Alick lsaacs

Notionol Institute for Medico1 Research. Mill Hill. London. England

I. Introduction . . . . . . . . . . . . . . . . . 1
.
A Definition . . . . . . . . . . . . . . . . . 1
B. Viral Inhibitory Substances Recovered from Virus Infections . . . 2
.
C Techniques of Assaying Interferon . . . . . . . . . . 3
I1. Production by Different Cells and Viruses . . . . . . . . . 4
.
A Production by Cells of Different Animal Species . . . . . . 4
B. Production by Different Varieties of Cells . . . . . . . . 5
C. Production of Interferon by Inactivated Virus . . . . . . . 6
D . Production of Interferon by Live Virus . . . . . . . . . 7
E. Senstivity of Different Viruses to the Antiviral Action of Interferon . 8
F. Factors Concerned with the Production of Interferon . . . . . 9
.
111 Properties . . . . . . . . . . . . . . . . . . 10
.
A Physicochemical Properties . . . . . . . . . . . . . 10
.
B Biological Properties . . . . . . . . . . . . . . 12
.
IV Purification . . . . . . . . . . . . . . . . . . 14
V. Mode of Action . . . . . . . . . . . . . . . . 15
A. Site of Action in Virus Growth Cycle . . . . . . . . . 15
B. Conditions Required for Action of Interferon . . . . . . . 17
C. Mechanism of Action . . . . . . . . . . . . . . 18
D. Effects of Interferon on Cells . . . . . . . . . . . . 21
E . Production and Action of Interferon at the Cellular Level . . . . 23
.
VI Interferon and Recovery from Virus Infection . . . . . . . . 26
.
A Recovery from Virus Infection in Vitro . . . . . . . . . 26
.
B Recovery from Virus Infection in the Chick Embryo . . . . . 28
C. Recovery from Virus Infection in Adult Vertebrates . . . . . 29
.
D Virus Virulence . . . . . . . . . . . . . . . . 30
VII. The Role of Interferon in Normal Cells . . . . . . . . . . 33
.
VIII Interferon as a Possible Therapeutic Agent . . . . . . . . . 34
References . . . . . . . . . . . . . . . . . . 35

.
I INTRODUCTION
.
A Definition
Interferon is the name that was given to an antiviral substance pro-
duced by the cells of many vertebrates in response to virus infection .
It appears to be of protein or polypeptide nature. it is antigenically
distinct from virus. and it acts by conferring on cells resistance to the
multiplication of a number of different viruses.
1
2 ALICK ISAACS

B. Viral Inhibitory Substances Recooered from Virus Infections

Interferon derived its name from virus interference, since it was first
isolated and characterized during a study of this phenomenon (Isaacs
and Lindenmann, 1957). However, similar substances were previously
observed, although they were not characterized. @rskov and Andersen
(1938) found that within a short time of the initiation of vaccinia1 in-
fection of the rabbit skin local “antibody” could be detected at the site
of infection at a time when none could be found in the serum. In
retrospect, grskov (personal communication, 1962) feels that this was
interferon, not antibody, Card (1944)studied tissue immunity in mouse
encephalomyelitis and observed an interfering factor that was separable
from the virus. He found that a suspension of brain from mice infected
with Theiler’s mouse encephalomyelitis virus was able to inhibit the
growth of virulent virus in fresh mice; the inhibitory factor seemed to
be cell-bound and did not act by combining with the challenge virus.
Lennette and Koprowski (1946) found that infected cultures of chick
and mouse embryo tissue when freed of virus showed a very weak
viral inhibitory action which they thought could not explain the viral in-
terference they observed. Nagano and Kojima (1954) studied a similar
experimental situation to that of grskov and Andersen (1938) and they
also found a virus-inhibitory substance separable from the infecting
virus in extracts of infected rabbit skin. However, these authors were
unable to decide whether the inhibition found was an immunological
or an interference effect; indeed it is difficult in experiments in animals
to distinguish how much of the viral inhibitory action found might be
due to specific antibody, to inactivated virus or viral antigens capable
of inducing virus interference, or to interferon. Thus the recent experi-
ments of Matumoto et al. (1959) show that infection of mice with
neurotropic Rift Valley fever v i r u s protects them against the virulent
pantropic variant; protection is slight if the two viruses are injected to-
gether but increases the longer the interval between the two. Their
results suggest that viral interference may have played a more importarit
role in inducing protection when the two viruses were injected together,
but with lengthening interval of time between the two, specific immunity
may have become more important. To distinguish these it has become
necessary to carry out experiments in chick embryos or in tissue culture,
in order to exclude antibody, and to allow adequate characterization
of the virus inhibitory substances found.
In this review, no attempt will be made to summarize studies on &US
interference which have been thoroughly covered by Henle (1950),
Schlesinger ( 1959), and Wagner (1960). Nor will any attempt be
 INTERFERON 3

