Some aspects of immunity and disease during influenza A virus infection


Material Information

Some aspects of immunity and disease during influenza A virus infection
Physical Description:
v, 60 leaves : ill. ; 29 cm.
Cogliano, Robert Christopher, 1946-
Publication Date:


Subjects / Keywords:
Influenza A virus -- immunology   ( mesh )
Influenza A virus -- pathogenicity   ( mesh )
Immunology and Medical Microbiology Thesis Ph.D   ( mesh )
Dissertations, Academic -- immunology and medical microbiology -- UF   ( mesh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 1979.
Bibliography: leaves 56-59.
Statement of Responsibility:
by Robert Christopher Cogliano.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000891228
oclc - 25097453
notis - AEJ9676
sobekcm - AA00004909_00001
System ID:

Full Text








I would like to thank Dr. Parker A. Small, Jr., for his support

and patience. Scientifically, his imaginative, creative and critical

thinking were always a standard for which to strive. His nature to

take time to care and try to offer advice on personal matters was much


My thanks also to Bob Yetter, a friend when needed, a source of

impeccable common sense when none was to be found elsewhere and one

of the best trivia experts around.






S ii

. iv

The Virus .
The Disease .
Immunity to Influenza .


Viruses . .
Intranasal Inoculation of V
Virus Adaptation in Mice. .
Assays . .
Tissue Cultures .
Statistical Analysis .
Serum Antibody Production .


Antiimmunoglobulin Productio
Interferon Assay .
Neutralization Titers .
X-rays . .
Scanning Electron Microscopy

rus into Animals.

n .

. .o ., .
. .
. ,. ., ., .


Introduction . . .
Results . . .
Mechanisms of Immunity in Ferret Organ Cultures
Bladder Immunity . .
Ciliary Activity in Tracheal Organ Cultures .
Discussion . . .








Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy



Robert Christopher Cogliano

December 1979

Chairman: Parker A. Small, Jr.
Major Department: Medical Sciences (Immunology and Medical Microbiology)

In order to study immunity to influenza virus infection in the

upper respiratory tract we have tested organ cultures from normal and

convalescent ferrets for resistance to infection. We then attempted to

delineate immune mechanisms acting within the tissue.

Ferret tracheal organ cultures prepared from animals previously

infected intranasally with influenza A virus required approximately 130

times more homologous virus (A/PR/8/34(HON1) or A/Port Chalmers/l/73

(H3N2)) to become infected in vitro than similar cultures from normal

ferrets. Also, these cultures from convalescent ferrets required about

10 times more heterologous virus (A/PR/8/34 (HON1) or Sendai) to become

infected in vitro than similar cultures from normal animals challenged

in vitro with the heterologous virus. We conclude that these tracheal

rings are specifically immune.

Possible mechanisms of resistance of the tracheal organ cultures

were examined. Serum antibody given intraperitoneally to normal ferrets

did not bring about immunity in cultures derived from these animals.

Rabbit antiferret gamma chain did not significantly ablate resistance

of cultures when put on cultures prior to virus challenge. Rabbit anti-

ferret alpha chain treatment gave ambiguous results. Neither neutraliz-

ing antibody nor interferon could be measured in Day 4 culture super-


Once tracheal rings are infected, they continue to shed viruses for

at least 60 days, the longest period any cultures were kept. Virus shed-

ding in the intact ferret lasts normally 5-7 days. Thus, recovery in the

intact ferret seems to be dependent upon factors which are not present,

or at least not functional, in the tracheal explant. This is consistent

with the hypothesis that recovery is dependent upon systemic rather than

local phenomena.

Bladder tissue from normal and previously infected ferrets was

also cultured and challenged with homologous and heterologous virus. The

bladder from previously infected ferrets exhibited specific immunity,

although the immunity was more variable.

Concurrent studies aimed at determining the therapeutic value of

serum antibody for the treatment of lethal viral pneumonia in mice showed

that serum antibody given before infection or on Days 3, 4 and 5 after

infection was life-saving but given later was ineffective in preventing

death. Since radiological change could first be detected 6-7 days after

infection we conclude that serum antibody is not likely to be useful for

treatment of animals with lethal viral pneumonia.


This dissertation deals primarily with aspects of immunity to

influenza virus and secondarily with pathogenesis and recovery from

influenza infection. The introduction begins with a discussion of the

virus, specifically its structure, replication and antigenic nature.

This is followed by a description of the disease caused by the virus and

then a summary of our knowledge of immunity. Lastly, we report the

specific questions which we addressed in this research.

The Virus


The influenza virus is taxonomically designated as belonging to

the Orthomyxoviridae family (Melnick, 1977). Structurally, influenza

virus contains a segmented RNA genome which is associated in the nucleo-

capsid with an RNA transcriptase consisting of 3 proteins and a nucleo-

protein. Exterior to the genome is the virus coded matrix protein and

an envelope which is derived from the host cell during budding. Imbedded

in the envelope of the virus are two types of virus specific glycoprotein

spikes, the hemagglutinin (HA) and the neuraminidase (NA). An eighth

protein is the nonstructural protein (NS) for which no definite role has

been established.


The RNA genome of influenza A contains 8 segments (Ritchey' et al.

1976) and the virus replicates as a class V virus according to the Baltimore

classification (Baltimore, 1971). Briefly, the negative stranded RNA

enters the cell. The segments must first be transcribed to plus stranded


PRNA to be able to act as messengers. There appears to be two types of

plus strands, one which is poly-adenylated at the 3' end and one type

that is not poly-adenylated. It has been suggested that the poly-A

form acts as messenger for viral proteins while the form without poly-A

acts as template for production of progeny viral RNA (Hay and Skehel,

1979). The nucleocapsid is assembled in the cytoplasm of the host cell

while surface viral proteins are inserted into the cell membrane at var-

ious sites. Concurrently, host proteins are largely excluded from the

membrane at these specific sites. From these sites at the cell surface

progeny virus bud by some unknown mechanism taking with them the nucleo-

capsid and cell membrane with imbedded viral proteins. One recent report

which merits mention but has as yet not been confirmed or explained is

that DNA has also been found in the influenza virus and comprises between

5-8% of the total nucleic acid (Kantorovich-Prokudina et al., 1978).

Antigenic Characteristics

Influenza viruses are classified into serotypes and each serotype

is further classified into subtypes. The serotype depends upon antigenic

determinants found in the nucleoprotein. Three major serotypes have so

far been identified A, B and C. Types A and B cause the majority of

disease in man. The subtypes within each serotype are defined by anti-

genic determinants on the HA and NA. It is the characteristic genetic

variability of the segments of viral RNA coding for these two viral

proteins which causes the unpredictable "shifts" and "drifts" in antigenic

character. Shift is found only in type A but drift is found in both

types A and B. These shifts and drifts in viral antigens result in the

well known pandemics and epidemics, respectively. Shift refers to a major

change in antigenic determinants of a surface protein, for example, the

shift of the hemagglutinin from H2N2 to 113N2. This change has been shown

to be radical in nature. By mapping tryptic digests of these viral pro-

teins it can be shown that the amino acid sequences from 112 differ greatly

from those of H3 (Webster and Laver, 1975). These types of changes

have been hypothesized to happen by reassortment events after co-infection

of host cells. Animal influenza viruses have been implicated as a source

of these new and different antigenic determinants. The presence of these

new antigens on a virus then leaves the population with decreased immunity

to this virus (Rasmussen, 1964). A drift, for example a drift of H3N2

Port Chalmers (1973) to H3N2 Texas (1977), is a small change in antigenic

character (perhaps as little as a single amino acid change). This is

demonstrated, again, by tryptic digests of viral surface proteins (Webster

and Laver, 1975). Drift probably occurs by single mutational events at

a locus within the RNA segment coding for the protein.

The Disease

Influenza, although it has been with us for centuries, remains as

perhaps the last great pandemic of man. During the period from 1968 to

1979, more than 150,000 excess deaths are estimated to have occurred

during epidemics of influenza A in the United States (MMWR, 1979). These

excess deaths occur primarily among the young, the chronically ill and

older individuals.

The disease in humans is commonly a rhino-tracheobronchitis with

accompanying headache, malaise, fever, anorexia and myalgia. In uncom-

plicated cases recovery occurs in a few days to a week. In a few cases,

however, the virus can cause a pneumonia which quickly runs its course.

Prognosis is poor in this instance. Viremia may be found only in these

severe cases (Naficy, 1963). Supportive care has proven useful in some

cases but more often than not death ensues in 3-7 days after onset of

illness. Prior to widespread use of antibiotics, bacterial pneumonia

was a frequent complication following influenza infections and even

today remains the major cause of influenza related deaths. Bacterial

infection after influenza was most probably the major cause of death

during the 1918 pandemic which claimed 20 million excess deaths world-

wide. It has been estimated, however, that a full 20% of the 20 million

excess deaths were directly due to viral pneumonia with death occurring

in 6 or less days (Crosby, 1976).

