 10 infectious units are introduced into abacterial cell. When two different, but related,phages infect the same cell, offspring of bothtypes and genetic recombinants resembling

neither parent may be produced. In mixed infec-tions with completely unrelated pairs, one phagemay be dominant and completely inhibit theother (Delbriick and Luria, 1942). The numberof phage particles of each type produced in themixedly infected cell depends on their geneticrelationship, which of them is dominant, the mul-tiplicity of each type of infectious particle perbacterial cell, and the times at which infection isinitiated with each phage. This phenomenon wasoriginally conceived as inhibition of adsorptionand penetration of the superinfecting phage,hence the term mutual exclusion. It is now clearthat "exclusion" does not take place at the sur-face of the host cell but does so intracellularly,i.e., the suppressed phage is "excluded" frommultiplying rather than penetrating. The "de-pressor effect" noted during mutual exclusionsimply refers to the fact that the yield of domi-nant phage is also reduced by the action of theexcluded superinfecting phage (Delbriuck, 1945).Convincing proof that mutual exclusion and thedepressor effect are often, if not always, intra-cellular phenomena is furnished by studies of in-duced lysogenic bacteria (Weigle and Delbriuck,1951). If E. coli K12, lysogenic for phage lambda,is irradiated and then superinfected with T5phage, the yield of each phage will be reducedappreciably. Obviously, suppression of inducedlambda, which was inside the cell at the start ofthe experiment, could not have been caused byits failure to penetrate.The conditions for demonstrating mutual ex-

clusion (including the depressor effect) of bac-terial viruses have been succinctly summarizedby Adams (1959): "If a bacterial strain is sus-ceptible to two distinguishable phages, it is pos-sible to study the results of mixed infection ofsingle cells with the pair. If the two infectingphages are not related, the usual result is mutualexclusion; one phage or the other multiplies butnot both. If one phage is clearly dominant underconditions of simultaneous mixed infection, it ispossible to transfer the advantage to the secondstrain by giving it a few minutes head start. Themechanism of mutual exclusion is not known butit clearly does not involve interference with ad-sorption, interference with penetration or com-petition for a unique key enzyme." Except forthe time relationships, none of these conditionsdifferentiates mutual exclusion of bacteriophagesfrom interference between animal viruses.

1960] 153

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

ROBERT R. WAGNER

Having cited some of the negative evidence forthe mechanism of mutual exclusion, it seemsworth while to consider whether the interferingagent can be identified as a constituent of theparental phage particle. It is safe to dismiss theexternal protein coat of the interfering phagebecause it does not penetrate the bacterial cellwall (Hershey and Chase, 1952). Ostensibly, weare left with intact phage DNA as the interferingprinciple, or more accurately, the factor thatinitiates the process of mutual exclusion. Thismay well be the case, but there are several cogentreasons for examining this hypothesis moreclosely. The first of these, as previously men-tioned, is that prophage DNA, which contains allthe genetic information for phage production,may not inhibit multiplication of an unrelatedsuperinfecting phage (Lwoff, 1953; Jacob andWollman, 1957). This, of course, may be a func-tion of inaccessibility of prophage DNA on thebacterial chromosome. If the lysogenic bacteriumis induced, the temperate phage will excludesuperinfecting unrelated phage (Weigle and Del-briuck, 1951). Of greater significance, perhaps, isthe fact that "T2r+ phage heavily irradiated withultraviolet light is nearly as potent as activephage in stimulating the exclusion of superinfect-ing phage T2r" (Dulbecco, 1949b). Ultravioletirradiation damages almost exclusively the geneticmaterial of phage and inhibits its capacity, atleast temporarily, to participate in nucleic acidmetabolism (Cohen and Arbogast, 1950; Her-shey et al., 1954). Therefore, if the DNA mole-cule of the interfering phage is in fact responsiblefor inducing mutual exclusion, its inhibitory ac-tion cannot be ascribed to production of compet-ing virus DNA by the resistant bacterial cell.

Further presumptive evidence that the infec-tious component of phage is not identical with itsinterfering activity is provided by studies withparental phage DNA labeled with p32. Lesley etal. (1951) have demonstrated that DNA of super-infecting phage is rapidly degraded and expelledfrom the cell. The degradation occurs within afew minutes and is apparently related to en-hanced DNase production by the bacterial cell.However, mutual exclusion takes place even ifbacterial DNase is inhibited by streptomycin(French et al., 1952) or reduced concentrations ofmagnesium (Hershey et al., 1954). Under condi-tions of enzyme inhibition the superinfectingphage DNA is retained within the cell. In addi-

tion, DNA of excluded T2 phage is not incorpo-rated into the progeny of unrelated T1 and T7phages (French et al., 1952). The studies of Her-shey et al. (1954) are of interest in another respect.Their experiments suggest that phage DNA istransferred from parent to offspring in largepieces rather than as constituent nucleotides,whereas the injected DNA of superinfectingphage is degraded in toto.These studies do not rule out the possibility

that phage DNA is the agent responsible forinitiating the process that leads to mutual exclu-sion. They merely indicate that the interferingagent within the resistant bacterium cannot beequated with prophage DNA of the bacterialchromosome, degraded DNA, intact DNA com-ponents that bear genetic information, or bacte-rial DNase. Thus, some doubt remains of thevalidity of the thesis that mutual exclusion de-pends entirely on genetic or metabolic competi-tion between two incompatible molecules ofphage DNA.

Recently published reports suggest a possiblealternative approach to a chemical analysis ofbacteriophage interference. Hershey (1955) hasidentified internal protein-like constituents ofT2 phage that comprise about 3 per cent of totalphage protein. Unlike the external proteins ofthe head and tail, the internal components pene-trate susceptible cells along with phage DNA.Further analysis (Hershey, 1957) reveals at leasttwo internal protein-like substances, one a poly-peptide formed by incorporation of host celllysine and the other, called substance A, which isderived from precursor arginine. Substance A isincorporated into phage progeny, is dialyzableand can enter bacterial cells in the absence ofphage DNA. Levine et al. (1958) have shownthat an internal protein of disrupted phage isheat stable and immunologically distinct fromproteins which comprise the external coat. Of thegreatest interest is the finding of Ames et al.(1958) that T4 phage contains substantial quan-tities of the polyamines putrescine and spermi-dine derived from host putrescine. It appearsthat these polyamines may be identical to Her-shey's substance A and can combine with phageDNA, presumably by reacting with negativelycharged phosphate groups of the DNA molecule.In addition, a different polyamine, spermine, ispresent in salmonella phages PLT-22 and 98.With these studies in mind, Levy and Wagner

154 [VOL. 24

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

VIRAL INTERFERENCE

(1959, unpublished data) have attempted to inter-fere with multiplication of phage T7 by pretreat-ing E. coli B with "shockates" of purified T2phage. Thus far, these efforts have been unsuc-

cessful, possibly because the interfering agentmay be the lysine-containing basic polypeptidecomponent of the internal phage protein ratherthan the polyamines. The fact that the free poly-peptide does not adsorb on bacterial cells (Her-shey, 1957) poses the difficult problem of design-ing an experimental model to study its potentialinterfering activity. Nevertheless, this very ten-tative postulate, that internal phage proteinsmay participate in the process of bacteriophageinterference, warrants future consideration. Pos-sibly DNA or ultraviolet-irradiated DNA of theinterfering phage stimulates the bacterial cell toproduce polyamines or basic polypeptides. Thesecompounds within the resistant cell may thencombine with and inactivate DNA of the super-

infecting phage.