made to evaluate the role of interferon in virus interference beyond

drawing attention to the evidence described by Henle et al. (1959) and
Isaacs (1959) that many examples of virus interference can be accounted
for by the production of interferon by cells in response to contact with
the interfering virus. Although work on interferon began as an attempt
to find an explanation of viral interference, an early observation was
that once formed, interferon was rapidly liberated from cells and could
be found in much higher concentration in the extracellular fluid than
within cells (Isaacs and Lindenmann, 1957; VilEek, 1961; Bader, 1962).
This raised the possibility that interferon was capable of protecting not
only the cells initially infected but also neighboring cells. Thus, atten-
tion was soon directed toward considering the possible role of interferon
in cellular resistance to virus infection, in general, and in the processes of
recovery from virus infection in particular. These themes will therefore
be dealt with in Section VI of this review in place of a consideration
of virus interference.
C. Techniques of Assaying Znterferon
Many different techniques are used to assay interferon. The most
generally used, however, measure the degree of inhibition in the ability
of treated cells to produce virus after infection. This is measured either
as a diminution of the yield of virus from treated cells or as a diminution
in the abililty of treated cells, when infected, to initiate the production
of a viral lesion (e.g., a plaque in cell monolayers). The technique
which was first used (Isaacs and Lindenmann, 1957) was to measure
the reduction in the yield of influenza virus hemagglutinin from pieces
of chick chorioallantoic membrane infected in vitro by the technique
of Fulton and Armitage (1951). This method was based on the finding
of a linear relationship between the degree of virus inhibition produced
and the concentration of interferon used (Isaacs et al., 1957). This type
of method has now been largely replaced by a plaque assay method in
which the concentration of interferon that will produce a 50%reduction
in the plaque count in a cell monolayer is measured (Wagner, 1960). In
assays with Merent experimental systems a linear relation has been
found between the degree of reduction of the plaque count and the
logarithm of the concentration of interferon over quite a wide range
of concentrations, so that the end point of the assay can, if necessary, be
determined by interpolation. Gifford et a2. (1963) have developed an
assay of this kind based on the method of Postlethwaite (1960) for
producing plaques with vaccinia virus without using an agar overlay.
When the logarithm of interferon concentration was plotted against
the reduction in plaque count, an S-shaped curve was formed which
1 ALICK ISAACS

was linear over a certain range of concentrations of interferon. They

also found a linear relationship when the relative average plaque
diameter or the total plaque area was plotted against the logarithm
of interferon concentration. A third method of assay is based on the size
of the zone of protection produced in a sheet of virus-infected cells
when a cup containing interferon is placed over the agar overlay.
Porterfield (1959) showed that there was a linear relationship between
the concentration of interferon, plotted logarithmically, and the area
of the protected zone. Another technique used is to measure the degree
to which a culture of cells is protected against the cytopathic action of
virus as judged microscopically ( Sellers and Fitzpatrick, 1962). This
assay gives a linear relationship between the logarithm of the concentra-
tion of interferon and the logarithm of the amount of virus inhibited.
Sueltenfuss and Pollard (1963) have developed a very sensitive assay
which is based on inhibition of the development of the inclusions
produced by psittacosis virus, as judged by fluorescence microscopy of
cells stained with acridine orange. These are the basic methods most
commonly used in assaying interferon; the review by Porterfield (1963)
gives a more detailed description of the techniques used.
11. PRODUCTION BY DIFFERENT CELLSAND VmusEs
Production of interferon was studied first in chick cells infected
with inactivated influenza virus, but it soon became clear that similar
substances were produced by the cells of many animal species in
response to infection with a variety of different viruses.
A. Production by C e h of Different Animal Species
Among the animal species whose cells have been shown to produce
interferon in vitro are chickens, ducks, mice, rats, guinea pigs, hamsters,
rabbits, ferrets, dogs, sheep, pigs, cows, monkeys, and man. Table I
of the review by Ho (1962b) gives many references to work describing
production of interferon by different cell-virus systems. Production
during the course of infection in uivo has been demonstrated in chick
embryos, mice, and rabbits, but has been much less studied than produc-
tion in uitro.
The fact that birds produce interferon raises the question of how
early in evolution such a mechanism might have arisen. Virus inter-
ference has been found among bacterial and plant viruses but it is not
known whether it is mediated by substances similar to interferon, al-
though one report has appeared indicating that an interferon-like
substance was produced by Pseudonumas aemgirwsa infected with
 INTERFERON 5