Immunity to Influenza

That an important part of the immunity to influenza is specific

is well known. Animals with prior infection with a subtype of influenza

(e.g. HON1) will at least for a period be solidly immune to challenge

with that same virus but are still susceptible to infection with a dif-

ferent type A influenza virus (e.g. H3N2). Specific immunity can be

either cell mediated (CMI) or antibody mediated (Ab). Further, the

immune response may be either local or systemic in nature. The obvious

question then arises: Which is these, local Ab, local CMI, systemic Ab

or systemic CMI is most responsible for the specific resistance to rein-


Antibody Mediated Immunity

If we turn first to Ab, most researchers can be divided into two

schools: those that conclude that systemic Ab protects the animal against

infection and those that believe that serum antibody (Sab) does not cause

protection but only correlates with it and that local antibody (Lab) is

responsible for resistance to reinfection. In the former group Stuart-

Harris (1973) reported that serum antibody was protective in man. Both


Allan et al. (1971) and Rott. et al. (1974) reported serum antibody to

be protective in chickens. More recently, Virelizier (1975) concluded

that in mice serum antibody was most important for protection against

reinfection by influenza. McLaren et al. (1974), using ferrets in

challenge experiments, found serum antibody against the hemagglutinin

to correlate best with protection against infection. Serum antihemag-

glutinin antibody is thought to be the effector in the above examples.

However, Murphy et al. (1972), working with human volunteers and Schulman

et al. (1968), working with mice,have concluded that serum antineuramini-

dase antibody can also be protective against infection.

There are also those that have observed the reverse situations,

that is, that Sab is not protective against infection. Morris et al.

(1966) demonstrated that influenza infection in humans is not prevented

by high levels of serum antibody. Small et al. (1976) passively immunized

ferrets so that their hemagglutination-inhibition titers (HAI) were 1000

or greater, but were unable to prevent influenza infection. They con-

cluded that serum antibody played no role in preventing infection of

ferrets. So it appears there is evidence both supporting and refuting

serum antibody as an important immune mechanism in preventing influenza


Many investigations support the idea that antibody produced locally

in the respiratory tract and distributed on the mucosal surface is the

primary mode of protection. Richman et al. (1974) have shown .that the

local antibody response is specific. Francis (1943) infers that nasal

wash antibody specific for infecting influenza virus is the most important

determinant of specific resistance to influenza. Frazekas de St. Groth

and Donnelley (1950a), studying the antibody response to mice receiving

graded doses of influenza virus (live or inactivated) by various immuni-

zation routes concluded that "serum antibody does not give any informa-

tion on antibody that is found at the local site." In further work

(1950b) they conclude that serum antibody does not coincide with immunity

but local antibody does. Beare et al. (1969) concluded that a factor,

probably local antibody, exerted a considerable influence on human

resistance to infection by influenza virus. Waldman and Coggins (1972)

concluded that a vaccine's ability to stimulate nasal secretary antibody

was the deciding factor as to whether it could protect humans against

influenza infection. Potter et al. (1972) showed production of nasal

antibody after nasal infection of ferrets but not after immunization

with killed virus. They postulated this was the reason for better pro-

tection of these animals upon subsequent challenge with virus. All of

the above studies show only a correlation between local antibody and

protection but do not prove that the local antibody is the cause of

immunity. However, Barber and Small (1978) in experiments using ferrets

with surgically created tracheal pouches, which allow study of infection

and protection in two separate sites, conclude that systemic factors do

not play a role in preventing upper respiratory tract infection with

influenza and that, therefore, local factors are primary in prevention

of infection.

Again there is evidence to support the opposite view, that is,

that local antibody, is not responsible for immunity to influenza. In a

recent report Jennings et al. (1978), using medical student volunteers,

specifically concluded that nasal wash antibody does not correlate with


Much of the confusion as to which is more important for protection,

serum antibody or local antibody, may stem from two problems of experi-

mental design. First, those that look only at Sab and not Lab can only

show correlation with the one parameter and are unable to rule out the

other parameter, Lab. Secondly, different workers study different sites

of infection having different immune mechanisms in operation. Virelizier

(1975) and Loosli (1953) while concluding that Sab protects against influ-

enza infection were studying the lung. While Sab probably does protect

the lung in their animal system, it probably does not protect the upper

respiratory tract, the tissue affected in the vast majority of cases of

human influenza. Ramphal et al. (1979) have shown in mice that serum

antibody given prior to infection will protect the lung from pneumonitis

but will not protect the trachea, thus separating the upper and lower

respiratory tract immunologically (see Appendix B).

Cell Mediated Immunity

The possible role of CMI in immunity to influenza has only recently

begun to be examined. However, it has been shown to be important in both

recovery and pathogenesis. In studies where nude mice (Sullivan et al.

1976),or mice treated with antilymphocyte serum (Suzuki et al. 1974), were

infected with influenza virus significant prolongation of survival time

was noted. Also, virus was not eliminated from these mice. This suggests

that, in severe infections, CMI is responsible for some pathogenicity of

the virus and that-T-cells are required for.recovery from influenza

infection. This, however, has not been shown to be the case in studies

by Yap et al. (1979) where they conclude that influenza immune T-cells

inhibit rather than contribute to pulmonary pathology. However, they do

agree that CMI may play an important role in recovery from murine influenzal


Cytotoxicity is seen by many as the major cell mediated effector

mechanism influencing the outcome of influenza disease. Yap and Ada

(1977) have shown that in mice this cytotoxicity is probably due to T-

cells. Yap and Ada (1978), speaking specifically about systemically

derived cytotoxic T-cells found in the lungs of mice, conclude that

these cells can bring about recovery and are part of the consolidation

process. Ennis et al. (1978) also conclude that cytotoxic T-cells are

derived systemically and that they are part of the immunological and

pathological response to virus infection.

It has been shown in vitro that cytotoxicity against influenza

infected cells displays the same histocompatibility restriction as has

been shown for other viruses (Braciale, 1979). However, it is not

specifically known if H-2 antigens exist in vivo on the epithelial cells

which are infected by influenza virus.

The mechanism of cytotoxicity is unknown. It has been suggested

that both antibody and cells may be involved and that this might there-

fore be an antibody-dependent cell cytotoxicity (Lucas and Barry, 1977,

Greenberg et al., 1977). Further, cytotoxic cells have been shown

generally to be of the Ly-23 subclass and that this activity can be

amplified by the Ly-1+ subclass of T-cells (Cantor and Boyse, 1975).

It has become clear that CMI reactions during influenza infection

are not as specific as humoral responses. Doherty et al. (1977) have

shown in mice least two populations of cytotoxic T-cells arise

during influenza infection, one specific and one cross-reactive. Also,

during secondary response, the cross-reactive population may be preferen-

tially stimulated. Biddison et al. (1979) have shown that cross-reactive

cytotoxic T-cells exist in man and that secondary response T-cells are

preferentially cross-reactive. What is the cause of this cross-reactivity?

Webster and Hinshaw (1977) suggest that M-protein, although not expressed

on the surface of the virion or on cell surfaces, may be responsible for

the cross-reactive T-cells. M-protein is virtually the same in all type-A

influenza (the cross-reactivity does not extend across serotypes).

Cretescu et al. (1978) have shown that humans do make anti-M-protein

responses during infection, so evidently the M-protein can be recognized

by the immune system at some stage during infection. In fact Ada and

Yap (1977) and Braciale (1977) report that M-protein is expressed on the

cell surface. Other reports show that another antigen which is shared

by all subtypes within a serotype, the nucleoprotein, is expressed at

the surface of infected cells during early events (Virelizier, 1977).

This may, therefore, account for the cross-reactivity of T-cell cyto-


Local CMI responses have been shown in a few instances. Henney and

Waldman (1970), and Waldman and Henney (1971) demonstrated that pulmonary

CMI is relatively independent of the circulating response. Cambridge et

al. (1976), using migration inhibition and cytotoxicity assays, observed

in mice that the local response by pulmonary node lymphocytes was both

greater and faster than spleen cells after influenza infection, thus

suggesting the presence of local CMI. Wyde and Cate (1978) found that

during murine influenza an increase in cell cytotoxicity was found in

regional lymph node cells before peripheral blood lymphocytes. They infer

from this that the cytotoxic response found in bronchioalveolar washes

at Day 6 after infection is a local CMI response. In humans, Jurgensen

et al. (1978) show that CMI in the respiratory tract was best stimulated

by aerosol immunization while subcutanetous immunization stimulated

primarily systemic CMI. They presented this to support the importance

of local CMI during influenza infection.

Most definitive work proves upper respiratory tract immunity is

local (Barber and Small, 1978). Further it has been shown that CMI is

not specific enough in nature to explain the specific immunity seen in

influenza. These two statements taken together suggest local Ab as

the probable mechanism of resistance. However, reported work which

implicates Lab is not direct. We therefore assessed the role of antibody,

both local and systemic in vitro. Further, we assessed the role of Sab

in pathogenesis and recovery in vivo. We wish to relate our findings to

three areas: Are these factors (1) important in resistance?, (2) import-

ant in causing pathology seen during influenza infection? and (3) import-

ant in recovery from the disease?