III. INTERFERENCE BETWEEN ANIMAL VIRUSES

It is a rather remarkable fact that so much in-formation about the kinetics of viral interferencein animal cells was obtained prior to availabilityof an adequate experimental model. Until therecent advent of refined cell-culture methods, theonly host systems available for semiquantitativestudies were the intact chick embryo and Mait-land-type cultures of embryonic tissues. Thechief drawback of these experimental models isthe inability to determine the exact number ofcells available for infection and the exact numberof virus particles that participate in productionof progeny. Nevertheless, several ingenious esti-mates have been made by direct counting of al-lantoic cells (Fazekas de St. Groth and Cairns,1952) and of influenza virus particles (Isaacs,1957). It is a tribute to the pioneers in this fieldthat much of their data on interference betweenanimals viruses is being substantiated by cell-culture methods.

A. Site of Viral InterferenceThe first problem confronted by virologists

interested in the mechanisms of the interferencephenomenon was whether the primary reactiontook place at the cell surface or intracellularly.The solution may seem obvious in retrospect butfor some time the issue was clouded by the knownaction of influenza virus on erythrocyte receptors.

It seemed plausible to consider that receptor-destroying enzyme (RDE) might by the responsi-ble factor, despite evidence that only slight andtransitory resistance to infection can be producedby excessive quantities of RDE obtained fromcholera filtrate (Stone, 1948). Isaacs and Edney(1950) demonstrated that heat-inactivated virusdevoid of RDE is an effective interfering agent,whereas formalinized virus with enzymatic ac-tivity is not. Similarly, incomplete influenza virus(noninfectious hemagglutinin derived from HeLacells) may retain receptor-destroying activitydespite a markedly reduced capacity to initiateinterference (Paucker and Henle, 1958). Schle-singer (1951) has also shown conclusively thatdestruction of surface receptors by influenza virusdoes not account for its capacity to interfere witheastern equine encephalomyelitis (EEE) virus.Resistance to the toxic action of influenza virusesinduced by cholera filtrate (Wagner, 1952) provedto be completely unrelated to its enzymatic ac-tivity (Group6 et al., 1954) and is almost un-doubtedly attributable to its content of bacterialendotoxin. It seems clear, parenthetically, thatendotoxin-induced resistance to viral infectionbears only superficial resemblance to the inter-ference phenomenon (Wagner et al., 1959).

It is, of course, conceivable that interferencebetween myxoviruses may represent alteration ofcellular receptor sites by a mechanism other thanenzymatic destruction. Baluda (1959) is the chiefproponent of the thesis that interfering virus pre-vents adsorption of superinfecting homologousvirus on the surface of host cells. Almost all otherinvestigators have presented evidence to the con-trary. To cite but a few examples, Henle et al.(1947) and Isaacs and Edney (1950) have shownthat superinfecting virus adsorbs on "interfered"cells and disappears. By implication, it can beassumed that both interfering and superinfectingvirus penetrate the resistant cell. The effective-ness of influenza virus as a homologous interfer-ing agent is readily demonstrable even when it isadministered after infection is established (Henleand Rosenberg, 1949). Moreover, Levine (1958)could detect no difference in the degree to whichwestern equine encephalomyelitis (WEE) viruswas adsorbed on susceptible chick embryo cellsor on those rendered resistant to superinfectionby prior exposure to Newcastle disease virus(NDV). Therefore, unless the interfering agentprevents virus release from infected cells, for

15519601

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

ROBERT R. WAGNER

which result there is no evidence, its inhibitoryaction must be on the intracellular phase of virusmultiplication (Henle, 1950).

It is appropriate at this point to examine someof the evidence for interaction between two vi-ruses within a single cell. Baluda (1957), amongothers, has shown that at least one virus particleper cell is required to induce interference. How-ever, more than one infectious unit can enter acell. If the infecting dose is excessive, the yieldof infectious virus will be diminished. The prog-eny resulting from large inocula of infectious in-fluenza virus is often composed of a preponder-ance of noninfectious incomplete virus (vonMagnus, 1951). Although there is no assurancethat only multiply infected chick allatoic cellscan produce incomplete virus (Fazekas de St.Groth and Graham, 1954), it is almost certainlytrue that the yield of noninfectious hemagglutininfrom HeLa cell cultures depends on the number ofvirus particles that infect each cell (Henle et al.,1955).

Genetic studies furnish more conclusive evi-dence for interaction of two animal viruses withina single cell. In mixed infections of chick embryocells with two distinguishable myxoviruses, someof the resultant progeny may be genetic recom-binants (Burnet and Lind, 1951) or phenotypi-cally mixed heterozygous variants (Granoff, 1959).Cross reactivation of two viruses rendered non-infectious by ultraviolet irradiation has also beendemonstrated (Gotlieb and Hirst, 1956). It seemsunlikely that these phenomena can be attributedto anything but double infection of a single cell.Except for difficulty in demonstrating reciprocalrecombination of influenza viruses (Hirst andGotlieb, 1955), the analogy to genetic interactionand phenotypic mixing of bacteriophages isclearly apparent.The often demonstrated fact that a doubly

infected cell can produce progeny resemblingneither parent has raised the interesting possi-bility that interference might at times be illusory.Presumably, genetically recombined or pheno-typically mixed viruses could be formed whichare incapable of infecting test hosts or cell culturesused to detect their presence. However, this seemsextremely unlikely as a general occurrence in in-terference systems. Rates of recombination of re-lated animal viruses are probably very low. Fur-thermore, the most marked degrees of interferenceoften occur in mixed infections with completely

dissimilar viruses which have not been found toundergo genetic interaction.

B. Role of Virus Nucleic Acids in InterferenceLet us next consider the possibility that nu-

cleic acids of two interfering viruses might beantagonistic even if they are incapable of geneticinteraction. This theory of competitive inhibi-tion of incompatible virus nucleic acids has prob-ably had the greatest vogue. It is potentially sup-ported by the important finding that ribonucleicacids (RNA) of plant and animal viruses are in-fectious even after their protein coats have beenstripped off with phenol (Gierer and Schramm,1956; Colter, 1958). Not only is the protein coatunessential, but it may actually serve to preventadsorption of enteroviruses and penetration oftheir RNA into insusceptible cells (Holland et al.,1959). In addition, Le Clerc (1956) has shownthat infected cells treated with ribonuclease haveimpaired capacities to produce influenza virus,although this effect could conceivably be causedby injury to cellular rather than to viral RNA.

