the pathogenesis of respiratory viral · thepathogenesisofrespiratory viral infection dramaticsince...

Postgraduate Medical Journal (February 1979) 55, 122-127.

The pathogenesis of respiratory viral infection

R. B. HEATHM.D., F.R.C.Path.

The Joint Department of Virology of the Medical Colleges of St Bartholomew's and The London Hospitals

SummarySendai virus is used as a model to study the histologicalresponse and the mechanism of recovery of mice frominfection. Although mice infected intranasally do notappear ill, the ciliated columnar cells of the bronchialepithelium undergoes rapid necrosis followed laterby repair. Interferon is produced at an early stage ofinfection and antibody formation follows later.Cyclophosphamide given to mice infected with Sendaivirus abolishes the humoral antibody response but notinterferon production; the mice develop pneumoniclesions. They cannot eliminate virus and they die.On the other hand, deprivation of 'T' lymphocytes bythymectomy, irradiation and bone marrow recon-stitution renders mice more susceptible to Sendaivirus but they do not die. The humoral response is thusapparently helped by 'T' lymphocytes and interferonproduction is not vital for recovery.

IntroductionSendai virus is a common murine respiratory

pathogen. It is endemic in many colonies of labora-tory mice including those of certain highly reputablecommercial breeders. It is an interesting virus,because if it is inoculated intranasally into suscept-ible mice there is no overt manifestation of clinicalillness. However, histological and virological examin-ation of inoculated animals reveals that they haveexperienced an upper respiratory tract infectionfrom which they recover itormally. Sendai virusinfection of mice is thus an excellent model of acuterespiratory viral infection of man. It is quite differentfrom the adapted influenza infection of mice whichinduces viral pneumonic illness: a pathology whichmercifully is rarely seen in man.

In this paper, some of the studies of Sendai virusinfection of mice which have led to and are leadingto a better understanding of the pathogenesis ofacute respiratory viral infection will be brieflyreviewed.

The pathology of the infectionAs mentioned above, inoculation of mice with

Sendai virus does not produce overt evidence ofdisease. However, the fact that these inoculatedmice are ill can be deduced from the finding that they

gain weight at a significantly slower rate than domock inoculated control mice. Histological examin-ation of lungs removed at intervals after inoculationreveals the characteristic features of acute respira-tory viral infection (Robinson, Cureton and Heath,1968). Within 24 hr of inoculation, early damage topatches of the airway mucosa is apparent. Immuno-fluorescence staining for viral antigens shows thatthis early damage is associated with viral replicationwhich is confined to the mucosa; there is no involve-ment of alveolar cells. By the third day of theinfection the involved areas of the mucosa arecompletely destroyed and this is associated withintense inflammatory changes in the underlyingmucosa (Fig. lb). The damaging effect of the virusdeclines towards the end of the first week of theinfection. By the beginning of the second week itis apparent that the infection is over, for at thattime repair of damaged areas of mucosa can beseen. The submucosa underlying these areas is nowcongested with mononuclear cells of the lymphoidseries (Fig. lc). All of the damaged mucosa iscompletely repaired by the end of the second week(Fig. Id). The pathological changes of the airwaysthat are seen in the model system are essentiallysimilar to those caused by influenza in mice(Andrewes, Laidlaw and Smith, 1934), ferrets(Francis and Stuart-Harris, 1938) and man (Hers,Masurel and Mulder, 1958).The remarkable feature of these changes is not

the infectious process itself but the fact that it is soabruptly stopped soon after initiation and, inparticular, is prevented from extending to involvethe whole of the susceptible mucosa. This clearlyindicates that respiratory viral infection inducespowerful local defence responses, which rapidlybring about its termination. Investigations relatedto evaluation of these defences will now be brieflydescribed.

Defence against the primary infectionVirus growth and both interferon and antibody

production occurring during the course of theinfection can be determined by assay of serum andsuspensions of lungs taken at intervals after inocu-lation (Robinson et al., 1968). The data obtained

0032-5473/79/0200-0122 $02.00 ©1979 The Fellowship of Postgraduate Medicine

copyright. on July 21, 2021 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.55.640.122 on 1 February 1979. D

ownloaded from

http://pmj.bmj.com/


U.C>cn

oI

mt

oo

x v

<

EIe0

0

cd

co0Cu

_cC._

.U

·~i~

._

\^»> -

--'I w) °Pre8-

;SR s

f-

- >

r··

s-

0

I o

U)

.

t

o

CiS0s

Q C-

0

Cd.Cauco

'Y C

123

:-

I,



j.bmj.com

/P


j.55.640.122 on 1 February 1979. D

ownloaded from

http://pmj.bmj.com/

R. B. Heath

can be correlated with the histological findingsdescribed above. The findings are summarized inFig. 2. Virus growth in the lungs is found to increaserapidly, reaching a peak level on the third to fourthday of the infection. Then, perhaps not surprisingly,virus titres decline rapidly and virus can no longerbe detected after the eighth day.