bacteriophage (Mercer and Mills, 1960). The nature of the repressor

that is responsible for some cases of immunity to superinfection shown
by lysogenic bacteria (Jacob, 1959) is not yet known, but the fact
that the immunity tends to be specific toward the infecting phage does
not favor the suggestion that the repressor might function in the same
way as interferon.
B. Production by Diferent Varieties of Cells
No systematic study has been made of the production of interferon
by cells from different organs, but no striking differences in the behavior
of cells have been found in in uitro or in uiuo studies. Thus, in uiuo,
production of interferon has been observed in the mouse brain and
lungs and in the rabbit skin, and in uitro, in chick chorionic and
allantoic cells, human amnion cells, calf, dog, monkey, and human kidney
cells, human thyroid cells, and human leucocytes. Until now, no dif-
ferences have been observed between the behavior of epithelial cells or
fibroblasts.
Certain lines of tumor cells were thought at first to be poor producers
of interferon (e.g., Henle et al., 1959) but this may be due to the fact
that many tumor cell lines are very insensitive to the antiviral action
of interferon, even to that produced in the same cells. Thus, Ho and
Enders (1959a,b) found that HeLa cells produced interferon which they
could assay on primary human amnion cells but not in HeLa cells.
Similar findings were reported for KB cells by Chany (1961), for HeLa
cells by VilEek ( 1962), and for a human amnion cell line by Mayer
(1962).However, this is not an invariable finding since Cantell ( 1981a)
and Isaacs et al. (1961b) have found that certain lines of HeLa cells
show some sensitivity to the action of interferon, although less than
that of primary human thyroid cells, in the case of one cell line studied.
It was suggested by Isaacs et al. (1961b) that this behavior of tumor
cells might reflect metabolic differences from normal cells, and it would
be interesting to study this question in lines of HeLa cells differing
in sensitivity to interferon.
Embryonic cells have been used extensively to produce interferon,
but chorioallantoic cells of 6-day chick embryos were found to produce
only about one-tenth as much interferon as the cells of ll-day embryos
after treatment with irradiated influenza virus ( Isaacs and Baron, 1980).
Also suckling mice infected intranasally during the first day of life
with parainfluenza 1 (Sendai) virus produced more virus but less
interferon than did 4-week old mice similarly infected (Sawicki, 1961).
The question of whether it might generally be found that cells show
6 ALICK ISAACS

increased production of interferon with aging of the animal of their

origin, or aging f n vitro, requires further investigation.
C. Production of Interferon by Znuctivated Virus
First studies of interferon were carried out with inactivated myxo-
viruses. Among the viruses shown to produce interferon were influenza
A and B, Newcastle disease, and fowl plague viruses, inactivated by
irradiation with ultraviolet (UV)light, heating at 5 6 O or at 37OC., but
not by treatment with formaldehyde (Burke and Isaacs, 1958b). Other
viruses shown to induce the production of interferon when used in-
activated are mumps (Cantell, 1961a), Rous sarcoma virus (Bader,
1962), vaccinia ( Glasgow and Habel, 1962), and herpes simplex (Wad-
dell, 1962). Incomplete influenza virus, produced by repeated passage
at high virus concentration, has been shown to induce interference
(von Magnus, 1954) and to induce production of interferon when
inoculated on the chick chorion at a site where virus multiplication
does not occur ( Burke and Isaacs, l958a). Interferon induced by different
viruses shows no evidence of specificity, i.e., it is not most active when
tested against the homologous virus (Lindenmann et al., 1957).
Ho and Breinig (1962) have found that Sindbis virus heated at 56OC.
for 4 hours did not induce production of interferon but was able to
“sensitize” cells so that they now produced interferon when infected
with live Sindbis virus. A number of reports have appeared indicating
absence of interferon production by arboviruses and enteroviruses
when used inactivated (e.g., Ho and Enders, 1959b). With one arbo-
virus, inactivation by deoxycholate was found to produce a virus still
capable of inducing interference but no interferon could be detected
(Henderson and Taylor, 1961). However, the fact that interferon was
not detected makes it difficult to conclude that none was produced since
the conventional tests measure only excess interferon liberated from
cells. Before concluding that a virus once inactivated does not produce
interferon it will be necessary to examine different types of inactivation,
since it is known that if influenza virus is heated too much (Isaacs and
Lindenmann, 1957) or over-irradiated (Burke and Isaacs, 1958a), it
loses its ability to produce interferon. The results of Ho and Breinig
suggest that, at least with one virus, prolonged heating may have
reduced its ability to stimulate the production of good titers of extracel-
lular interferon while retaining its ability to sensitize cells to respond to
infection by live virus by producing interferon. Iduenza virus more
gently inactivated by heat was able to induce production of interferon
and was found to sensitize cells to respond to infection by live virus
by producing a rapid synthesis of interferon (Burke and Isaacs, 1958b).
Recently, VilEek (1963) has studied production of interferon in
 INTERFERON 7

