Postgraduate Medical Joiurnal (September 1978) 54, 581-586.

General features of persistent virus infections

C. A. MIMSB.Sc., M.D., F.R.C.Path.

Department of Microbiology, Guy's Hospital Medical School, London SEJ 9RT

SummaryPersistent virus infections are discussed from the viruspoint of view in terms of the bodily sites in which theinfection persists. Glands and body surfaces arethought to be significant because they give the virusprotection at the topographical level from immuneforces, and because they are appropriate sites for theshedding of virus to the exterior. Germ cells are rele-vant sites because infection can thus be transmittedvertically from generation to generation in the host.The central nervous system, however, is generally a'dead end' from which there is no shedding to theexterior. Persistence in blood may be relevant whencontinued arthropod transmission becomes possible.Most persistent viruses infect lymphoreticular tissues,and this is interpreted by suggesting that it results inan impaired immune response to the infecting virus,which in turn favours persistence.

It is suggested that the biological function of virustransformation and the integration of viral into hostcell DNA is that it enables the infection to persist inthe host and undergo reactivation. Papovaviruses,adenoviruses and oncornaviruses are considered fromthis point of view.

IntroductionThere are three groups of viruses that seem to

have an assured future. First, the respiratory viruses,which are evolving rapidly and entering theirgolden age as populations increase and give analmost unlimited supply of susceptible hosts livingin the crowded conditions that favour transmission.Vaccines for all but the most pathogenic are animpossible dream because of the great variety ofserotypes; chemotherapy seems a long way away,and the best hopes for control perhaps depend on thecontrol of air quality and air movement in publicplaces. Second, the viruses and the Chiamydia trans-mitted by the venereal route. Their transmission isgreatly favoured by current habits of having multiplesexual partners, generally without mechanical(contraceptive) barriers to infection. These infectiousagents also seem to be evolving rapidly and enteringa successful period with unprecedented numbers ofhosts engaged in harmless venery, suffering in the

case of Chlamydia what we must call occasionalcolds in the urethra or cervix.The third and most interesting group of success-

ful viruses are those that persist in the body for longperiods after initial infection, often for life.

Bodily sites of persistenceLet us consider the sites of the body in which

viruses persist, and see whether these sites are ofany particular significance (Table 1).

Nervous systemFrom the point of view of the infectious agent,

there must be routes of discharge to the outsideworld if fresh hosts are to be infected. Herpes sim-plex and varicella-zoster viruses persist in dorsalroot ganglia, and later in life, after reactivation inthe ganglia, they spread down peripheral nerves tothe body surface and initiate a lesion (cold sore orzoster) which is a source of infection for otherindividuals. In contrast, viruses that persist in thecentral nervous system would appear to have noopportunity to be shed to the outside world. Theirpersistence in this site can therefore be regarded asa 'dead end' from the virus point of view, except inthe rare disease kuru where transmission of thecausative agent takes place during cannibalisticconsumption of infected brain. Such infectionsinclude subacute sclerosing panencephalitis (SSPE),progressive multifocal leucoencephalopathy (PML)and visna in sheep. The significance of this site ispossibly that antiviral forces are less effectivelymobilized in the central nervous system, thusfavouring persistence. Perhaps the virus properties(non-productive infection, etc.) that are relevant topersistence elsewhere in more 'useful' sites in thebody (see below) happen also to favour persistencein the central nervous system.

Glands and body surfacesThe viruses that persist in glands and other body

surfaces discharging to the outside world are clearlyable to spread to fresh individuals. This is as truefor rabies virus in the acinar cells of the salivarygland of a vampire bat as it is for cytomegalovirus

0032-5473/78/0900-0581 $02.00 ©) 1978 The Fellowship of Postgraduate Medicine

copyright. on July 7, 2020 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.54.635.581 on 1 Septem

ber 1978. Dow

nloaded from

TABLE 1. Sites of infection of some persistent viruses

Site Virus Species Significance of site

Measles (SSPE) Man Immune forces lessrBrain JC virus (PML) Man effective?