It would be of great interest to learn whethervirus RNA can also interfere with virus multipli-cation. To the writer's knowledge the only evi-dence that parental virus RNA might be capableof initiating interference is indirect. Notwith-standing, inactivation studies with ionizing andultraviolet irradiation are not incompatible withthe thesis that the interfering property resides inthe nucleoprotein fraction of influenza virus(Powell and Pollard, 1956; Powell and Setlow,1956). Experiments by Tyrrell and Tamm (1955)suggest that 2,5-dimethylbenzimidazole, an anti-metabolite of nucleic acid, inhibits the interferingaction of heat-inactivated influenza virus. Itshould be noted, however, that some of theirdata can be interpreted as showing a direct effectof the benzimidazole compound on the cell aswell as on the interfering virus RNA. Of greatersignificance, perhaps, is the finding that incom-plete influenza virus from undiluted egg-passagematerial had progressively diminished interferingactivity and that incomplete virus produced inHeLa cells had no capacity to interfere whatso-ever (Paucker and Henle, 1958). Loss of theinterfering, property is correlated with deficiencyof internal S antigen, which, in turn, has beenshown to be related to nucleic acid content ofinfluenza virus (Ada and Perry, 1956). Therefore,reasoning by indirection, in order to initiate the

156 [VOL. 24

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

VIRAL INTERFERENCE

process that leads to interference, an influenzavirus particle must contain nucleic acid. Not onlyis the RNA-deficient incomplete virus an ineffi-cient interfering agent, it is also genetically de-fective and cannot participate in cross reactiva-tion (Gottlieb and Hirst, 1956).

It is reasonable to assume that competitive in-hibition will result if RNA of two viruses shouldenter the same cell. If, for example, the virusesare genetically related but not identical strains ofinfluenza, the yield of each might be reduced anda small proportion of the progeny could emergeas genetic recombinants or mixed phenotypes.To carry further the analogy to mutual exclusionof bacteriophage, the presence in the same hostcell of RNA molecules of two completely unre-lated viruses, such as NDV and WEE, shouldresult in a reduced yield of both (Levine, 1958)without genetic crossing. The extent to which themultiplication of each infectious component issuppressed would perforce depend on which viruswas dominant in the cell under study, on the ratioof the different types of parental particles in themixture, and on the interval of time between in-fection with each virus of the pair. It is onlylogical to suggest, as does Schlesinger (1959),"that interference in this system may involvedirect competition for cellular constituents (orfor limited sites?) required for replication of bothviruses."

Attractive as it may seem, there is some reasonfor questioning the validity, or at least the uni-versal applicability, of this thesis. Most of theinconsistencies are also cited by Schlesinger(1959). Foremost, perhaps, is the incontrovertiblefact that the infective and interfering propertiesare differentially susceptible to ultraviolet irradi-ation (Henle and Henle, 1947). If the inactivatedvirus particle is unable to impart the genetic in-formation required for production of new prog-eny, it is difficult to conceive how a sufficientquantity of "interfering" RNA can be formed tocompete with the superinfecting virus. It mayalso be paradoxical that the RNA components oftwo virus particles can cooperate in the produc-tion of recombinant progeny as well as competewith each other. In addition, the capacity of anRNA virus (influenza) to interfere with a DNAvirus (vaccinia), cited by Isaacs (1959) as anexample of heterologous interference, raises theintriguing question of whether competitive in-hibition can occur between nucleic acids with

presumed dissimilar metabolic pathways. Also ofnote is the finding by Schlesinger and Kuske(1959) that the interfering activity of influenzavirus is not reversed by treating the "interfered"cells with ribonuclease. The capacity of influenzavirus to protect mice against infection with equineencephalomyelitis viruses (Vilches and Hirst,1947) does not appear to qualify as an exampleof competitive RNA inhibition as the mechanismof the interference. Influenza virus multiplies inmouse brain largely in an incomplete form (Schle-singer, 1954) and, therefore, the resistant cells areconceivably deficient in "competing" influenzaRNA, although these cells appear to produce Santigen not incorporated into incomplete virus.To the reviewer's knowledge no studies have beenreported on the capacity of "naked" virus RNAto act directly as an interfering agent. However,Paucker and Henle (1958) made an unsuccessfulattempt to render cells resistant to infection byexposing them to internal nucleoprotein S antigenof influenza virus, although, as they point out,there was no assurance that the S antigen evenadsorbed on the cells.The conflicting evidence on the role of RNA in

viral interference can, perhaps, best be resolvedby postulating that cells form interfering sub-stances other than virus RNA. This hypothesisdoes not imply that RNA of the interfering virus,either in an intact or noninfectious form, is un-essential for initiating the processes that lead tocellular resistance. As noted subsequently, solubleinterfering substances, distinguishable from nu-cleic acids, are formed by cells exposed to irradi-ated and nonirradiated interfering viruses.

C. InterferonsExisting theories of the mechanisms of viral

interference must be reevaluated in the light ofthe discovery of interferon by Isaacs and hisassociates (Isaacs and Lindenmann, 1957; Lin-denmann et al., 1957; Isaacs et al., 1958; Burkeand Isaacs, 1958; Isaacs and Burke, 1958; Isaacs,1959). Interferon, as originally described by theseinvestigators, is a nonsedimentable productformed by interaction of inactivated influenzavirus and living cells. When transferred to nor-mal chick embryo cells, it renders them resistantto infection with myxoviruses and vaccinia virus.Reports by other investigators suggest that sim-ilar products of infected cells may interfere withthe viruses of 17 D yellow fever (Lennette and

19601 157

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

ROBERT R. WAGNER

Koprowski, 1946), poliomyelitis (Ho and Enders,1959), vesicular stomatitis (Henle et al. 1959),EEE (Wagner, 1959, unpublished data), and avariety of other neurotropic viruses (Porterfield,1959). Rather than attempting to review thepublished reports, manuscripts in press, and per-sonal communications, the significance of thesefindings will be evaluated by summarizing somestudies on a similar substance currently underinvestigation in the writer's laboratory. The dataare in general accord with the results obtained byothers, and represent confirmation and extensionof the work of Isaacs and his colleagues.Our interest in interferon was stimulated by the

observation that allantoic fluid infected with theWS strain of influenza A virus had a profoundinhibitory effect on multiplication of EEE virusin monolayer cultures of chick embryo fibro-blasts. A marked degree of interference betweenthese virus pairs had been noted previously bySchlesinger (1951) and others at a time whenquantitative methods for virus and cell assayswere not available. We found that infected al-lantoic fluid containing 108 EID50 of WS viruscompletely inhibited the cytopathic effect of 109plaque-forming units (pfu) of EEE virus. Thenext step was to subject the same infected al-lantoic fluid to several cycles of high speed cen-trifugation which reduced its content of influenzavirus to 100 EID50. The supernatant of thisfluid had exactly the same capacity to interferewith EEE virus as did the original infected fluid.This simple procedure of merely centrifuging in-fected allantoic fluid to prepare our interferonobviated the necessity for using ultraviolet-irradi-ated virus and surviving tissue fragments or cellcultures which have been required for preparationof other interferons. Although supernatant fluidsfrom different pools of infected allantoic fluidhave varied somewhat in potency, none has beendevoid of interferon activity, and the titers asmeasured by dilution have generally been 10 to100 times greater than those reported from otherlaboratories. It should be emphasized that this isprobably a reflection of the sensitivity of theassay method in which EEE virus is used ratherthan other test viruses. Our preparation of inter-feron does not inhibit plaque formation by NDVin monolayers of the same chick embryo cells,whereas infectious influenza virus does so readily.Isaacs and Westwood (1959) and Porterfield(1959) have recently reported that Arbor viruses

are far more sensitive to the action of interferonthan are myxoviruses.