Mature ferrets were obtained from Marshall Research Animals, Inc.,

North Rose, New York, and housed in individual cages under conditions

which prevent cross-infection. Five-week old A/J mice were obtained

from Jackson Laboratory, Bar Harbor, Maine. Goats were obtained from

the University of Florida. New Zealand white rabbits were obtained from

Kel Farms, Gainesville, Florida.


Influenza viruses used were A/PR/8/34(HON1) and A/Port Chalmers/

1/73(H3N2). Large stocks of both viruses were obtained by injecting

viruses into allantoic cavities of ten-day old embryonated chicken eggs

which were then incubated for 3 days at 360C at which time allantoic

fluid was harvested, pooled and stored at -85 C in 1 ml aliquots. The

HON1 virus had a chick erythrocyte hemagglutination (HA) titer of 1280

and contained 107.2 50% egg infectious doses/ml (EID50). The H3N2 virus

had an HA titer of 160 and contained 10 EID /ml. Sendai virus was

propagated in the same manner and had an HA titer of 512 containing
10 EID 0/ml.

Intranasal Inoculation of Virus Into Animals

Ferrets were anesthetized with 0.5 cc of Ketaset (Ketamine hydro-

chloride, Bristol Laboratories) and infected with 0.1 ml of undiluted

virus in each naris. Mice were anesthetized with 0.2 cc nembutal

(6 mg/ml sodium pentobarbitol, Abbott Laboratories, Chicago) and infected

with 0.05 cc of undiluted virus per naris.

Virus Adaptation in Mice

To obtain virus which was able to cause lethal pneumonia in mice,

H3N2 influenza virus was passage successively in mouse lung. Mice were

anesthetized, infected and three days later they were sacrificed via

cervical dislocation. The lungs aseptically removed and macerated in

5 mls of Lebovitz's L-15 medium. Homogenates of lung tissue were

centrifuged to clear debris and supernatants were used to infect the

next group of mice.


Virus was detected by inoculation of samples into the allantoic

cavity of embryonated chicken eggs that were 10 days old, as previously

described (Barber and Small, 1974). HA and HAI titers were performed

with a microtiter kit using disposable microtiter plates (Cooke Engineer-

ing, Alexandria, Virginia) as described by Sever (1962). Sera used for

HAI assays were first absorbed with kaolin and chicken RBCs and heated

at 560C for 30 minutes as described previously (Barber and Small, 1974).

Tissue Cultures

Ferrets were anesthetized and exsanguinated by cardiac puncture.

Trachea and bladder were aseptically removed and placed into sterile

100 x 15 mm petri dishes containing approximately 20 mis of Hanks'

balanced salt solution with 100 units/ml penicillin and 100 vg/ml strepto-

mycin. While in the petri dish the trachea was cut into individual rings
and the bladder was cut into pieces approximately 3 x 3 mm Tissue

pieces were then put individually into 35 x 10 mm petri dishes with 3 mis

of the L-15 medium (glutamine [.3 mg/ml], 10% fetal calf serum, gentamicin

[5 pg/ml], streptomycin [100 pg/ml], penicillin [125 units/ml] and

mycostatin [100 units/ml]). The cultures were incubated overnight at

35 C. The following day (Day 1) the 3 mis of medium was removed and

replaced with 3 mis of the medium containing the challenge virus at the

proper dilution (Figure 1). If antiimmunoglobulin (a-Ig) treatment was

done, the a-Ig serum was added after tissues were cut up but still in

100 x 15 mm dishes on Day 0. The a-Ig serum was added in equal parts with

medium and cultures were incubated at 350C. The next day the individual

pieces of tissue were put into 35 x 10 mm petri dishes with media without

antiimmunoglobulin and cultures were immediately challenged with virus

as before.

Statistical Analysis

Viral and antibody titers were compared using the t-test (Men-

denhall, 1975) and mortality compared using Fisher's exact test

(Siegel, 1956).

Serum Antibody Production

Ferrets previously infected with H3N2 influenza virus were inoc-

ulated IM with 0.5 cc of A/Port Chalmers/1/73/N3N2 killed vaccine (MRC-11).

Two weeks later they were bled via cardiac puncture and serum was col-

lected. Goat serum antibody was obtained from a goat given the same

A/Port Chalmers vaccine IM weekly for three weeks. The goat was bled

via the external jugular. The sera were heat inactivated at 560C for

30 minutes and then filter sterilized before use. In some experiments

very high titer antibody was needed. Thiswasprepared by three succes-

ive 33% ammonium sulfate precipitations and centrifugations at room temper-

ature. After the last spin the precipitate was redissolved in saline and

the solution was then dialyzed against saline for two days. They were

then heat inactivated and filter sterilized.


2 .


w I

I I: I

I- *I

z z

-21 -19 -17T 4 -2 0 I 2 4 6 8 60

Z- Z > >

0 f *d

Experimental protocol; experimental sequence in days.

jijurj 1.

Antiimmunoglobulin Production

Rabbit Antialpha Serum

Preparation of rabbit antialpha serum was performed by Richard

Kris in our laboratory. Surgery was performed on ferrets to canulate

the bile duct. Bile was collected over a period of a few days. The

bile was subjected to preparative electrophoresis on acrylamide (3% aga-

rose) at pH 6.3, 5 milliamps for 18 hours. The gammaglobulin fraction,

collected in the buffer at the top of the gel, was then concentrated by

pressure dialysis. This was then gel filtered over an Ara-22 column

(Pharmacia) and the 11S fractions were pooled. One-half of a mg of

this in complete Freund's adjuvant (CFA) was then injected subcutaneously

into a rabbit which had previously been injected with ferret IgA prepara-

tions more than one year prior to this time. Two weeks later the rabbit

was bled from the ear artery and serum was collected. This serum was

then subjected to affinity chromatography by passing it over a sepharose

4B CNBr column (Pharmacia) with ferret IgG attached. The resulting serum

was shown by ouchterlony analysis to have anti-IgA activity but no activity

toward other immunoglobulin classes.

Rabbit Antigamma Serum

Again, preparation of rabbit antigamma serum was largely performed

by R. Kris. Ferret serum was treated with KBr and centrifuged to elim-
inate lipoproteins. The supernatant was dialyzed against PBS (103 M).

The pseudoglobulin fraction was passed over a DEAE column and the 7S

fractions were pooled. One-half of a mg of this 7S fraction in CFA was

injected subcutaneously into a rabbit. Three weeks later the rabbit was

boosted with 0.5 mg in Freund's incomplete adjuvant. Three weeks later

the rabbit was bled for serum. The anti-IgG serum was subjected to

affinity chromatography by passing it over a sepharose 4B CNBr column with

IgA attached. The resulting serum was shown by ouchterlony analysis to

have activity against ferret IgG but not ferret IgA.

Interferon Assay

Ferret kidneys were removed aseptically and put into Gey's A solu-

tion. The kidney capsules were stripped and the kidneys were minced

with scissors. The minced tissue was washed twice with Gey's solution.

The tissue was put into a trypsinization flask with 100 mls Gey's A,

(Streptomycin 125 pgms/ml, penicillin 250 units/ml), 0.1% trypsin and

0.04% versene. The flask was put into the cold (40C) overnight. The

next day the flask was put on a magnetic stirring device at low speed

and 370C for 30 minutes. The loose cells were decanted and centrifuged

at 1000 rpm for 15 minutes. The supernatant was poured off and the pellet

resuspended in medium. This was centrifuged and resuspended as before.

Cell density was adjusted to approximately 106 cells/mi and 20 mis were

dispensed into 75 cm3 culture flasks which were then incubated at 37 C.

Cultures were fed after three days with MEM (10% FCS and 0.03 M IIEPES)

and thereafter twice weekly. After monolayers were confluent they were

washed once with Gey's A solution and treated with trypsin/versene solu-

tion until the cells came off the plastic. The culture flask was washed

twice with 10 mis of medium. Cell density was adjusted to 2.5 x 105

cells/ml and 1 ml was dispensed into each well of a 24 well plate. The

next day samples to be assayed were incubated with ferret antiinfluenza

virus antibody (HAI = 1000) for 30 minutes at 37 C. Samples were then

diluted serially and put onto cell cultures overnight. Fluid was withdrawn

and 0.1 ml of a dilution of VSV (giving between 30 and 300 plaques/well

on controls) was dispensed to each well for 1 hour. Methyl-cellulose

overlay was put on and two days later removed. Cell monolayers were then

stained with crystal violet for 20 minutes. Plaques were counted and 50%

plaque reduction titers were calculated.

Neutralization Titers

Samples were heat inactivated at 56 C for 30 minutes. Samples were

then diluted serially and equal amounts of samples and virus (100 EID50/

mls) were incubated for 1 hour at room temperature. Ten-day old embryon-

ated chicken eggs were injected with 0.1 ml of incubated mixtures. Eggs

were incubated at 36 C for two days, then harvested and hemagglutination

assays performed and results recorded. That dilution of the sample that

gave 50% neutralization of the virus was determined to be the neutrali-

zation titer.


Mice were x-rayed using Dupon Extremity A film (10 x 12 inches)

and an Extremity I cassette. The machine used was a Phillips XF 3001.