Nervous system ) Visna virus Sheep

tGanglia Herpes simplex Man Shedding of virus toVaricella-zoster Man J exterior

rSalivary gland (or Cytomegalovirus Man, Mouse )oropharyngeal EB virus Mansite) Rabies virus Vampire bat

Glands and body Cytomegalovirus Man Shedding of virus tosurfaces Kidney tubules BK, JC virus? Man exteriorPolyoma virus Mouse etro

Epidermis Wart virus Man

Feather follicle Marek's disease virus Chicken

Ovum Leukaemia, mammary tumour, Mouse Mechanism for verticalGerm cells LCM viruses transmifon

LSperm Mammary tumour virus Mouse transmission

CGrowth liver Hepatitis B Man

Blood t ? growth site Simian haemorrhagic fever Patas monkey Mosquito transmission?Blood ? growth site virus t

LGrowth macrophages Equine infectious anaemia Horse 9

Lymphoreticular tissue Many examples Many species Mechanism for weakeningantiviral forces

in the tubular epithelial cells in the kidney of an

infected child. Warts often persist for many months,and large amounts of virus are present in the skinscales that are shed from the surface of the wart.Glands and surface epithelia are often infected, ofcourse, by non-persistent 'hit and run' viruses, butthe histological structure of these tissues favourspersistence (Fig. 1). Circulating antibody fails toreach the lumenal surface of infected cells wherevirus is being released, and IgA antibodies, whilethey can attach to the virus surface and limit infec-tion of other cells, are not capable of killing infectedcells. Sensitized lymphocytes cannot attack theinfected cell surface because in the absence of localinflammation or cell damage, they are unable totraverse the basement membrane and the epithelialcell sheet.

Germ cellsThe viruses that persist as integrated genomes in

germ cells, the viral genome being replicated asthe cell divides, are in a very powerful position. Allcells arising from the fertilized ovum will be infected,and thus all descendant generations, giving a type ofvertical transmission in which it may be difficult tosay which genes are viral and which are host. Thisis an ideal state of parasitism, it being of no conse-

quence to the viral parasite whether it is present as afully infectious entity or a nucleic acid remnant.If there is an 'Adam and Eve' origin for the species,then all individuals in the species will be infectedab initio but if not, an initial period of horizontaltransmission between individuals would be neces-sary. As argued elsewhere (Mims, 1975) it is con-ceivable that during the course of evolution manyretroviruses have trod this path to integration intothe host genome, because there is no selection againstdefectiveness and the presence of mere fragments ofthe viral genome as long as most individuals areinfected.

BloodIt might seem a more difficult problem for an

infecting virus to persist in the blood, there beingpowerful mechanisms for the removal of circulatingforeign particles, but there are a number of examples.In many of these it seems likely that the infectionhas induced the formation of non-neutralizing orlow avidity antibodies which favour persistence.One possible significance of this site is that it repre-sents a long term source of infection for bitingarthropods, who could thus continue to transmitthe infection to fresh individuals. The persistentviraemia established in the Patas monkey by simian

C. A. Mims582

copyright. on July 7, 2020 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.54.635.581 on 1 Septem

ber 1978. Dow

nloaded from

Introduction to persistent infections

secretory or eKeretorY glandMam6ary gland

tub

Feater folileleBil duct

i. . -:

Ifected cellsFIG. 1. Viral infection of cell surfaces facing the exterior.

TABLE 2. Persistence of viruses in lymphoreticular tissues

Lymphoreticularsite Virus Host

Lymphocytes Adenoviruses ManEB virus ManCytomegalovirus Man etc.Leukaemia viruses Mouse etc.Visna virus SheepAfrican swine fever virus PigLCM virus Mouse

Macrophages Lactic dehydrogenasevirus Mouse

Aleutian disease virus MinkEquine infectious anaemia

virus HorseAfrican swine fever virus PigLCM virus Mouse

haemorrhagic fever virus could have this significance,as would a viraemia that reappeared long afterinitial infection by any arbovirus, following re-activation in the host. Both hepatitis B (man) and

equine infectious anaemia (a retrovirus infection ofhorses) are known to persist in the blood of the hostfor as long as 20 years. There have been suggestionsthat equine infectious anaemia is mechanicallytransmitted by biting arthropods. Could hepatitisB have arisen in this way, with spread via needlesand scratches playing a secondary role in developedcountries where there is lower exposure to bitingarthropods? On the other hand, two other virusesthat persist in the blood, lactic dehydrogenase virus(mice) and Aleutian disease virus (mink) are notknown to be transmitted by arthropods, so perhapsthe search for significance can at times be un-rewarding.