Also of inestimable convenience was the findingthat the concentration of interferon is more im-portant than the challenge dose of EEE viruswhen assayed on chick embryo monolayers by theplaque-inhibition method. No significant differ-ence could be detected in the capacity to inhibit1 or 109 pfu. Thus, we are able to standardize ourtest system by using a constant input of 40 to 50pfu, a convenient number for counting. The titerof interferon is read as the 2-fold dilution thatreduces the number and size of plaques by ap-proximately half. The error of the assay methodon replicate plating is about 50 per cent, whichcompares favorably with most serologic tests. Offurther advantage is the fact that diluted inter-feron can be added to the test cultures simultane-ously with the EEE virus without loss of inhibi-tory activity, provided that the cell layers arenot washed.

It was next incumbent upon us to determinewhether interferon is a cell product rather thandegraded virus. Four lines of evidence, parallelto those cited by Isaacs, suggest that interferonis not derived from influenza virus particles perse: hyperimmune sera with high titers of anti-V(antihemagglutinin prepared in rabbits) or anti-S(soluble CF antibody prepared in guinea pigs) do

TABLE 1Evidence that interferon is not

influenza virus protein

WS VirusInfected

Allantoic Fluid

WholeSupernatant

Supernatant

Supernatant

Sediment

Treatment

NoneAnti-V anti-body

Anti-S anti-body

Red bloodcells ad-sorbed

Red bloodcellseluted

LogEIDsoTiter ofWsVirus

8.0<1.0

2.5

1.5

7.5

InterferenceTiter*

Unheated| 65 C

256256

192

192

16

256

192

0

* Reciprocal of 2-fold dilution that produces 50per cent inhibition of 40 to 50 plaque-formingunits (pfu) of eastern equine encephalomyelitis(EEE) virus.

158 [VOL. 24

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

VIRAL INTERFERENCE

40 96 -

1024 -a

256-

0

Z 64-

-J

16 -z

cnI 4-w

I--

C

/\INFLUENZA ,"/ VIRUS ,'

/ s~~~o'I ,,~~~~~~~

/+_ I NTERFERON

-//I_

I24 48 72

HOURS AFTER INFLUENZA INFECTION

Figure 1. Presence of hemagglutinin and inter-feron in the allantoic fluid of the same chick em-

bryos at intervals after inoculation with infec-tious WS influenza virus.

not neutralize interferon activity, interferon doesnot adsorb on chicken erythrocytes, and the in-terfering actions of purified influenza virus andsemipurified interferon can be differentiated byheat lability. These data are summarized intable 1.A fourth, and perhaps more convincing, form

of evidence was obtained by studying the rate atwhich infected allantoic cells produce influenzavirus and interferon. In this experiment, thehemagglutinin and interferon content of allantoicfluids were measured at intervals after infectionwith WS influenza virus. Figure 1 demonstratesthat large amounts of virus appear in the allan-toic fluid before any interferon can be detected.To rule out the possibility that interferon was

formed by thermal inactivation of virus in ovo,

samples of allantoic fluid from the early stages ofinfection were incubated further at 37 C in vitro.This resulted in diminished rather than increasedinterfering activity of the virus. Thus, it appears

from this and the preceding study that allantoiccells infected with influenza virus produce twointerfering agents, heat-labile virus and heat-stable, nonsedimentable interferon.

Before interferon could be implicated as an

intermediary substance responsible for interfer-ence with EEE virus, definite evidence of itsintracellular site of action was required. It was

readily ascertained that interferon does not in-

activate extracellular EEE virus. This was donesimply by demonstrating that infectivity is com-pletely restored following dilution of a noninfec-tious mixture of EEE virus and interferon. Next,it was found that interferon reacts with host cells.This was accomplished by determining that therate at which interferon adsorbs on chick embryofibroblasts coincides, at least approximately, withthe rate at which these cells develop resistance toinfection with EEE virus. The longer its periodof contact with cells, the more interferon is ad-sorbed and the greater the degree of cellular re-sistance to virus challenge. After prolonged con-tact, susceptibility cannot be restored by washingthe cells, partial evidence that interferon does notmerely adsorb on the surface but penetrates thecell. If the resistant cells are disrupted by alter-nate freezing and thawing, none of the adsorbedinterferon can be recovered. This fact suggeststhat interferon is rapidly "eclipsed" (metab-olized?) and that, unlike virus, it does not stim-ulate the cell to produce more interferon.The next question that arose was whether in-

terferon renders cells resistant to infection byaltering their capacity to adsorb EEE virus. Al-though some difficulty was encountered in ob-taining reproducible virus adsorption curves, nodifferences could be detected in the degree towhich EEE virus adsorbs on susceptible cellsand on cells rendered resistant to infection bytreatment with interferon.

Conclusive proof that interferon acts by in-hibiting intracellular synthesis of EEE virus wasobtained by comparing curves of virus growth insusceptible and resistant cells. In these experi-ments, cultures of chick embryo fibroblasts weretreated with interferon for various periods of

TABLE 2Effect of interferon on multiplication of eastern

equine encephalomyelitis (EEE) virus in chickembryo fibroblasts

Time at Which Cells Were Treated with Average Yield of256 Units of Interferon EEE Virus per cell

Pfu*Controls (no interferon) 50000 time (simultaneous) .......... 252 hr before infection ........... 2-312 hr before infection .......... <11 hr after infection............. 2-10

* Plaque-forming units.

1591960]

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

ROBERT R. WAGNER

time before or after infection with a multiplicityof approximately 5 EEE virus particles per cell.A summary of the results is shown in table 2.The striking effect of interferon on virus multi-plication is clearly evident. These data also sup-

port the contention that the inhibitory action ispartially dependent on the duration of contactwith cells. It also seems safe to assume that ex-

posure to interferon for 12 hr can suppress thevirus-producing capacity of almost every cell inthe culture and that a significant proportion ofthe resistant cells are incapable of forming any

virus. The most important information derivedfrom these experiments is that interferon affectsintracellular virus by inhibiting its multiplicationafter adsorption on cells. Evidence that most ofthe virus had penetrated the cells 1 hr after infec-tion was obtained by finding that treatment withimmune serum did not appreciably reduce thenumber of infective centers.

Perhaps the most important problem for futurestudy is to determine whether interferon preventssynthesis of EEE virus by direct inactivation ofeclipsed parental virus RNA or indirectly alterscellular metabolism of virus nueleic acid and pro-

tein. Although we have no satisfactory quantita-tive data as yet, it does not appear that prolongedexposure to interferon significantly affects thegeneration time, plating efficiency, or carbohy-drate metabolism2 of resistant cells. It will be ofconsiderable interest to ascertain whether inter-feron is capable of inactivating infectious RNAof EEE virus separated from its protein coat byphenol extraction (Wecker and Schafer, 1957).

It is perhaps all too apparent that any furtherspeculation about the mechanisms by which in-terferon inhibits virus multiplication must besupported by analysis of its chemical nature andreactivity. Unfortunately, only the most prelim-inary information on the physicochemical prop-

erties of interferon is available at present. Theresults of some of these studies are summarizedin table 3. The data indicate that interferon is a

protein-like compound, probably of relatively lowmolecular weight, that is somewhat more stableto heat than the preparation originally described

2 The implication of this statement is that in-terferon does not appear to decrease the metabolicactivity of the cells. Since this review was pre-

pared, Isaacs (1960 and personal communication)has reported increased aerobic glycolysis of chickfibroblasts treated with interferon.