Films were takn at 40 KV, 16 MAS, and 1/20. Developing was accomplished

with a Kodak RP X-omat automatic processor, model M6A-N. Animals were

held extended and in place by extending a string across a styrofoam board

and allowing the animals to grab hold of the string and pulling them back

by the tail. The styrofoam board was approximately 1 inch thick because

thicker boards cause loss of clarity in the radiograph. This board was

rested on top of the cassette as the picture was taken. Individuals

holding the tails of the mice were always dressed in protective clothing

and wore a film badge for monitoring levels of radiation exposure.

Scanning Electron Microscopy

Trachea samples for scanning electron microscopy (SEM) were

placed in a buffered fixative composed of 2.5% glutaraldehyde, 0.1 M

sodium cacodylate, and 0.1% CaC12 (pH 7.4) and allowed to fix for at

least 24 hours before further preparation. They were then removed

from the fixative and dehydrated in graded concentrations of acetone

(70 to 100%). Specimens were critical point dried in a Bomar SPC 900/Ex

critical point drying machine (Bomar Corp., Tacoma, Washington) coated

with gold-palladium in a Hummer II shadowing machine (Technics, Alexandria,

Virginia) and examined with a Novascan 30 electron microscope (Semco,

Ottawa, Canada).



Organ Culture

We began our studies by examining local immune mechanisms by use

of ferret organ cultures. Organ cultures consist of fragments of tissue

taken from the adult animal and maintained in vitro. Cell proliferation

does not normally occur during the short time the cultures are in use

and the tissue retains many of its in vivo characteristics. Organ cultures

can provide a high degree of sensitivity and economy in determining factors

at work in tissues (Schmidt and Maassab, 1974).

Many uses have been found for organ cultures, including the study

of resistance to reinfection. Heuschele and Easterday (1970) used organ

cultures derived from chicken trachea to study resistance to reinfection

by Newcastle Disease Virus (NDV). They suggested a potential role for

organ cultures to study mechanisms of formation and/or secretion of local

antibody. Finkelstein et al. (1972) again using chicken tracheal organ

cultures demonstrated that resistance to NDV was probably partially

mediated by interferon. Resistance after intratracheal immunization

correlated only with tracheal antibody and could be partially blocked by

exogenously applied rabbit antichicken globulin. This suggested antibody

was present on mucosal surfaces which could have been produced locally.

Schmidt and Maassab (1974) found specific resistance to influenza virus

in tracheal organ cultures from chickens. They suggest that this immunity

was at least partially derived from a secretary immune system.

Organ cultures have been shown to be an excellent system for

studying pathogenesis of respiratory infections. Klein and Collier

(1974) studied pathogenesis of human parainfluenza type 3 virus infection

by using hamster tracheal organ cultures. Human organ cultures have been

used successfully to study the cytopathic effect of this same pathogen

(Craighead and Brennan, 1968).

The relationships between influenza strains that infect humans and

animals can be studied in organ cultures. In other words, they offer a

way of determining the host ranges of the different strains in a precise

and economical manner. This is important as it relates to the hypothesized

recombinational events occurring in simultaneously infected host cells

causing the creation of a new strain of influenza virus capable of causing

pandemics (see general Introduction). Schmidt et al. (1974) studied human,

swine, equine and avian influenza A viruses and tested them for infectivity

in chicken, ferret, equine and porcine tracheal organ cultures. Results

showed that the homologous tissue was much more easily infected than hetero-

logous tissue.

Organ cultures have also been shown to be an excellent way for

screening potential live vaccines. Mostow and Tyrrell (1973) have used

ciliary activity in tracheal organ cultures as a measure of virulence and

therefore usefulness of vaccines. Vaccines which cause a small decrease

of ciliary activity have been shown to be extremely avirulent in vivo

while vaccines that are highly virulent thus causing marked decreases in

ciliary activity in organ culture are likewise virulent in vivo.

Ferrets and Influenza

We have elected to use ferrets for our studies of influenza, because

influenza infection in ferrets appears to resemble the human infection and

because ferret trachea are relatively long and contain a large number of

individual rings. Further, the size of the animal makes it easy to handle.

The ferret also lends itself to the organ culture system because of the

ease with which their tissues are infected by influenza virus. Bang and

Niven (1958) reported the cultivation of influenza virus in ferret nasal

mucosa. In perhaps the most extensive studies performed so far, Basarab

and Smith (1970) showed the growth patterns of influenza virus in cultures

of ferret organs. They reported virus replication in cultures of ferret

nasal mucosa, lung, trachea, oviduct and bladder.


Immunity in Ferret Organ Cultures

The basic experimental design is shown in Figure 1. In all experi-

ments ferrets, subsequently identified as convalescent, were infected 21

days before being killed. Nasal washes were taken 2 and 4 days after

this infection. Serum was taken at the time of infection and of killing.

Animals were proven to be infected by both virus isolation from nasal wash

(one exception noted in Table 1) and a 4-fold or greater rise in HAI anti-

body titre to the virus. At Day 0 normal and convalescent ferrets were

killed and their tissues put into organ cultures. Serum taken from normal

ferrets at the time of killing (Day 0) had no detectable HAI antibody

(<1:8) to influenza virus. Cultures were challenged with varying dilutions

of influenza virus on Day 1. The number of organ cultures/dilution varied

in different experiments depending upon the number obtained from each



b Virus Isolationa on Day
Virus Challenge Rings 0 4 6 8 10 12 14 17 19
Normal Ferret

Control 1 0 0 0 0 0 0 0 0 0
2 0 0 0 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0 0
4 0 0 0 0 0 0 0 0 0
5 0 0 0 0 0 0 0 0 0

10- 0 + + + + + + + +
2 0 + + + + + + + +
3 0 + + + + + + + +
4 0 + + + + + + + +
5 0 + + + + + + + +

-4 0 + + + + + + +
0 1 0 + + + + + + + +
2 0 + 0 + + + + + +
3 0 + + + + + + + +
4 0 + + + + + + + +
10 1 0 + + + + + + + +
2 0 + + + + + + + +
3 0 + + + + + + + +
4 0 + + + + + + + +
5 0 + + + + + + + +
6 1 0 + + + + + + + +
2 0 + + + + + + + +
3 0 + + + + + + + +
4 0 + + + + + + + +
4 0 + + + + + + + +

5 0 0 0 0 0 0 0 0 0

10-7 1 0 0 0 0 0 0 0 0 0
2 0 0 0 0 0 0 0 + 0
3 0 + + + + + + + +
4 0 + + + + + + + +
c 0 + + + + +
5 0 0 + + + + + 0 0

TABLE 1--Continued

Infected Rings (Day 8)
21 25 31 37 40 Total Rings OCID50



+ + +



+ 3/5
- +

TABLE 1--Continued

b Virus Isolationa on Day
Virus Challenge Rings 0 4 6 8 10 12 14 17 19
Convalescent Ferret

-2 1 0+ + + + +
10 1 0 + + + + + + + +


a+ = at least one of the duplicate eggs show HA 3 days after inoculation
with sample.
Dilution of stock virus containing 107.2 EID50/ml of A/PR/8/34 (HON1).
CRing which changed from not infected to infected.

ND = Not done.

TABLE 1--Continued

Infected Rings (Day 8)
21 25 31 37 60 Total Rings OCID0

+ +

10 5EID
10 EID50s/ml


+ + + +
+ 0 + +

0 0 0 0

+ + +

animal. Fluids were completely drained and replaced every other day and

supernatants were assayed for virus on Days 1, 4, 6 and 8 and in one experi-

ment periodically up to 60 days.

Table 1 shows data from Experiment IV. In general, once a culture

was infected it usually remained so throughout the experiment. Three of

the 55 cultures differed significantly from this pattern. Transient con-

versions from positive to negative occurred occasionally but were probably

due to the egg assay system which apparently can have false negatives but

not false positives. When a large number of false negatives appeared on

one day it probably can be attributed to improper handling of samples

(e.g. Day 31). Since very little variation occurred from day to day, Day 8

was selected to calculate an OCID50 (50% organ culture infectious dose).

OCID50S were calculated by the method of Reed and Muench (1938). For the

remaining experiments cultures were usually carried through Day 10 and

OCID50s calculated for Day 8.

Table 2 summarizes the results of Experiment IV (already shown in

Table 1) and Experiment V, in which tracheal organ cultures from 5 normal

and 5 convalescent ferrets were compared for relative resistance to influ-

enza infection. As can be seen, convalescent ferret trachea required

between 140 and 200 times more virus to infect them than did normal cul-

tures. This difference was statistically significant and hence the

experiments demonstrate the presence of immunity.