Lymphoreticular tissueThe numerous persistent viruses that infect lym-

phoid tissue or macrophages are listed in Table 2.Scrapie, although it is not known to be a virus, couldbe added to this list. These viruses might seem atfirst sight to derive no benefit from persisting in

583

copyright. on July 7, 2020 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.54.635.581 on 1 Septem

ber 1978. Dow

nloaded from

C. A. Mims

lymphoreticular tissue. There is generally no routeto the outside world, except in arbovirus infectionswhen virus reaching the blood is ingested by thetransmitting arthropod, so can this site of infectionbe regarded as a 'dead end' like the nervous system?Perhaps an explanation should be sought in theeffect of growth in lymphoreticular tissue on immuneresponses to the infecting virus. Viruses, like othermicroorganisms, are commonly taken up by macro-phages and many are then inactivated. Macro-phages prepare antigenic components for deliveryin small amounts to lymphocytes, and in doing soexercise a controlling function in the induction ofthe immune response. Strict controls over the move-ment and handling of antigens are probably vitalin the quiet backwaters of lymphoid tissue wheremacrophages and lymphoid cells come closely to-gether and generate the immune response. But wheninfectious virus reaches these quiet backwaters andmultiplies in macrophages or lymphocytes theorderly and well controlled process is almost certainto be disturbed. Large amounts of viral antigenwould be liberated in the midst of viable immuno-competent cells which may themselves be infected,and serious defects in the resulting responses are tobe expected, which may permit the establishmentof the persistent infection. In other words, to evadethe immune forces the virus invades the sites wherethese forces are generated and controlled. Themechanisms are not clear, but possibilities wouldinclude specific T cell deletions (the growth of virusin clonally expanding cells resulting in abnormalfunction in these cells), abnormal macrophagehandling of virus antigens, effects via suppressorcells, and so on. Certainly immunosuppression is acommon phenomenon in persistent virus infections,suggesting interference with immune function, andin some instances, such as EB and cytomegalovirusinfection in man, there are auto-immune phenomenasuggesting a degree of 'immunological chaos'.

Significance of transformation by virusesVery little is understood about persistence of

infection in the body at the virus-cell level. Quiteprobably adenoviruses, for instance, differ in thisway from herpes simplex, and this in turn from otherviruses of the herpes group. One very importantmechanism by which a virus can evade antiviralforces and become persistent is by ceasing to pro-duce virus particles and remaining in the host cellin the form of viral nucleic acid. Many persistentviruses show a striking reluctance to come out of thecell. The leukaemia viruses for instance possess areverse transcriptase that enables them to form aDNA replica of their RNA genome, -which thenbecomes part of the host cell genome. This is acentral mechanism for the persistence of these and

of other retroviruses such as visna and equineinfectious anaemia virus.The papovaviruses, which include SV40, polyoma,

wart virus and the recently discovered humanpolyoma-like viruses (JC and BK viruses) also havethis ability to integrate into host cell DNA. Quiteearly in the development of modern virology it wasfound that SV40 and polyoma viruses, which arenatural infections of monkeys and mice respectively,caused transformation of cells in culture. VirusDNA was incorporated into host cell DNA, andthe cells developed stable alterations in their proper-ties, behaving in many ways like tumour cells. Theapparent relevance of this phenomenon for theunderstanding of cancer led to a great surge inresearch, mostly at the virus-cell level. These aresmall DNA viruses whose genome codes for nomore than 7 or 8 polypeptides, and in recent yearssophisticated molecular virological and geneticstudies have been made. Indeed, work on polyomahad advanced to the stage at which a partial mappingof the viral genome is possible. It is now clear,however, that these viruses rarely cause tumoursunder natural circumstances in the host species,and the ability to transform cells and cause cancer isthus irrelevant from the virus's point of view. Yetthe striking facility with which papovavirus DNAbecomes stably incorporated into the host cellgenome is likely to have some significance. In fact,it would enable these viruses to persist in host cellsand enjoy protection from immune and other anti-viral forces. Persistence in the host, if accompaniedby the likelihood of reactivation later in life withthe reshedding of virus to the exterior, would be ofimmense biological significance. Could it be thatthis is the 'function' of papovavirus integration intohost cell DNA, just as it is for leukaemia virus andother retroviruses? Adenoviruses also can integrateinto host cell DNA, and transform cells, yet withoutcausing tumours under natural circumstances. Here,tob, the function of integration may be to facilitatepersistence.