TABLE 3Some physicochemical properties of interferon

Property Result

Ultracentrifugal sedi- Unaffected at 100,000 Xmentation G for 4 hr

Dialysis Not dialyzable(NH4) 2SO4 precipita- 90% precipitated attion 60% saturation

Enzyme susceptibility Destroyed by trypsinor chymotrypsin

Enzyme resistance Not affected by RNase,DNase, papain, plas-min, or RDE

Heat stability Only 90% destroyed at85 C for 1 hr; com-pletely stable at 70 Cfor 1 hr

pH stability Stable at pH 3-11,> 90% destroyed atpH 12.5, 67% de-stroyed at pH 1.0

Adsorption on bento- Completely adsorbednite

Elution from bentonite Only 25% eluted bypyridine at pH 9 orabove

UV absorption spec- No UV absorption attrum any wave length

Abbreviations: RNase, ribonuclease; DNase,deoxyribonuclease; RDE, receptor-destroying en-zyme; UV, ultraviolet.

by Isaacs and his colleagues. (A recent personalcommunication from Isaacs indicates that heatstability is influenced by pH of the suspendingmedium.) The failure of interferon to absorbultraviolet light at wave lengths 260 to 290 mgusuggests that it contains no tryptophan, tyrosine,phenylalanine, or nucleic acid constituents. Itsmarked stability in acid and the fact that it canbe eluted from bentonite only at pH 9 or above isconsistent with the thesis that interferon is abasic protein or polypeptide. Additional studiesindicated that it is not lysozyme and that its bio-logical activity could not be stimulated or blockedby calf thymus histone, RNA, DNA, RNase, orDNase. Therefore, we are left with the tentativehypothesisthat interferon resembles a basic proteinof the histone variety.

It is of interest that basic protein moieties ofnucleohistones are biologically much like nucleicacids in their tendency to exhibit species rather

160 [VOL. 24

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

VIRAL INTERFERENCE

TABLE 4Species specificity of interferons

Interference with EasternEquine Encephalomyelitis

(EEE) VirusSource of Interferons E__Virus

Chick Duck Mousecells cells brain

Chick embryo allantois.... + 0 0Duck embryo allantois 0 + NT*Mouse brain .............. 0 NT* +

* NT = not tested.

than organ specificity (Brachet, 1957). Studies ofthe specificity of interferon prepared in variousspecies of animals furnish additional circumstan-tial evidence of its biological resemblance to his-tones. In these experiments chick embryos, duckembryos, and mice infected with WS influenzavirus served as three different sources of inter-feron. Infected allantoic fluids and mouse brainsuspensions were centrifuged at 100,000 X G tosediment most of the influenza virus and thesupernatant fluids were then tested for their ca-

pacity to interfere with EEE virus in chick or

duck embryo cell cultures. Comparative studieshave not yet been made in mouse embryo cellsbecause they do not support multiplication of thechick-adapted strain of EEE virus used in theseexperiments. A general pattern of specificity ofinterferons prepared in different animal species isshown in table 4. These findings are closely anal-ogous to those recently reported by Isaacs andWestwood (1959) who found little, if any, cross

resistance to vaccinia infection of rabbit kidneyor chick membranes treated with interferons pre-pared in the heterologous species.Mention should be made of the difficulties

encountered in demonstrating the effect of inter-ferons in intact animals. In the foregoing experi-ments, definite but only slight resistance to cere-

bral infection with EEE virus could be elicited bypretreatment with mouse brain interferon. Isaacset al. (1958) noted limited protection against vac-

cinial infection of chorioallantoic membranes ofintact chick embryos previously injected withinterferon. Death of chick embryos infected withEEE virus can also be prevented by treatmentwith interferon, but the most potent preparationsafforded only a 30-fold reduction of LD50 titer ofthe challenge virus (Wagner, 1959, unpublisheddata). By comparison, chick embryo cell cultures

could readily be made to resist 109 pfu of EEEvirus. This discrepancy can undoubtedly be at-tributed to considerable dilution of interferon inthe extraembryonic and extracellular fluids of theintact host, and to the large number of cells po-tentially susceptible to infection.

Despite the technical difficulties inherent insuch an approach, the observation that interfer-ence can in fact be "passively transferred" tointact animals may be of considerable theoreticalsignificance. These studies call to mind the ques-tion (Schlesinger, 1959) whether viral interferenceis purely a local cellular phenomenon or can occurat a tissue site distant from a primary infectionwith interfering virus. Burnet and Fraser (1952)addressed themselves to this question in theirstudies of resistance to cerebral infection in chickembryos. They found that prior allantoic infec-tion with influenza virus prevented cerebral hem-orrhages in embryos challenged intravenouslywith neurotropic influenza virus. Noting thatembryo brain tissue contained an insufficientamount of interfering virus to account for theprotective effect, they postulated the existence ofa "limiting factor" in the circulation of the re-sistant embryos. In retrospect, this "limitingfactor" could be interferon, which, being a muchsmaller molecule than the influenza virus, couldconceivably pass more readily from the allantoiccavity to the circulating blood and thence tocerebral blood vessels. Confirmation of these hy-potheses is being sought in this laboratory (Hookand Wagner, 1958; Grossberg, Hook, and Wag-ner, 1959, unpublished data). We have come tothe same conclusion expressed by Burnet andFraser: that resistance to hemorrhagic encephalo-pathy in chick embryos cannot be explainedsolely in terms previously considered as "clas-sical" interference between two viruses infectingthe same cells. Although certain technical diffi-culties must still be surmounted before obtainingconclusive proof, it appears from preliminarystudies that allantoic injection of interferon af-fords chick embryos slight protection againstcerebral hemorrhages caused by intravenous in-jection of neurotropic influenza virus. Gledhill(1959 and personal communication) also is seek-ing to determine by passive transfer studieswhether interference with ectromelia by mousehepatitis virus is caused by the presence of aninterferon in the blood of resistant mice.

19601 161

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

ROBERT R. WAGNER

D. Resistance to Superinfection of PersistentlyInfected Cell Cultures

Virologists have been intrigued by the knowl-edge that animal viruses can persist in tissues orcell cultures for long periods of time without pro-ducing overt manifestations of infection. Thequestion as to whether some of these persistentinfections represent examples of true lysogenyremains unanswered. The conditions for estab-lishment of persistent viral infections vary con-siderably but often depend on factors such asvirulence of the virus (Sabin, 1954), temperatureof incubation and growth rate of cells (Bang etal., 1957), and the presence in the supportingculture media of antibody or other antiviral sub-stances (Ginsberg, 1958). More pertinent to thepresent discussion is the report by Chambers(1957) that persistence of WEE virus in culturesof L cells could not be attributed to alterationsof the virus or the cells, or to environmental fac-tors. It is her contention that this chronic infec-tion can best be explained by autointerference.