Immunity in Ferret Tracheal Organ Cultures with
Homologous and Heterologous Virus

To determine if the immunity demonstrated in tracheal organ cultures

was specific for the virus which had been used to infect the animal (homo-

logous virus), half of the cultures were challenged with the homologous

) IUn
o o
SO U 0
o o P+.H
0 0 0 00 .* H
4 V V 0 0
CT O c c -
v v V) v r u

r : VI -0

(l 0 O 0 OC
0o 0, H
0 *-- 0 0 (A t
0 -H X X

o ,- a+ o1 o
-- 4-0 0 0 X 0 >

C ) o 0 o C *H *
U) 0 (* +- 0 bO 0 *Hq C)
+ C -4 -4 0 0 4-

0qH 0 0 0- O
H 0 CD 0 0 En

ri r +q> 41

0 0 L e o 4
'-' >
So ?-4 0
S.i 0 0 4\ *(-
C 0 +1 4-J 9 t4 t o
LI) ,-q --4 9-- '-' 0 2 PV) 0 to

Sl U t H O -
LI 0 0 -0 0 4-.

o o a D e 0
H1 0 0f I C d, -4 %
t- C' U (NJ cj +. C' k ed -
LO >* U4- 0 :

u& 4-) : C

C O $1 0
-4 40 +1 0) > 0 .I)

*fl 'H O O o $4 '4 >
0 01 0 0 p4 o '
I M )O 0 0 1- -, 0 0
So 0 4-4 +

S 2 o 1o c. c ) n o
LI) -H C 0 --4
V)* r- N a -H 4 o u

;:U ^= a 0 0 0Q
S0 C r-4
H 0 04 J ) dH U Ci

C4 4 o) O 0 -
4 T 0 ) 1 0 -1)0

( 4-4 CT 0 0 uP o
0 rz; 0)0 0 >
r-4 n-r 0 0 ( C Or M
o*' 0 0 4 0 0 r
4-' *H 0 o 4
0 0 -4

: -H o- j o o *H 40 0 ,
.-HC) 0 C'. 0I0
C VU) 4j 0 >
F -4 c O C
cr3 bo : 0 Ln

-rj 0 O 4 O
E >4- O U0 U-4

t H H O +O a
H 0 w 0 0 0
S- 4 0H U
i r (0 0C '
** H \a 0 -H H

4Jt (N C 44 4-4 4-)

?i 0 0 Ho
S4- CD CD H 0 4
L/)O [) '4- .44 .
$44 4-4 0 0 0,A-
P4 4 '-4 4-Q 4- 0

virus and the other half with a different virus (heterologous virus).

The results are shown in Table 3. Rings challenged with homologous

virus (H3N2) were between 20 and 200 times more resistant to infection

in culture than normal ferret tracheal organ cultures. Cultures chal-

lenged with heterologous virus (HON1 or Sendai) showed no significant

difference in one experiment and only 6-16 times more resistance in two

experiments when the OCIDs50 were compared with those from normal rings.

The immunity therefore appears to be largely specific, although a variable

amount of nonspecific immunity is sometimes observed.

Mechanisms of Immunity in Ferret Organ Cultures

Having now shown that ferret tracheal organ cultures derived from

convalescent animals are specifically immune to challenge, we wish to

delineate the mechanisms involved. Since the immunity is largely specific,

the mechanism could be mediated by specific antibody and/or lymphocytes.

However, the character of CMI, as stated before, is not totally specific

but rather cross-reactive in nature. With this information it seemed

more likely that antibody was the mechanisms of specific immunity in

tracheal organ cultures and some other mechanism responsible for the non-

specific aspect. We therefore wished to test the possibility that either

serum antibody or local antibody was responsible for the specific effect.

Effect of Passive Administration of Serum Antibody
On Organ Culture Immunity (in vivo)

The protocol diagrammed in Figure 1 was followed except instead of

infecting ferrets at Day -21, animals received a total of 50 ml of either

immune ferret serum (HAI 2048) or normal ferret serum (HAI <16) in two

doses at Day -7 and Day -1. A third group of ferrets received no treat-

ment. Ferrets in the immune serum group had HAI titers of 64, 64, and



Number Animal/ OCID Log b,c (Mean SD)
Experi- of Dilution Virus Normal
ment Animals of Virus Homologous Heterologous Homologous

VII 4 4 HON1
4 H3N2 1.8 (0)

IX 6 2 HON1
2 H3N2 3.95(0.4)

XI 8 2 Sendai
2 113N2 2.2 (0.7)

Half the number of animals listed were normal and half were convalescent.

OCID50 obtained by using total infected and total uninfected cultures for an
entire group of animals.

COCID = 50% organ culture infectious dose as determined by the Reed-Muench
50% endpoint method (1938).

Antilog of (Log0 OCID50 convalescent -Logl0 OCID5 normal).

probability that difference occurred by chance.

TABLE 3--Continued

b,C Sc
OCID5 Log0 b (Mean SD)
Normal Convalescent
Heterologous Homologous Heterdogous










Ratio of




Weighted average




128 at sacrifice. Titers of normal and no treatment groups were all 8 or

less. Table 4 illustrates the results. Both trachea and bladder cultures

were assayed for immunity. The results show that OCID50s of organ cultures

derived from animals receiving normal serum (NS) did not significantly

differ from OCID50s of organ cultures from animals receiving immune serum

(IS). Further, the IS group did not differ significantly from the no

treatment group. We conclude from this that serum antibody does not cause

specific immunity in ferret organ cultures.

Antialpha Treatment of Organ Cultures

To test whether tracheal organ culture immunity is caused totally

or in part by IgA in the tissue, organ cultures were exposed to rabbit

antiferret alpha chain 24 hours prior to challenge with virus. Briefly,

ferrets were infected and 21 days later sacrificed along with normal

uninfected ferrets. Trachea were cut into rings and placed in a petri

dish with antialpha or rabbit normal serum overnight. Cultures were

challenged and sampled periodically as before. Results are shown in

Table 5. Organ cultures from convalescent ferrets exposed to normal serum

showed 16-fold immunity over organ cultures from normal animals also ex-

posed to normal serum (p = <.0. Organ cultures from convalescent ferrets

exposed to antialpha showed no significant immunity relative to normals.

However, it is also apparent that when cultures from normal ferrets were

treated with-the antialpha they became 200 fold more resistant to infec-

tion (p <.01). This result is unexplainable and therefore makes the

interpretation of the results of antialpha treatment ambiguous. We can

make no clear interpretation of these data.



OCID50 Log0(Mean SD)

of b c
Virus Animals NORx NS IS p p

Trachea H3N2 9 3.5 .1 2.9 .4 2.9 .8 NSd NS

Bladder H3N2 9 3.3 .9 3.0 1.0 3.5 .5 NS NS

aFerret anti-H3N2 antibody (64, 64, 128)

IS vs. NoRx

cIS vs. NS

dNot significant



Number of OCID
Treatment Animals (logl0 _0SD) Pa

Convalescent 3 4.9 .2 NSb
Normal 3 4.0 .6-
Rabbit Normal Serum
Convalescent 3 5.1 + .4 0 <.01
Normal 3 6.3 .2_

Convalescent 3 4.5 + .4- <.02
Normal 3 5.7 0
Rabbit Normal Serum NS
Convalescent 3 3.7 .6-
Normal 3 5.8 + .3 <.01

aStudents t test (small sample)

Not significant

Antigamma Treatment of Organ Cultures

To test whether tracheal organ culture immunity might also be

ablated by antigamma treatment we carried out the same protocol again

except that we substituted the use of rabbit antiferret gamma chain

antisera for the antialpha used above. Table 5 shows that convalescent

cultures treated with rabbit normal serum have 130-fold immunity

(p = <.01) compared to normal cultures treated with normal serum. Con-

valescent cultures treated with antigamma showed 16-fold immunity

(p = <.02) when compared with normal cultures. Further, convalescent

cultures treated with antigamma were not significantly different from

convalescent cultures treated with normal serum. This suggests that IgG

may not play a significant role in immunity of trachea] organ cultures.

Neutralizing Antibody in Tracheal
Organ Cultures

Day 4 cultures which received no virus were assayed for neutral-

izing antibody to the homologous virus (A/Port Chalmers) and a hetero-

logous virus (A/PR8). We were not able to show any significant neutral-

izing capability in the supernatants of the cultures to either strain.

This is shown in Table 6.

Interferon in Tracheal Organ Cultures

Interferon assays were performed on culture supernatants from

Day 4 cultures. Table 6 shows that no significant interferon activity

was found (<3 units/ml) in any culture.

Bladder Immunity

To get an indication of whether specific immunity was restricted

to the respiratory tract, a second anatomically distinct site was tested.

Basarab and Smith (1970) had shown that ferret bladder tissue was



Interferon Neutralizing Antibody
Treatment (units/ml) H3N2 HONi


Convalescent < 3 < 8 < 8

Normal < 3 < 8 < 8


Convalescent < 3 < 8 < 8

Normal < 3 < 8 < 8

Normal Serum

Convalescent < 3 < 8 .< 8

Normal < 3 < 8 < 8

susceptible to in vitro infection by influenza virus. Therefore, cultures

of trachea (results shown in Tables 2 and 3) and bladder were taken from

ferrets in some experiments. Table 7 shows the OCID50s for the normal

and convalescent bladder in four experiments. In Experiment V bladder

immunity is demonstrated since convalescent bladder tissue required

approximately 200 times more virus to infect than did bladder from

normal ferrets (statistically significant at p < 0.10). From Experiment

VII it appears that nonspecific immunity exists since it took 20 and 16

times more virus, respectively, to infect convalescent tissues than to

infect normal tissues with heterologous and homologous virus. Levels of

significance varied with a p value of 0.10 for heterologous and a value

of < 0.10 for homologous; hence this is not a conclusive experiment.