FIG. 2. Diagrammatic summary of virus growth, inter-feron production, serum antibody levels, inflammatorychanges and mononuclear cell infiltration of the sub-mucosa occurring during the course of Sendai virusinfection of mice. -= Virus titres. ---- =Inter-feron titres ......= Antibody titres. (Reproduced bypermission of the Journal of Medical Microbiology).

There are a number of host defence factors whichcan play some part in virus eradication that ulti-mately leads to the termination of the infection.Sendai virus is a good interferon inducer and it canbe seen that this antiviral substance together withthe non-specific inflammatory changes appear veryearly in the infection. Both those factors fade atapproximately the same time, which is soon after theviral eradication process has commenced. Incontrast, specific antibody cannot be detected ineither serum or lung suspensions until the eighthday of the infection; that is at a time when the viraleradication is nearly completed. The late appearanceof detectable antibody should not be interpreted asindicative of a delayed immune response. The reasonfor this is that it can be shown that a considerableincrease in the number of antibody containing Bcells has occurred in the submucosa by the second

day (Blandford, Cureton and Heath, 1971) andthere is indirect evidence that these cells are produc-ing specific anti-Sendai antibody by the third dayof the infection (Blandford and Heath, 1972).Although it cannot be said that there is a syn-chronous appearance of the various host responsesto the infection, it is apparent that they are allinduced at roughly the same time early in theinfection.

The relative importance of the various host defencesAlthough it is probable that both the non-specific

defences (interferon and inflammation) and thespecific immune response all play some part interminating these infections, it was of interest toattempt to find out which had the predominant role.Pertinent information was obtained by treatingSendai-infected mice with the immunosuppressantdrug cyclophosphamide (Robinson, Cureton andHeath, 1969). The effect of this treatment was

o

00

4)

c

3

tn

c

-J

c

ta

4,

*00

c.0aEe4C,

Day of infection

FIG. 3. Effect of cyclophosphamide on virus growth,interferon and serum antibody production in Sendai-infected mice. = Untreated. - -- = 400 mg/kg..... =200mg/kg. -. -= 100 mg/kg. -- = 50 mg/kg.(Reproduced by permission of the Journal of MedicalMicrobiology).

124



j.bmj.com

/P


j.55.640.122 on 1 February 1979. D

ownloaded from

http://pmj.bmj.com/


dramatic since it converted what is normally a mildinapparent infection into a fatal pneumonic illness.The sequence of events occurring in this alteredinfection is summarized in Fig. 3. The most obviousdifference between the disease process in normal andcyclophosphamide-treated mice is the inability of theimmunosuppressed animals to eradicate virus fromtheir lungs. The apparent cause of this is the absenceof the immune response, which manifests as non-appearance of serum antibody, and also by thehistological finding that mononuclear cells do notappear in underlying areas of affected mucosa. Incontrast to this, the non-specific defence mechan-isms show no diminution but are exaggerated.Interferon production, for example, continuesuntil the death of the animal and inflammatorychanges are not only more intense in the submucosabut have now extended to the alveoli. These exagger-ated non-specific defence mechanisms on their ownare obviously quite incapable of eradicating thevirus.

The humoral immune responseThe appearance of large numbers of immuno-

globulin-containing lymphocytes in the submucosaon the second day of the infection (Blandford et al.,1971), the demonstration that these cells are produc-ing specific antibody by the third day (Blandfordand Heath, 1972) and the disastrous results that

o

0

0'OP

1-3

1.0

0*5

0.1

Spleen

O- Lungs

/_//__,1227 34 49 70//0 2 4 6 8 12 27 34 49 70

Days after infection

FIG. 4. Ratios of IgG- and IgA-containing cells in thespleen and respiratory submucosa of mice at variousintervals after infection with Sendai virus. (Reproducedby permission of Immunology).

ensue if this submucosal response is prevented bycyclophosphamide, all suggest that B lymphocytesare of prime importance in terminating the infection.A study of the various immunoglobulin classes

produced by these B cells was of interest (Blandfordand Heath, 1974) and the results are summarizedin Fig. 4. Before inoculation, the few IgA- and IgG-containing cells in the submucosa were present in theexpected ratio of 5 : 1. However, at the height of theinfection, when there was an appreciable increase inthe numbers of both kinds of cell, the ratio wasapproximately 1:1. IgA-containing cells did notagain appear in excess until the thirtieth day afterinoculation. These findings clearly show that thedominant humoral response to this damaging viralinfection is the local induction of IgG rather thanIgA class antibody.