Persistence of myxoviruses in stable cell linesand resistance of these cultures to superinfectionhave been the subjects of a comprehensive seriesof reports published from Henle's laboratory(Henle et al., 1958; Bergs et al., 1958; Deinhardtet al., 1958). Their findings are deemed to beparticularly pertinent to this discussion of theinterference phenomenon and will be summarizedbriefly. The chronicity of the infections withmyxoviruses is illustrated by the failure to affecta ''cure") even after long exposure to immuneserum, although specific antibody does suppressthe virus temporarily. The cells in these per-sistently infected cultures grow and divide some-what more slowly than uninfected cells and ex-hibit increased aerobic glycolysis and concomitantaccumulation of lactic acid (Green et al., 1958).However, these factors do not explain an extraor-dinary degree of resistance to superinfectionwith vesicular stomatitis virus (VSV). The latentmyxovirus itself does not exhibit any significantchange in its biological properties, nor do unin-fected clones from these chronically infected cul-tures. VSV readily adsorbs on and penetratesresistant cells, indicating that their failure tosupport multiplication of superinfecting viruscannot be attributed to alteration of surface re-ceptor sites. The most interesting fact is that theamount of myxovirus present in the chronicallyinfected cultures is insufficient to account for re-

X ........ . ....... ........ .............-% gfS I *:. E I..._. ...........

...........I t.....,., ......... .

- IVIRUS

'f INTERFERON

0° TIME-

Figure 2. Theoretical schema of the potentialequilibrium between virus and interferon pro-duced by cells in a persistently infected culture.Infected cells are constantly being replaced bygrowth of susceptible uninfected cells at a reducedrate. If the production of virus outstrips that ofinterferon, all the susceptible cells will die. Ifinterferon production is greater than virus multi-plication, the infection will be cured.

sistance to superinfection. Fewer than 10 per centof the cells contain even a single myxovirus par-ticle at any one time and no incomplete virus canbe detected. Therefore, it is necessary to postulatethe existence in these cultures of an interferingagent other than the persistent myxovirus itself.A soluble substance, produced in cell culturesafter exposure to myxoviruses, that inhibits mul-tiplication of VSV and other viruses has recentlybeen identified in Henle's laboratory as an inter-feron (Henle et al., 1959; and personal communi-cation).

Henle and his colleagues should not be heldaccountable for the following interpretation oftheir findings, for which the reviewer takes fullresponsibility. It seems entirely possible that acell infected with influenza virus can produceeither new virus progeny or interferon, or both.Thus, an equilibrium might be established in thepersistently infected culture between these twoantagonistic products of cell infection. If thevirus is temporarily in the ascendancy, it maystimulate certain cells to produce interferon. Ifthe rate of interferon formation becomes exces-sive, it might result in a decreased virus titer,thus removing the stimulus for further productionof interferon. Consequently, the concentration ofinterferon, which is not a self-replicating sub-stance, should decline and the virus increase. Inthis way it is conceivable that a persistent lowgrade infection can be established in cell cultures

162 [VOL. 24

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

VIRAL INTERFERENCE

by virtue of cyclic production of both virus andinterfering substance in response to viral infec-tion. A diagrammatic representation of this hy-pothetical situation is shown in figure 2. Ob-viously, this theory would be subject to widerevision should further studies of persistent in-fections of cell cultures reveal complete absenceof interferons in fluid and cellular phases of thecultures. Thus far, however, there seem to be noobvious inconsistencies in studies of persistentinfections of cell cultures with other viruses.

IV. SUMMARY AND THEORY

Interference appears to be one mechanism bywhich the bacterial or animal cell can defend it-self against viral infection. It is unlikely that re-sistance to superinfection takes place at the sur-face of the cell, but it almost certainly does sointracellularly. There is insufficient evidence toimplicate genetic or metabolic factors as explana-tions for competitive antagonism between nu-cleic acid moieties of two viruses within the samecell. However, the nucleic acid of the interferingvirus may well be essential for initiating the cel-lular response that leads to interference. If this bethe case, the virus contains in its nucleic acid thepotential information for its own destruction,mediated by the cellular defenses of the host.Certain interfering viruses stimulate the cell toelaborate protein-like substances of nonviral ori-gin that prevent superinfection with homologousor heterologous viruses. These substances, theinterferons, can be secreted by an infected celland transmitted to other cells, thereby renderingthem resistant to infection. Presumably, similarevents can transpire in a persistently infected cellculture or an intact animal. It can be predictedwith confidence that a considerable amount offuture research will be directed toward giving thehost an added advantage by passive transfer ofthe antibiotic-like interfering substances, theinterferons.

V. REFERENCES

ADA, G. L. AND PERRY, B. T. 1956 Influenzavirus nucleic acid: Relationship between bio-logical characteristics of the virus particleand properties of the nucleic acid. J. Gen.Microbiol., 14, 623-633.

ADAMS, M. H. 1959 Bacteriophages. Intersci-ence Publishers, Inc., New York.

AMES, B. N., DUBIN, D. T., AND ROSENTHAL, S.

M. 1958 Presence of polyamines in certainbacterial viruses. Science, 127, 814-816.

BALUDA, M. A. 1957 Homologous interferenceby ultraviolet-irradiated Newcastle diseasevirus. Virology, 4, 72-96.

BALUDA, M. A. 1959 Loss of viral receptors inhomologous interference by ultraviolet irradi-ated Newcastle disease virus. Virology, 7,315-327.

BANG, F. B., GEY, G. O., FOARD, M., AND MINNE-GAN, D. 1957 Chronic infections producedin cultured cell strains by the virus of easternequine encephalomyelitis. Virology, 4, 404-417.

BERGS, V. V.. HENLE, G., DEINHARDT, F., ANDHENLE, W. 1958 Studies on persistent in-fections of tissue cultures. II. Nature of theresistance to vesicular stomatitis virus. J.Exptl. Med., 108, 561-572.

BERTANI, G. 1953 Infections bacteriophagiquessecondaires des bacteries lysogenes. Ann.inst. Pasteur, 84, 273-280.

BERTANI, G. 1956 The role of phage in bacterialgenetics. Brookhaven Symposia Biol., 8,50-56.

BRACHET, J. 1957 Biochemical cytology. Aca-demic Press, Inc., New York.

BURKE, D. C. AND ISAACS, A. 1958 Some fac-tors affecting the production of interferon.Brit. J. Exptl. Pathol., 39, 452-458.

BURNET, F. M. AND FRASER, K. B. 1952 Studieson recombination with influenza viruses in thechick embryo. I. Invasion of the chick embryoby influenza viruses. Australian J. Exptl.Biol. Med. Sci., 30, 447-458.

BURNET, F. M. AND LIND, P. E. 1951 A geneticapproach to variation in influenza viruses. 4.Recombination of characters between the in-fluenza virus A strain NWS and strains of dif-ferent serological subtypes. J. Gen. Micro-biol., 5, 67-82.

CHAMBERS, V. C. 1957 The prolonged persist-ence of western equine encephalomyelitisvirus in cultures of strain L cells. Virology,3, 62-75.

COHEN, S. S. AND ARBOGAST, R. 1950 Chemicalstudies in host-virus interactions. VIII. Themutual reactivation of T2r+ virus inactivatedby ultraviolet light and the synthesis of de-oxyribose nucleic acid. J. Exptl. Med., 91,637-650.

COLTER, J. S. 1958 Nucleic acid as the carrierof viral activity. Progr. Med. Virol. 1, 1-35.

DEINHARDT, F., BERGS, V. V., HENLE, G., ANDHENLE, W. 1958 Studies on persistent in-fections of tissue cultures. III. Some quanti-tative aspects of host cell-virus interactions.J. Exptl. Med., 108, 573-589.

16319601

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

ROBERT R. WAGNER

DELBRUCK, M. 1940 The growth of bacterio-phage and lysis of the host. J. Gen. Physiol.,23, 643-660.