The last two experiments (IX and XI) show specific immunity. In experi-

ments which include tissues from a total of 14 animals it took 500 and

40 times more homologous virus to infect convalescent tissues than to

infect normal tissues (significance p = 0.025 and p = < 0.005). At the

same time, it took 10 and 0.25 times more heterologous virus to infect

convalescent cultures than it took to infect normal cultures (significance

level of p = 0.10 and p = < 0.005). Thus 3 of the 4 experiments suggest

specific immunity in bladder to homologous virus.

Ciliary Activity in Tracheal Organ Cultures

We attempted to correlate ciliary activity with infection of tracheal

organ cultures. Organ cultures were observed under 100X power using an

inverted microscope. Ciliary activity of rings was qualitatively evalu-

ated by determining the percentage of the inner circumference of the

ring showing detectable activity and by the intensity of ciliary beating.

We observed a significant decrease of both parameters by activity of



Pieces of
Number Animal/ OCID 0 Log0, (Mean SD)
Experi- of Dilution Virus Normal
ment Animals of Virus Homologous Hfeterologous Homologous

2 or 3












aHalf the number of animals listed were normal and

bOCIDs obtained by using total infected and total
entire group of animals.

half were convalescent.

uninfected cultures for an

COCID = 50% organ culture infectious dose as determined by the Reed-Muench
50% en~point method (1938).

Antilog of (Log10

OCID50 convalescent -Log10 OCID50 normal).

e21 or 22 animals shown to be convalescent by both criteria: (a) Virus
shedding as determined by assay of nasal wash; (b) 4-fold increase in specific
Ab titre after infection. (One convalescent animal in Experiment V was not
tested for (a) due to lost samples.

Probability that difference occurred by chance.

TABLE 7--Continued

OCIDn5 Log,* b,c (Mean SD)c Ratio of Infectivityd
Normal 0 Convalescent Homologous Heterologous f
Heterologous Homologous Heterologous Challenge Challenge p

2.52(0.5) 200 <0.10

0.2(0) 1.5(0.35) 20 0.10
4.0(0) 16 <0.10

0.7(0) 1.7(0.7) 10 0.10
5.0(1.0) 500 <0.025

2.3(0.25) 1.7(0.2) .25 <0.005
4.6(0.9) 40 <0.005

Weighted average 190 8

both infected and noninfected tracheal organ cultures during the first

ten days. Upon further manipulation, this decrease in activity was not

as apparent as originally seen. It was observed that upon agitating the

tissue in it's media there was a significant increase in activity. With

this method we were unable to conclude that there was a significant dif-

ference in ciliary activity between infected and noninfected cultures.

We wished to look at the surface of the rings to determine if

infection caused cell loss and if we could detect a difference between

infected and noninfected rings. To determine the amount of desquamation

occurring in tracheal organ cultures we employed SEM to observe rings

at various times, up to 28 days, after infection in culture. No more

than 50% loss of ciliated cells was seen in any culture on any day.

Further, there seemed to be no difference between infected and noninfected

cultures. We also compared the ciliated surface of trachea taken from

animals infected in vivo. Animals were sacrificed on Days 0, 3, 4 and

5. No loss of ciliated cells was noted. These results suggest that in

our organ culture system neither ciliostasis nor desquamation can be used

as an indicator of infection of cultures. The possible explanation is

that this virus is not able to cause desquamation of ciliated epithelium.


Organ cultures were used to study immunity in tissues of ferrets

exposed to influenza A virus. Tracheal organ cultures from ferrets con-

valescent from influenza infection required about 130 times more homologous

virus to become infected than cultures from normal ferrets. It took only

about 9 times more heterologous virus to infect convalescent cultures

than to infect normal cultures. Immunity was therefore largely specific.

Bang and Niven (1958) briefly reported without giving their data or

experimental details that ferret mucosal tissue from convalescent animals

was not resistant to infection. It is not clear why their results dif-

fered from those presented here.

Bladder tissue cultures were used to test if the specific immunity

was localized in the respiratory tract or was more widespread. Experi-

ments show that bladders from convalescent ferrets were about 190 times

more resistant to challenge with homologous virus than normal bladder,

and that convalescent tissues were only about 8 times more resistant to

heterologous challenge. Bladder specific immunity could be explained in

at least two ways: (a) the specific immunity is caused by systemic

factors, or (b) it is a local response caused by either antigenemia or

homing to bladder mucosal tissue of specific lymphocytes stimulated in

the respiratory tract. Basarab and Smith (1970) did show that influenza

virus could replicate in vivo in bladders of ferrets. And it has been

shown that during severe influenza infection virus can be recovered from

urine of patients (Naficy, 1963).

Since the immunity in both trachea and bladder appears to be largely

specific, it could be mediated by antibody and/or lymphocytes. However,

it has recently been demonstrated in cytotoxic studies using influenza-

infected target cells (Effros et al., 1977; Zweerink et al., 1977) that

CMI may be less specific than is required to account for the specific

protection measured in challenge experiments. Therefore, antibody seems

to be the more likely mechanisms for prevention of influenza in ferrets.

Using tracheal organ cultures we tested which of the following

hypotheses might be true: (1) Systemic serum antibody could have been

responsible and simply be trapped in the mucous secretions and/or in the

tissue itself or (2) the immunity could be locally produced in submucosal

immunocompetent cells. Organ cultures derived from animals receiving IP

injections of Sab were not significantly more resistant than organ cultures

derived from normal ferrets. This suggests that Sab plays no role in

tracheal organ culture immunity. The second alternative of locally pro-

duced antibody was tested and we showed that antigamma serum did not

ablate the in vitro resistance of tracheal rings. Antialpha serum seemed

to ablate the response but results from antialpha treatment of normal

rings made a clear interpretation impossible. Although we were not able

to clearly implicate IgA in resistance we were able to show that IgG does

not contribute significantly to resistance. This agrees well with

studies which quantitate immunoglobulin levels in the respiratory tract.

Waldman et al. (1973) showed that IgA levels increased dramatically as

you ascend from the lower respiratory tract to the upper. Further,

histological studies show a preponderance of IgA and IgM positive plasma

cells in the lamina propria of secretary tissues (Bienenstock et al.,

1978). However, little IgM is found in secretions (Waldman et al., 1970)

so it seems likely that IgA is the primary immunoglobulin able to act in

the upper respiratory tract.

An attempt was made to measure interferon and neutralizing antibody

in supernatants of organ cultures. Both assays, done with samples taken

from cultures on Day-4, were negative. Interferon levels were less than

3 units/ml and neutralizing antibody titers were less than 8 in all cases.

This does not necessarily mean that these substances were not present in

measurable quantities in the tissue or in close proximity to the rings,

but we were not able to find significant amounts once dilution occurs

in the total volume of the media on each culture.

We observed that ciliary activity could not be an accurate indicator

of infection of tracheal organ cultures since agitation of cultures

seemed to significantly restore activity which had been seemingly lost

over the first ten days. This suggests that mucous could entirely cover

surface areas which then would appear to have lost ciliary activity, or

mucous could slow the beating of cilia thus causing a further decrease

in activity. We are unable to say why these observations differ from

those of Mostow and Tyrrell (1973).

Further study using SEM showed that desquamation was not complete

in tissue from tracheal organ cultures and, in fact, no culture showed

greater than 50% loss of ciliated cells. Also, tracheas taken from

infected ferrets showed no loss of ciliated cells. We are unable to say

why these observations do not agree with those seen in mice by Ramphal

et al. (1979). Possibly our egg grown virus is not virulent enough to

cause total desquamation. Or perhaps ferret trachea is more resistant to

loss of ciliated cell than mouse trachea.

It is difficult to assess the value of in vitro studies in the

in vivo situation, especially when animal models are used to study human

disease. In the case of organ cultures used for this study two important

factors are offered for consideration. First, Rosztodzy et al. (1975)

have shown that human fetal tissues have similar susceptability to

challenge in culture with influenza virus as ferret tissues. Secondly,

Mostow and Tyrrell (1973) have shown that attenuated human influenza

viruses have similar activity in ferret organ cultures to that found in

vivo in humans. These studies suggest that ferret organ cultures are

a tool for studying influenza in humans.

Turning from prevention of infection to recovery from infection,

recent studies showed that mice with deficient CMI (nude mice; Sullivan

et al., 1976) or mice treated with ALS (Suzuki et al., 1974) shed virus

over longer periods than did normal mice, suggesting CMI may play a

critical role in recovery. In Experiment IV we showed that once a

ferret tracheal organ culture was infected it remained so; that is, it

did not recover. If CMI is responsible for recovery, it would follow

that CMI was not functional in the tracheal organ culture.

Irrespective of the mechanism of prevention or recovery from

influenza, it seems that the ferret tracheal organ culture enables one

to separate the two mechanisms. Immunity to reinfection can be demon-

strated in the same piece of tissue that lacks the ability to recover.