The cellular immune responseNot all of the mononuclear cells accumulating

in the submucosa could be shown by immuno-fluorescence to have Ig surface markers and it isnot unreasonable to assume that some of these areT cells. Indeed it can be shown by 51Cr releasemethods that T cells sensitized to Sendai antigenscan be detected in the spleen early in the infection,with their activity reaching a peak on the sixth day.It can reasonably be presumed that cells with similaractivity appear in the respiratory mucosa. The cyto-toxic activity of normal and sensitized spleen cellsagainst Sendai-infected target cells is a complexphenomenon (Anderson et al., 1977) which will notbe detailed here. It was, however, intriguing to findout whether these sensitized T cells played any rolein the primary defence against the Sendai virus. It is,for example, well established that T cell-mediatedimmunity is essential for efficient recovery frominfection by poxviruses and herpesviruses but thistype of immunity has been shown to be of noimportance in recovery from enteroviral infection.The role of cell-mediated immunity (CMI) inrecovery from respiratory myxoviral infection is notclearly understood. Suzuki, Ohya and Ishida (1974)have shown that administration of either anti-lymphocyte serum or antilymphocyte globulinreduces the mortality of mice infected with apneumonic strains of influenza. The findings suggestthat T cells have an immunopathological effecton the lung infection. Virelizier (1975) has shownthat thymus-deprived mice show increased sus-ceptibility to infection with PR8 virus and he hasattributed this to the absence of the helper effect ofT cells on antibody production.

It was clearly of interest to observe the effect ofT cell deprivation on the Sendai infection which isrestricted to the mucosa. Accordingly, mice werethymectomized, irradiated and reconstituted with

125copyright.

on July 21, 2021 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.55.640.122 on 1 F

ebruary 1979. Dow

nloaded from

http://pmj.bmj.com/

R. B. Heath

bone marrow cells (TXBM) and subsequentlyinfected with Sendai virus. The results of this experi-ment (Anderson and Heath, to be published) werebriefly as follows: The thymus-deprived micewere found to be more susceptible to the infectionsince they developed pneumonitis. The immuno-suppressant effect was not nearly as marked as thatproduced by cyclophosphamide, as none of themice died and none became particularly ill. Itis not clear why the defences were defective in theseTXBM mice. The serum antibody response wasnormal and large numbers of plasma cells appearedin the submucosa. Thus, although the T cell-mediatedresponse is not essential for recovery, it neverthelesshas a role in defence.

The macrophageBacterial pneumonia is a well known complication

of viral infection of the upper respiratory tract.Possible reasons for this are disruption of the'mucociliary escalator' due to damage of the mucosalcells and the inflammatory oedema induced by thisdamage. Another possibility is that virus infectionimpedes the antibacterial activity of alveolarmacrophages.

It has been shown by immunofluorescence thatmouse peritoneal and alveolar macrophages canbe infected by Sendai virus but infectious virus isnot released from these cells (Mims and Murphy,1973). It has also been shown that Sendai-infectedmice are much less efficient than normal mice inclearing deliberately inoculated Haemophilusinfluenzae organisms from their lungs (Degr6 andGlasgow, 1968). This impaired clearance wasassociated with the development of pneumoniawhich could not be induced by either virus orbacterium when each was inoculated alone.

Further studies with the Sendai model should becapable of more precisely defining the defect inanti-bacterial defences occurring in upper respiratorytract infection.

DiscussionSendai virus infection of mice is a useful model of

acute respiratory viral infection of man since thedisease process is restricted to the lining of theairway. It is thus more satisfactory than the morecommonly used influenza infections which inducepneumonic illnesses. Fundamental concepts of thenature of acute respiratory viral infection can begained by use of this Sendai model. The dataobtained are probably applicable to all respiratoryviral infections. Indeed, virtually identical resultshave been obtained with the Kunz strain of influenzavirus, which grows well in the respiratory tract of

mice without producing pneumonic changes (Rautet al., 1975).

In terms of pathogenesis, the early stages of theinfection are unremarkable. Infected mucosal cellsare destroyed in a manner similar to that seen intissue culture systems infected by cytopathicviruses. What is remarkable is the fact that thisinfective process is rapidly stopped and the damagedmucosa repaired. The efficient defence processeswhich bring about the termination of the infectionare of considerable importance. This is because anunremitting destructive infection of the respiratorymucosa would so expose the alveoli to both patho-genic and commensal bacteria, that a fatal outcometo the disease would be probable.The effective defence of the respiratory system