DELBRUCK, M. 1945 Interference between bac-terial viruses. III. The mutual exclusion effectand the depressor effect. J. Bacteriol., 50,151-170.

DELBRfTCK, M. AND LURIA, S. E. 1942 Inter-ference between bacterial viruses. I. Inter-ference between two bacterial viruses actingupon the same host, and the mechanism ofvirus growth. Arch. Biochem., 1, 111-141.

DOERMANN, A. H. 1948 Lysis and lysis inhibi-tion with Escherichia coli bacteriophage. J.Bacteriol., 55, 257-276.

DULBECCO, R. 1949a The number of particles ofbacteriophage T2 that can participate in in-tracellular growth. Genetics, 34, 126-132.

DULBECCO, R. 1949b Mutual exclusion betweenrelated phages. J. Bacteriol., 63, 209-217.

FAZEKAS DE ST. GROTH, S. AND CAIRNS, H. J. F.1952 Quantitative aspects of influenza virusmultiplication. IV. Definition of constantsand general discussion. J. Immunol., 69,173-181.

FAZEKAS DE ST. GROTH, S. AND GRAHAM, D. M.1954 The production of incomplete virusparticles among influenza strains: experimentsin eggs. Brit. J. Exptl. Pathol., 35, 60-74.

FRENCH, R. C., GRAHAM, A. F., LESLEY, S. M.,AND VAN ROOYEN, C. E. 1952 The contribu-tion of phosphorus from T2r+ bacteriophageto progeny. J. Bacteriol., 64, 597-607.

GIERER, A. AND SCHRAMM, G. 1956 Der Infek-tiositht der Nucleinsaure aus Tabakmosaik-virus. Z. Naturforsch., 116, 138-142.

GINSBERG, H. S. 1958 A consideration of therole of inhibitors in latency and analysis ofpersistent adenovirus infections of mamma-lian cells. In A symposium on latency andmasking in viral and rickettsial infections, pp.157-168. Burgess Publishing Co., Minneapo-lis.

GLEDHILL, A. W. 1959 The interference ofmouse hepatitis virus with ectromelia inmice and a possible explanation of its mech-anism. Brit. J. Exptl. Pathol., 40, 291-300.

GOTLIEB, T. AND HIRST, G. K. 1956 The exper-imental production of combination forms ofvirus. VI. Reactivation of influenza virus afterinactivation by ultraviolet light. Virology,2, 235-248.

GRANOFF, A. 1959 Mixed infection studies withmutants of Newcastle disease virus (NDV).Federation Proc., 18, 571 (abst.).

GREEN, M., HENLE, G., AND DEINHARDT, F. 1958Respiration and glycolosis of human cells

grown in tissue culture. Virology, 5, 206-219.

GROUP1, V., HERRMANN, E. C., JR., AND DOUGH-ERTY, R. M. 1954 Protection of mice fromneurotoxic action of influenza virus by "heatinactivated" receptor destroying enzyme.Proc. Soc. Exptl. Biol. Med., 87, 636-37.

HENLE, G., GIRARDI, A., AND HENLE, W. 1955A non-transmissible cytopathogenic effect ofinfluenza virus in tissue culture accompaniedby formation of non-infectious hemmagglu-tinins. J. Exptl. Med., 101, 25-41.

HENLE, G., DEINHARDT, F., BERGS, V. V., ANDHENLE, W. 1958 Studies on persistent in-fections of tissue cultures. I. General aspectsof the system. J. Exptl. Med., 108, 537-560.

HENLE, W. 1950 Interference phenomena be-tween animal viruses: a review. J. Immunol.,64, 203-236.

HENLE, W. AND HENLE, G. 1947 The effect ofultraviolet irradiation on various propertiesof influenza viruses. J. Exptl. Med., 85, 347-364.

HENLE, W. AND ROSENBERG, E. B. 1949 One-step growth curves of various strains of in-fluenza A and B viruses and their inhibitionby inactive virus of the homologous type.J. Exptl. Med., 89, 279-285.

HENLE, W., HENLE, G., AND KIRBER, M. W. 1947Interference between inactive and activeviruses of influenza. V. Effect of irradiatedvirus on the host cells. Am. J. Med. Sci..214, 529-541.

HENLE, W., HENLE, G., DEINHARDT, F., ANDBERGS, V. V. 1959 Studies on persistentinfections of tissue cultures. V. Evidence forthe production of an interferon in MCN cellsby myxoviruses. J. Exptl. Med., 110, 525-541.

HERSHEY, A. D. 1955 An upper limit to theprotein content of the germinal substance ofbacteriophage T2. Virology, 1, 108-127.

HERSHEY, A. D. 1957 Some minor componentsof bacteriophage T2 particles. Virology, 4,237-264.

HERSHEY, A. D. AND CHASE, M. 1952 Independ-ent functions of viral protein and nucleic acidin growth of bacteriophage. J. Gen. Physiol.,36, 39-56.

HERSHEY, A. D., GAREN, A., FRASER, D. K., ANDHUDIS, J. D. 1954 Growth and inheritancein bacteriophage. Carnegie Inst. Wash.Yearbook, Publ. No. 53, 210-225.

HIRST, G. K. AND GOTLIEB, T. 1955 The ex-perimental production of combination formsof virus. V. Alterations in the virulence ofneurotropic influenza virus as a result of mixedinfection. Virology, 1, 221-235.

164 [VOL. 24

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

VIRAL INTERFERENCE

Ho, M. AND ENDERS, J. F. 1959 An inhibitor ofviral activity appearing in infected cell cul-tures. Proc. Natl. Acad. Sci. U. S., 45, 385-389.

HOLLAND, J. J., MCLAREN, L. C., AND SYVERTON,J. T. 1959 The mammalian cell-virus rela-tionship. IV. Infection of naturally insuscep-tible cells with enterovirus ribonucleic acid.J. Exptl. Med., 110, 65-80.

HOOK, E. W. AND WAGNER, R. R. 1958 Hemor-rhagic encephalopathy in chicken embryosinfected with influenza viruses. I. Factors in-fluencing the development of hemorrhages.Bull. Johns Hopkins Hosp., 103, 125-139.

ISAACS, A. 1957 Particle counts and infectivitytitrations for animal viruses. Advances inVirus Research, 4, 111-158.

ISAACS, A. 1959 Viral interference. Sympo-sium Soc. Gen. Microbiol., 9, 102-121.

ISAACS, A. 1960 Metabolic effects of interferonon chick fibroblasts. Virology, 10, 144-146.

ISAACS, A. AND BURKE, D. C. 1958 Mode ofaction of interferon. Nature, 182, 1073-1074.

ISAACS, A. AND EDNEY, M. 1950 Interferencebetween inactive and active influenza virusesin the chick embryo. I, II, and IV. Austral-ian J. Exptl. Biol. Med. Sci., 28, 219-230; 231-238; 635-645.

ISAACS, A. AND LINDENMANN, J. 1957 Virus in-terference. I. The interferon. II. Some prop-erties of interferon. Proc. Royal Soc. B. 147,258-267 and 268-273.

ISAACS, A. AND WESTWOOD, M. 1959 Inhibitionby interferon of the growth of vaccinia virusin the rabbit skin. Lancet, 2, 324-325.