Thus, polyoma virus persists for life in the infectedmouse, and with the aid of the factors illustrated inFig. 1, is shed from saliva and urine for long periods.It would be interesting to see whether polyoma virusreactivates, as shown by an increase in sheddingduring pregnancy, or in old age. BK and JC virusespersist in the human host after primary infectionand can later be reactivated and excreted in infectiousform in urine in the immunosuppressed renal allo-graft recipient. Epithelial cells in the kidney tubulesor ureter are likely sites for persistence (Coleman.1975) and the presence of IgM antibodies to BKvirus in many normal people (Flower, Banatvalaand Christie, 1977) suggests the continued release ofviral antigens. To take another example, human

584

copyright. on July 7, 2020 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.54.635.581 on 1 Septem

ber 1978. Dow

nloaded from

Introduction to persistent infections 585

wart virus also is reactivated in immunosuppressedrenal allograft recipients, leading to the appearanceof multiple warts. It seems possible that after initialinfection, which often appears to be subclinical(Lee and Eisinger, 1976), the virus persists, possiblyas integrated DNA, in stem cells in the epidermis.In the wart itself it is known that viral DNA ispresent in epidermal cells, but antigens and infec-tious virus are only formed as the cells keratinizeand approach the skin surface.

Importance of persistent virus infectionsWhy are persistent virus infections important?

(Table 3).

TABLE 3. Importance of persistent virus infections

1. Reactivation Renal transplant patientsHodgkin's diseaseOld age

2. Immunopathological Circulating immune complexesdamage Most good examples in animal

models3. Association with Leukaemia viruses

tumours Burkitt's lymphomaCarcinoma of cervix

4. Persisence enables Example of varicella-zosterparasite to be Prehistoric mansuccessful Isolated communities

1. They can reactivate, as shown for immuno-suppressed renal transplant recipients in Table 4.They also reactivate in certain diseases characterizedby immunosuppression, as illustrated by the zostereruptions seen in patients with Hodgkin's disease.A weakening of immune defences is also a feature ofold age, whether in man or mouse, and the frequencyof zoster rises sharply in patients over the age of50 years. It would be of very great interest to deter-mine whether the other persistent virus infections,for instance BK, JC, adenovirus or hepatitis B,reactivate in old age.

2. Persistent viruses can cause progressive disease.This is often immuno-pathological in nature, suchas the glomerulonephritis and vasculitis seen inlymphocytic choriomeningitis infection in mice andAleutian disease virus infection in mink. In thesepersistent infections circulating immune complexesare formed in antigen excess and are responsiblefor the lesions.

3. Persistent viruses can also cause tumours.Certain persistent herpes viruses can either causetumours experimentally or are strongly implicatedin naturally occurring tumours. The leukaemia andmammary tumour viruses cause tumours in certainspecies, depending on host genetics and hormonalstatus.

4. Persistence is important from the virus's pointof view, enabling it to be a successful parasite. Theexample of varicella-zoster makes this clear. Duringchildhood infection with chickenpox, virus ascendsto the dorsal root ganglia of sensory nerves supply-ing the affected skin areas and stays there in a non-infectious state after recovery and the elimination ofvirus from the rest of the body. The disease,chickenpox, disappears from the community. At alater stage, when infected individuals are old and theimmune responses that control the persistent in-fection haveweakened, the virus reactivates randomlyin one of the ganglia and spreads down the peripheralnerve to the skin, causing a crop of infectious vesicles.The fresh set of susceptible individuals who havearisen in the community can now be infected withchickenpox. Chickenpox thus maintains itself in smallcommunities without reintroduction from the out-side. In the palaeolithic period about 100 000 yearsago, and during most of human evolution, men havelived in small groups of 20-50 indviduals, contactwith other groups being infrequent. A virus such asvaricella-zoster can maintain itself in such a com-munity, whereas a virus such as measles or polio-myelitis could not do so. Hence persistent virusesare probably old associates of man. They are well

TABLE 4. Some common persistent virus infections of man

Virus group Virus Reactivation in immunosuppressedrenal transplant patients

Herpes viruses Herpes simplex +Varicella-zoster +Cytomegalovirus + +EB virus + Armstrong et al., 1976;

Spencer and AndersonAdenoviruses Adenoviruses + ? 1970; Krech et al.,

(Types 1, 3, 5?) 1975.