against viral infection is affected by a number offactors including the immune response, macro-phages, interferon and inflammatory changes. Itappears that the immune response is the mostimportant of these, particularly its humoral arm.It is certain that lymphoid cells of the B series, whichproduce antibodies in the immediate vicinity ofaffected areas of mucosa, have the predominantrole in terminating the infection. It is also possiblethat the B cells that produce IgG class antibodiesare more important that those that produce IgA.In the convalescent period, however, re-infection isprevented by secretory IgA in the bronchial mucus.Lymphoid cells of the T series become sensitized

to Sendai antigen during the course of the infectionThese cells do not seem to have an important rolein defence against the infection but they are nottotally unimportant as they appear to have a 'helper'effect on the humoral response.The Sendai model can be used to evaluate the

role of macrophages in acute respiratory viralinfection. Evidence has been obtained which suggeststhat Sendai virus can impair the antibacterialactivity of alveolar macrophages thus increasing theprobability of the development of secondarybacterial pneumonia.The crucial event in these common infections must

be the appearance of mononuclear cells of thelymphoid series in the respiratory submucosa.If their rate of accumulation at this site could beenhanced then the severity of the infection would bebound to be reduced. This would be particularly trueof symptomatic secondary viral infections. Immuno-potentiation might thus be an additional aim offuture respiratory viral therapy and prophylaxis.

AcknowledgmentsThe work referred to in this paper was undertaken by a

number of colleagues including Drs T. W. E. Robinson,R. J. Cureton, G. Blandford, S. Raut, Mrs Hurd and MrsAnderson.

126



j.bmj.com

/P


j.55.640.122 on 1 February 1979. D

ownloaded from

http://pmj.bmj.com/

The pathogenesis of respiratory viral infection 127

ReferencesANDERSON, M.J., BAINBRIDGE, D.R., PATTISON, J.R. &HEATH, R.B. (1977) Cell mediated immunity to influenzainfection in mice. Infection and Immunity, 15, 239.

ANDREWES, C.H., LAIDLAW, P.P. & SMITH, W. (1934) Thesusceptibility of mice to the viruses of humans and swineinfluenza. Lancet, ii, 859.

BLANDFORD, G., CURETON, R.J.R. & HEATH, R.B. (1971)Studies of the immune response in Sendai virus infectionof mice. Journal of Medical Microbiology, 4, 351.

BLANDFORD, G. & HEATH, R.B. (1972) Studies on the immuneresponse and pathogenesis of Sendai virus infection ofmice. I. The fate of viral antigens. Immunology, 22, 637.

BLANDFORD, G. & HEATH, R.B. (1974) Studies on the immuneresponse and pathogenesis of Sendai virus infection ofmice. II. The immunoglobulin class of plasma cells in thebronchial sub-mucosa. Immunology, 26, 667.

DEGRE, M. & GLASGOW, L.O. (1968) Synergistic effect inviral-bacterial infection. I. Combined infection of therespiratory tract in mice with parainfluenza and hemo-philus influenza. Journal of Infectious Diseases, 118, 449.

FRANCIS JR, T. & STUART-HARRIS, C.H. (1938) Studies on thenasal histology of epidemic influenza virus infection in theferret. Journal of Experimental Medicine, 68, 789.

HERS, J.F.Ph., MASUREL, N. & MULDER, J. (1958) Bacterio-logy and histopathology of the respiratory tract and lungsin fatal Asian influenza. Lancet, ii, 1141.

MIMS, C.A. & MURPHY, F.A. (1973) Parainfluenza virusSendai infection in macrophages, ependyma, choroidplexus, vascular endothelium and respiratory tract of mice.American Journal of Pathology, 70, 315.

RAUT, S., HURD, J., CURETON, R.J., BLANDFORD, G. &HEATH, R.B. (1975) The pathogenesis of infections of themouse caused by virulent and avirulent variants of an in-fluenza virus. Journal of Medical Microbiology, 8, 127.

ROBINSON, T.W.E., CURETON, R.J.R. & HEATH, R.B. (1968)Pathogenesis of Sendai virus infection in the mouse lung.Journal of Medical Microbiology, 1, 89.

ROBINSON, T.W.E., CURETON, R.J.R. & HEATH, R.B. (1969)The effect of cyclophosphamide on Sendai virus infectionon mice. Journal of Medical Microbiology, 2, 137.

SUZUKI, F., OHYA, J. & ISHIDA, N. (1974) Effect of antilym-phocyte serum on influenza virus infection in mice.Proceedings of the Society for Experimental Biology andMedicine, 146, 78.

VIRELIZIER, J.-L. (1975) Host defences against influenzavirus: the role of anti-haemagglutinin antibody. TheJournal of Immunology, 115, 434.



j.bmj.com

/P


j.55.640.122 on 1 February 1979. D

ownloaded from

http://pmj.bmj.com/

the pathogenesis of respiratory viral · thepathogenesisofrespiratory viral infection dramaticsince...

Documents