Hence it seems that prevention and recovery are mediated by different


The role of antibody in recovery will be addressed in

Appendix A.





As was stated before, the major cause of death related to influenza

is subsequent bacterial pneumonia. This is treatable by antibiotics.

However, in a small number of cases viral pneumonia is the direct cause

of death. Why in these few cases does the lung become involved to such

a great extent? Immune factors are evidently one of the major influences

in this determination. There is good support for the idea that Sab, if

present, is a factor which prevents lung involvement. This is a likely

explanation taking into account the intimate association of the lung

parenchyma and the pulmonary circulation. Loosli (1953) showed the

importance of circulating antibody in the prevention of death of mice.

Virelizier (1975) reported that HAI antibodys have an important role in

protection. More recently, Ramphal et al. (1979) reported that Sab pro-

tects the lung from pneumonitis but does not protect the ciliated epi-

thelium of the trachea. So it appears that if Sab against the infecting

influenza virus is present prior to infection, the lung will be protected

against viral pneumonia but the animal can still have an upper respiratory

infection. When drift from one subtype of influenza virus to another

subtype occurs very few cases of viral pneumonia are found. The reason

for this may be that Sab formed during infection by the closely related

strain is cross-reactive with the second strain of virus thus providing

lung protection. During a shift, when viral pneumonia is found to a

greater extent, cross-reacting Sab is not available because of the major


shift of viral antigens thus the lung is left unprotected and lung involve-

ment is more likely.

Since it has been proven that Sab does protect against death, by

decreasing viral pneumonitis, we wished to determine if Sab could be used

therapeutically to treat diagnosed or suspected viral pneumonia. If a

cure could be affected, we wished to determine how soon before death Sab

could be effective for treatment.


The virus used in these experiments was passage in mouse lung

11 times. A lethal dose 50 (LD50) was performed in mice and we found

that 0.05 cc of the fluid contained 10 3.9LD50s. In all subsequent experi-

ments this virus was diluted so that mice received approximately 10 LD50s.

In our first experiments we wished to ensure that the Sab we were

using could be life-saving upon subsequent infection with the eleventh

passage influenza virus. Briefly, one group of A/J mice were injected

intraperitoneally (IP) with a total of 0.5 cc of goat antiinfluenza

antibody (HAI 16000) in equal doses on two days. A second group received

no Sab. Both groups were infected on the day the first group received

its second dose of Sab. Animals were observed for deaths. Results

are shown in Table A-1. No deaths were recorded in the group receiving

Sab (0/5 dead) while those that did not receive Sab all died by two

weeks time (5/5 dead). We conclude that the dose of virus used is

lethal for our mice and that the dose of Sab was sufficient to protect

mice from lethal pneumonitis when given before infection.

We asked could Sab given after infection be life-saving? The

experiment consisted of infecting mice and then subsequently giving Sab

on specific days. In a preliminary experiment mice were infected and



Number of Animals Dead
Treatment Number of Animals Infected

No serum 5/5 -
p = .004
Saba 0/5 -

aSab given in two doses prior to infection,
0.25 cc/dose.

on Days 3, 4 and 5 after infection one group of these mice were given

Sab (0.75 cc total) and a second group was given nothing. Animals were

observed for death. Results for Experiment I are shown in Table A-2.

Mice not receiving Sab all died between Days 7 and 10 (4/4 dead) but

those receiving Sab on Days 3, 4 and 5 after infection all survived

for at least 21 days (0/5 dead). We conclude that Sab given on Days 3,

4 and 5 after a lethal influenza infection can be life-saving.

The next experiment was to determine how close to death mice could

be when administered Sab and still survive lethal infection. As before,

mice were infected and groups of mice were given (a) nothing, (b) Sab

on Days 3, 4 and 5, (c) Sab on Days 5, 6 and 7, or (d) Sab on Days 7, 8

and 9. Animals were observed for death. Results are shown in Table A-2

under column Experiment II. Again those mice that did not receive Sab

all died on Day 8 (7/7 dead). Mice that received Sab on Days 5, 6 and

7 all died between Days 5 and 8 (6/6 dead) and mice given Sab on Days 7,

8 and 9 all died between Days 7 and 9 (6/6 dead). The only group to be

affected by the Sab was the group receiving Sab on Days 3, 4 and 5 (1/7

dead). This one death occurred immediately after Sab injection on Day 4.

This was a significant decrease in mortality when compared with controls

(p = 0.002). This experiment was then repeated exactly as before but with

larger numbers of mice. Results are shown in Table A-2 under Experiment

III. Again, mice that did not receive Sab all died between Days 7 and 10.

Those receiving Sa' on Days 7, 8 and 9 all died between Days 6 and 10

after infection. In the group that received Sab on Days 5, 6 and

7 all but one animal died with deaths occurring between Days 6 and 8.

The one group that differed significantly from the control group was the




Treatment ment I Experiment II Experiment III

No Serum 4/4a 7/7 10/10 -

Sab b p = 0.000006
Days 3, 4, 5 0/5 1/7 1/13 -

Days 5, 6, 7 NDC 6/6 -p = .002 11/12

Days 7, 8, 9 ND 6/6 14/14

Mortality (number of animals dead/number of animals infected)

Dose of Sab 0.25 cc

CNot done

erouo that received Sab on Days 3, 4, and 5 where only a single death

was recorded (1/13 p = 0.000006).

The results so far obtained seemed to suggest if the lethal

infection by influenza could be diagnosed and treated by Days 3 or 4,

a cure could be affected. The determining factor in clinical diagnosis

of viral pneumonia, however, is often radiological evidence of consolida-

tion. We therefore attempted to discern if radiological changes were

evident by Day 3. If so, then it seemed possible that in a clinical

situation there might be a chance that the diagnosis might be made and

that subsequent treatment with Sab might benefit the patient. A control

group was infected with the virus and x-rayed on Days 3, 5, 7, and 9.

Photographs of representative radiographs are presented in Figure A-i.

It can be seen that no radiological evidence of lung involvement is

present until Day 7 or after. This indicates that, since by Day 7, Sab

cannot modify the lethal effect of infection with the virus that it

seems unlikely that Sab can be a useful treatment for lethal penumonitis.


We have shown that high titered Sab (HAI 16000) given to mice

before a lethal infection of virus (10 LD50s, which kills in 7-12 days)

can be life-saving. Further, when Sab is given after a similar infection

but before radiological changes appear in mouse lung the treatment is

also life-saving. However, if Sab is given after radiological changes

are noted, death occurs as if no treatment were given.

Because of our original success with Sab treatment of mice prior to

infection (Ramphal et al., 1979, see Appendix B) and because totally

effective treatment of lethal viral pneumonia is not available we had

Figure A-i. Radiological appearance of mouse lungs after
infection with lethal influenza virus (A/Port Chalmers, 10 LD50s).

A. Day 3

B. Day 5

C. Day 7

D. Day 9






5 ~: !!;~

i';~"~3iY"ilii~[lsBI~:: i~;i~' ~



hoped to show that Sab given after infection could be useful in a life-

threatening situation. However, clinically, x-ray evidence of consolida-

tion is the primary diagnostic tool in determining lung involvement dur-

ing viral disease. Because of this and because we were not able to

modify the lethal effects of the advanced disease state, therapeutic use

of Sab seems less promising.

It is probable at this stage of disease that immune processes

have begun to overwhelm the majority of lung tissue. Epithelial cells

of alveoli have been disrupted and cellular infiltrates and edema are

present and increasing. Alveoli become filled and at this point Sab

acting to neutralize virus or coat infected cells has negligible effect.

In fact, Sab may increase vascular pressure which in turn increases

edematous flow into alveolar spaces thus worsening the problem.



Reuben Ramphal, Robert C. Cogliano, Joseph W. Shands, Jr., and
Parker A. Small, Jr.

This paper reports studies showing the effects of serum antibody

upon influenza infection at two different sites: the trachea and lung.

Tracheal desquamation, pulmonary consolidation, death, and virus

shedding were examined after infection of mice with a lethal A/Port

Chalmers/l/73(H3N2) influenza virus. Immune serum administered intra-

peritoneally before infection prevented death and pulmonary consolidation

and also significantly lowered lung virus shedding as compared with

controls receiving normal serum. However, this protection did not;

extend to the ciliated epithelium of the trachea because serum antibody

did not prevent desquamation of the trachea or significantly decrease

viral yield from the trachea. These results indicate that serum antibody

is protective against severe pulmonary parenchymal disease but not for

disease of the ciliated epithelium.


Ada, G. L. and K. L. Yap. 1977. Immunochem. 14: 643-651.

Allan, W. H., C. R. Madeley, and A. P. Kendal. 1971. J. Gen. Virol.
12: 79-84.

Baltimore, D. 1971. Bact. Rev. 35: 235-241.

Bang, F. B. and J. S. F. Niven. 1958. Brit. J. Exp. Path. 39: 317-

Barber, W. H. and P. A. Small, Jr. 1974. Inf. Imm. 9: 530-536.

Barber, W. H. and P. A. Small, Jr. 1978. Inf. Imm. 21: 221-228.