Papovaviruses BK, JC +Wart? + J

? Hepatitis B +?9 Nagington, Cossart andf Cohen, 1977.

copyright. on July 7, 2020 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.54.635.581 on 1 Septem

ber 1978. Dow

nloaded from

586 C. A. Mims

adapted parasites capable of maintaining themselvesin small groups of individuals. During an expeditioninto the Amazon basin a number of isolated Indiantribes were discovered, many of whom had littleor no contact with the outside world. Serum frompeople in these tribes was found to have antibodiesto persistent viruses such as herpes simplex, varicella-zoster, EB, hepatitis B and cytomegalovirus, butthere were no antibodies to non-persistent 'hit andrun' viruses such as poliomyelitis, measles andinfluenza viruses (Black et al., 1974).

Finally, it is clear that when an infectious agentmultiplies in the host, induces immune responses,and these immune responses fail to disinfect thetissues, then some trouble is to be expected with thedevelopment of vaccines. One of the successfulvaccines against viruses of the herpes group is thatfor Marek's disease, which is a living attenuatedvirus vaccine. It causes a persistent infection in thechicken host and thus protects against infection withthe wild type virus. When the vaccine virus is liveand persists in tissues, there are worries about re-activation or tumour induction later in life. Perhapsthe live vaccine virus manages to persist in the bodybecause, like its 'wild' counterpart, it invadeslymphoreticular tissues as discussed above. If thisis so, the future lies in killed virus vaccines, and thesemay come when we are able to induce the right typeof immune response to the right virus antigen. Untilthen, man will keep his persistent viruses.

ReferencesARMSTRONG, J.A., EVANS, A.S., RAO, N. & Ho, M. (1976)

Virus infections in renal transplant patients. Infection andImmunity, 14, 970.

BLACK, F.L., HIERHOLZER, W.J., PINHEIRO, F., EVANS, A.S.,WOODALL, J.P., OPTON, E.M., EMMONS, J.E., WEST, B.S.,EDSALL, G., DOWNS, W.G. & WALLACE, G.D. (1974)Evidence for persistence of infectious agents in isolatedhuman populations. American Journal of Epidemiology,100, 230.

COLEMAN, D.J. (1975) The cytodiagnosis of human polyomavirus infection. Acta cytologica, 19, 93.

FLOWER, A.J., BANATVALA, J.E. & CHRISTIE, I.L. (1977) BKantibody and virus-specific IgM responses in renal trans-plant recipients, patients with malignant disease, andhealthy people. British Medical Journal, 2, 220.

KRECH, U., JUNG, M., PRICE, P.C., THIEL, G., SEGE, D. &REUTTER, F. (1975) Virus infections in renal transplantrecipients. Zeitschrift fur Immunitatsforschung, Allergieund klinische Immunologie, 148, 341.

LEE, A.K.Y. & EISINGER, M. (1976) Cell mediated immunity(CMI) to human wart virus and wart-associated tissueantigens. Clinical and Experimental Immunology, 26,419.

MIMS, C.A. (1975) The meaning of persistent infections innature. Bulletin of the World Health Organization, No. 52,747.

NAGINGTON, J., COSSART, Y.E. & COHEN, B.J. (1977) Re-activation of hepatitis B after transplant operations.Lancet, i, 558.

SPENCER, E.S. & ANDERSON, K.H. (1970) Clinically evident,non-terminal infections with herpes viruses and the wartvirus in immunosuppressed renal allograft recipients.British Medical Journal, 3, 251.

copyright. on July 7, 2020 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.54.635.581 on 1 Septem

ber 1978. Dow

nloaded from