ISAACS, A., BURKE, D. C., AND FADEEVA, L. 1958Effect of interferon on the growth of viruseson the chick chorion. Brit. J. Exptl. Pathol.,39, 447-451.

JACOB, F. AND WOLLMAN, E. L. 1957 Geneticaspects of lysogeny. In The chemical basis ofheredity, pp. 468-499. Edited by W. D. Mc-Elroy and B. Glass. Johns Hopkins Press,Baltimore.

LE CLERC, J. 1956 Action of ribonuclease on themultiplication of the influenza virus. Na-ture, 177, 578-579.

LEDERBERG, S. 1957 Suppression of the multi-plication of heterologous bacteriophages inlysogenic bacteria. Virology, 3, 496-513.

LENNETTE; E. H. AND KoPROWSKI, H. 1946 In-terference between viruses in tissue culture.J. Exptl. Med., 83, 195-219.

LESLEY, S. M., FRENCH, R. C., GRAHAM, A. F.,ANDVAN ROOYEN, C. E. 1951 Studies on therelationship between virus and host cell. II.The breakdown of T2r+ bacteriophage upon

infection of its host Escherichia coli. Cana-dian J. Med. Sci., 29, 128-143.

LEVINE, L., BARLOW, J. L., AND VAN VUNAKIS, H.1958 An internal protein in T2 and T4 bac-teriophages. Virology, 6, 702-717.

LEVINE, S. 1958 Dynamics of heterologous in-terference between viable viruses in chickembryo fibroblast monolayers. Virology, 5,150-167.

LINDENMANN, J., BURKE, D. C., AND ISAACS, A.1957 Studies on the production, mode ofaction and properties of interferon. Brit. J.Exptl. Pathol., 38, 551-562.

LWOFF, A. 1953 Lysogeny. Bacteriol. Revs.,17, 269-337.

LWOFF, A. AND GUTMANN, A. 1950 Recherchessur un Bacillus megatherium lysogene. Ann.inst. Pasteur, 78, 711-739.

PAUCKER, K. AND HENLE, W. 1958 Interferencebetween inactivated and active influenzaviruses in the chick embryo. II. Interferenceby incomplete forms of influenza virus. Vi-rology, 6, 198-214.

POWELL, W. F. AND POLLARD, E. C. 1956 Theeffect of ionizing radiation on the interferingproperty of influenza virus. Virology, 2, 321-336.

POWELL, W. F. AND SETLOW, R. B. 1956 Theeffect of monochromatic ultraviolet radiationon the interfering property of influenza virus.Virology, 2, 337-343.

PORTERFIELD, J. S. 1959 A simple plaque assayinhibition test for antiviral agents: applica-tion to assay of interferon. Lancet, 2, 326-327.

PUCK, T. T. AND LEE, H. H. 1955 Mechanismof cell wall penetration by viruses. II. Demon-stration of cyclic permeability changes accom-panying virus infection of Escherichia coli Bcells. J. Exptl. Med., 101, 151-175.

SABIN, A. B. 1954 Noncytopathogenic variantsof poliomyelitis virus and resistance to super-infection in tissue culture. Science, 120, 357.

SCHLESINGER, R. W. 1951 Studies on interfer-ence between influenza and equine encephalo-myelitis viruses. Arch. ges. Virusforsch., 4,501-517.

SCHLESINGER, R. W. 1954 Incomplete growthcycle of influenza virus in mouse brain. Proc.Soc. Exptl. Biol. Med., 74, 541-548.

SCHLESINGER, R. W. 1959 Interferences be-tween animal viruses. In The viruses: Bio-chemical, biological and biophysical properties,vol. 3. Edited by F. M. Burnet and W. M.Stanley. Academic Press, Inc., New York.

SCHLESINGER, R. W. AND KUSKE, T. T. 1959Cited in R. W. Schlesinger, The Viruses, vol.

16519601

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

ROBERT R. WAGNER

3, pp. 157-194. Edited by F. M. Burnet andW. M. Stanley. Academic Press, Inc., NewYork.

STONE, J. D. 1948 Prevention of virus infectionwith enzyme of V. cholerae. I and II. Aus-tralian J. Exptl. Biol. Med. Sci., 26, 49-64;287-298.

TYRRELL, D. A. J. AND TAMM, I. 1955 Preven-tion of virus interference by 2,5-dimethyl-benzimidazole. J. Immunol. 75, 43-49.

VILCHES, A. AND HIRST, G. K. 1947 Interferencebetween neurotropic and other unrelatedviruses. J. Immunol., 57, 125-140.

VISCONTI, N. 1953 Resistance to lysis fromwithout in bacteria infected with T2 bacterio-phage. J. Bacteriol., 66, 247-253.

VON MAGNUS, P. 1951 Propagation of the PR8strain of influenza A virus in chick embryos.II. Formation of "incomplete" virus followinginoculation of large doses of seed virus. ActaPathol. Microbiol. Scand., 28, 278-293.

WAGNER, R. R. 1952 Acquired resistance inmice to the neurotoxic action of influenzaviruses. Brit. J. Exptl. Pathol., 33, 157-167.

WAGNER, R. R., SNYDER, R. M., HOOK, E. W.,ANDLUTTRELL,C.N. 1959 Effect of bacterialendotoxin on resistance of mice to viral en-cephalitides. J. Immunol., 83, 87-98.

WATSON, J. D. 1950 The properties of X-rayinactivated bacteriophage. I. Inactivation bydirect effect. J. Bacteriol., 60, 697-718.

WECKER, E. AND SCHXFER, W. 1957 Eine infek-ti6se Komponente von Ribonucleinshure-Charakter aus dem Virus der AmerikanischenPferde-Encephalomyelitis (Type Ost). Z.Naturforsch., 12b, 415-417.

WEIGLE, J. AND DELBRtCK, M. 1951 Mutual

exclusion between an infecting phage anda carried phage. J. Bacteriol., 62, 301-318.

DISCUSSION

Referring to figure 2 it was indicated that Tyr-rell (Salisbury, England), has demonstrated thephenomenon of repeatedly emerging and disap-pearing cytopathic virus production in mouse kid-ney tissue culture infected with WS influenzavirus (Westwood, Porton, England).A factor inhibiting the cytopathic effect of sev-

eral viral strains has been demonstrated in cul-ture fluids of human kidney cells infected with achick embryo adapted strain of Type II polio-virus. This inhibitor, or type of interferon, isseparable from infective virus and may be adeterminant in chronic cell infection in vitro andin vivo. The development of chronic infectioncaused by such viruses may be associated withthe so-called "zone phenomenon," referring tothe ability of certain viruses to proliferate andcause cell destruction when in low concentrationbut not when in high concentration. For example,the type II poliovirus adapted to chick embryo(RMC virus) causes destruction of human am-nion cells in high dilution. Undiluted virus, how-ever, does not destroy the cells but produceschronic cell infection (Ho, Pittsburgh) (Ho, M.and Enders, J. F., An inhibitor of viral activityappearing in infected cell cultures, Proc. Natl.Acad. Sci. (U. S.), 45, 385-389, 1959).

In considering virus interfering substancesthere are two factors that are active; one is thevirus itself, which is heat labile and can be sep-arated from interferon, which is heat stable(Wagner, Baltimore).

166 [VOL. 24

on April 13, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from