Basarab, 0. and H. Smith. 1970. Br. J. Exp. Path. 51: 1-6.

Beare, A. S., D. A. J. Tyrrell, D. Hobson, C. H. L. Iowells, M. S.
Pereira, T. M. Pollock, and L. E. Tyler. 1969. J. Hyg., Camb.
67: 1-9.

Biddison, W. E., S. Shaw, and D. L. Nelson. 1979. J. Imm. 122: 660-

Bienenstock, J., R. L. Clancy, and D. Y. E. Perey. 1976. In: Lung
Biology in Health and Disease. Eds. C. H. Kirkpatrick and H. Y.
Reynolds, Vol. I, p 29-58.

Braciale, T. J. 1977. J.E.M. 146: 673-689.

Braciale, T. J. 1979. J.E.M. 149: 856-869.

Cambridge, G., J. S. Mackenzie, and D. Keast. 1976. Inf. Imm. 13:

Cantor, H. and E. A. Boyse. 1975. J.E.M. 41: 1390-1399.

Craighead, J. E. and B. J. Brennan. 1968. Amer. J. Path. 52: 287-300.

Cretescu, L., A. S. Beare, and G. C. Schild. 1978. Inf. Imm. 22:

Crosby, A. W. 1976. Epidemic and Peace 1918. Greenwood Press.

Doherty, P. C., R. B. Effros, and J. Bennink. 1977. Proc. Natl. Acad.
Sci. U.S.A. 74: 1209-1213.

Ennis, F. A., M. A. Wells, G. M. Butchko, and P. Albrecht. 1978.
J.E.M. 148: 1241-1249.

Effros, R. M., P. C. Doherty, W. Gerhard, and J. Bennink. 1977. J.E.M.
145: 557-567.

Fazekas de St. Groth, S. and M. Donnelley. 1950a. Austral. J. Exp.
Biol. 28: 45-60.

Fazekas de St. Groth, S. and M. Donnelley. 1950b. Austral. J. Exp.
Biol. 28: 61-75.

Finkelstein, M. S., M. McWilliams, and C. G. Huizenga. 1972. J. Imm.
108: 183-194.

Francis, T. 1943. Science. 97: 229-235.

Greenberg, S. B., B. S. Criswell, H. R. Six and R. B. Couch. 1977.
J. Imm. 119: 2100-2106.

Hay, A. J. and J. J. Skehel. 1979. Brit. Med. Bull. 35: 47-50.

Henney, C. S. and R. H. Waldman. 1970. Science. 169: 696-697.

Heuschele, W. P. and B. C. Easterday. 1970. J. Inf. Dis. 121: 486-

Jennings, R., R. J. Fenton, M. G. McEntegart, and C. W. Potter. 1978.
Med. Microbiol. Immunol. 166: 43-50.

Jurgensen, P. F., G. N. Olsen, J. E. Johnson, III, E. W. Swenson, E. M.
Ayoub, C. S. Henney, and R. H. Waldman. 1973. J. Inf. Dis. 128:

Kantorovich-Prokudina, E. N., N. P. Semyonova, S. S. Yamnikova, 0. N.
Berezina, and V. M. Zhdanov. 1978. J. Gen. Virol. 41: 405-409.

Klein, J. D. and A. M. Collier. 1974. Inf. Imm. 10: 883-888.

Loosli, C. 1953. In: Transactions of the Association of American
Physicians, Vol. 66, p 222-230.

Lucas, S. J. and D. W. Barry. 1977. Abstr. ASM. Abstr. #S232, p 317.

McLaren, C., C. W. P6tter, and R. Jennings. 1974. Med. Microbiol.
Immunol. 160: 33-44.

Melnick, J. L. 1977. Prog. Med. Virol. 23: 196-201.

Mendenhall, W. 1975. Introduction to Probability and Statistics.
Duxbury Press, Duxbury, Mass.

Morbidity and Mortality Weekly Report. 1979. 28: 231-239.

Morris, J. A., J. A. Kasel, M. Saglam, V. Knight, and F. A. Loda.
1966. New Eng. J. Med. 274: 527.

Mostow, S. R. and D. A. J. Tyrrell. 1973. Arch. Ges. Virusforsch.
43: 385-392.

Murphy, B. R., J. A. Kasel, and R. M. Chanock. 1972. New Eng. J. Med.
286: 1329-2332.

Naficy, K. 1963. New Eng. J. Med. 269: 964-966.

Potter, C. W., S. L. Shore, C. McLaren and C. Stuart-Harris. 1972.
Br. J. Exp. Path. 53: 168-179.

Ramphal, R., R. C. Cogliano, J. W. Shands, Jr., and P. A. Small, Jr.
1979. Inf. Imm. 25: 992-997.

Rasmussen, A. F. 1964. In: Newcastle Disease Virus. Ed. R. P.
Hanson, p 313-325.

Reed, L. J. and H. Muench. 1938. Amer. J. Hyg. 27: 493-506.

Richman, D. D., B. R. Murphy, E. L. Tierney, and R. M. Chanock. 1974.
J. Imm. 113: 1654-1656.

Ritchey, M. B., P. Palese, and E. D. Kilbourne. 1976. J. Virol. 18:

Roszt6czy, I., C. Sweet, G. L. Toms, and H. Smith. 1975. Br. J. Exp.
Path. 56: 322-328.

Rott, R., H. Becht, and M. Orlich. 1974. J. Gen. Viol. 22: 35-41.

Schulman, J. L., Manijeh Khakpour, and E. D. Kilbourne. 1968. J.
Virol. 2: 778-786.

Schmidt, R. C. and H. F. Maassab. 1974. J. Inf. Dis. 129: 637-643.

Schmidt, R. C., H. F. Maassab, and F. M. Davenport. 1974. J. Inf.
Dis. 129: 28-36.

Sever, J. L. 1962. J. Imm. 88: 320-330.

Siegel, S. 1956. Nonparametric Statistics for the Behavioral Sciences.
McGraw-Hill, New York.

Small, Jr., P. A., R. H. Waldman, J. C. Bruno, and G. E. Gifford. 1976.
Inf. Imm. 13: 417-424.

Stuart-Harris, C. 1973. Med. J. Aust. Special Suppl. 1: 42-46.

Sullivan, J. L., R. E. Mayner, D. W. Barry and F. A. Ennis. 1976. J.
Inf. Dis. 133: 91-94.

Suzuki, F., J. Ohya, and N. Ishida. 1974. Proc. Soc. Exp. Biol. Med.
146: 78-84.

Virelizier, J. 1975. J. Imm. 115: 434-439.

Virelizier, J. L., A. C. Allison, J. S. Oxford, and G. C. Schild. 1977.
Nature. 266: 52-53.

Waldman, R. H., S. H. Wood, E. J. Torres, and P. A. Small, Jr. 1970.
Amer. J. Epidem. 91: 575-584.

Waldman, R. H. and C. S. Henney. 1971. J.E.M. 134: 482-494.

Waldman, R. H. and W. J. Coggins. 1972. J. Inf. Dis. 126: 242-248.

Waldman, R. H., P. F. Jurgensen, G. N. Olsen, R. Ganguly, and J. E.
Johnson, III. 1973. J. Imm. 111: 38-41.

Webster, R. G. and W. G. Laver. 1975. In: The Influenza Viruses and
Influenza. Ed. E. D. Kilbourne, p 269-314.

Webster, R. G. and V. S. Hinshaw. 1977. Inf. Imm. 17: 561-566.

Wyde, P. R. and T. R. Gate. 1978. Inf. Imm. 22: 423-429.

Yap, K. L. and G. L. Ada. 1977. Immunology. 32: 151-159.

Yap, K. L. and G. L. Ada. 1978. Scand. J. Immunol. 7: 73-80.

Yap, K. L., T. J. Braciale, and G. L. Ada. 1979. Cell. Imm. 43: 341-

Zweerink, H. J., S. A. Courtneidge, J. J. Skehel, M. J. Crumpton, and
B. A. Askonas. 1977. Nature, Lond. 267: 354-356.


Robert Christopher Cogliano, son of Vincent and Mary, was born

in Boston, Massachusetts,on the third day of August, 1946. He was

raised with his 5 brothers and sisters in Pembroke, Massachusetts,and

attended college at Bridgewater State College, Bridgewater, Massa-

chusetts. Robert graduated in June of 1968 and entered military

service February, 1969. He served as a pilot in the United States

Air Force for five years. Robert started his graduate studies in

September, 1974, and will now go on to a position at New York Uni-

versity Medical Center.

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

/ /.

/ Parker A. Small, Jr Chairman
SProfessor of Immunology and Medical

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

L. William Clem
Professor of Immunology and Medical

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

George E. Gifford
Professor of Immunology and Medical

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

Paul A. Klein
Associate Professor of Pathology

This dissertation was submitted to the Graduate Faculty of the College of
Medicine and to the Graduate Council, and was accepted as partial fulfill-
ment of the requirements for the degree of Doctor of Philosophy.

December 1979

Dean, Collbge of Medicine

Dean, Graduate School

/ 7q

IIH111111Illl11111llllll l 111111 11111 11111
3 1262 08554 3675