chick cells induced by tick-borne encephalitis virus inactivated by

incubation for various periods of time at 37OC. He has concluded that
interferon production could be demonstrated only when live virus was
present. VilEek points to the fact that among viruses that have been
shown to induce production of interferon when used in the inactivated
form it has not yet proved possible to obtain infective viral nucleic
acid. Alternatively, in the viruses among which, until now, no clear
evidence of production of interferon by inactivated viruses has so far
been shown, it is readily possible to prepare infective viral RNA (ribo-
nucleic acid). This seems to be an interesting division among viruses,
although so far its significance is unknown.*
The findings quoted above concern production of interferon by
virus which has been rendered noninfective by a particular treatment.
The converse situation is infection by live virus of cells that are “insus-
ceptible,” implying that the cells are unable to support a complete
cycle of growth by a particular virus. Interferon production of this
kind has been found with influenza virus in chick chorionic cells
(Lindenmann et al., 1957) and by parainfluenza 1 and measles viruses
in human leucocytes (Gresser, 1961b). It seems clear, therefore, that
virus multiplication is not essential for production of interferon. The
question of which viral constituent stimulates cells to produce interferon
will be discussed in Section VII.
D. Production of Interferon by Live Virus
The term l i v e virus” is used to denote virus prepared in such a
way as to avoid as much as possible any loss of infectivity. However,
with animal viruses kept under optimal conditions, the majority of
the virus particles are incapable of initiating infection, the ratio of
infective particles to total virus particles being usually of the order
of 1 to 10. Since some strains of influenza virus grown in suspended
chick chorioallantoic membranes gave rise to good yields of interferon
within 6-12 hours of infection with inactivated virus, and poorer yields
of interferon at a later stage of infection with live virus, it is possible
that production of interferon by live virus is due largely to particles
in the virus population that are not undergoing multiplication. This
question cannot be resolved until methods are available for measuring
the yield of interferon from single cells. It is discussed further in
Sections II,F and VI.
The review by Ho (196213) gives in Table I a list of references to
production of interferon by different live viruses. Viruses shown to
’However, Gifford and Heller (1963)have now found good yields of interferon
on infecting chick cells with an arbovirus (Chikungunya virus) inactivated by in-
cubation for 23 hours at 35°C.
8 ALICK ISAACS

induce production of interferon include RNA and DNA (deoxyribo-

nucleic acid) viruses, all ranges of size from foot-and-mouth disease
virus (Dinter, 1960) to the pox viruses (Nagano and Kojima, 1958),
cytolytic viruses, e.g., arboviruses, and tumor viruses, e.g., polyoma
(Allison, 1961) . It seems justifiable to conclude, therefore, that produc-
tion of interferon is a very general response of cells to virus infection.
The yield of interferon differs greatly with different viruses grown in
the same cells or with a single virus grown in different cells. This is
discussed further in Section VI in relation to the problem of virus
virulence.

E . Sensitivity of Diflerent Viruses to the Antiviral Action of Intsrferon

In addition to the differences in the yield of interferon which they
can induce, viruses also differ in their sensitivity to the antiviral action
of interferon on cells. The two properties give the impression of being
related, since it is frequently found that viruses that give good yields
of interferon are sensitive to its antiviral action, and conversely, that
viruses that give poor yields of interferon are generally much less
sensitive to its antiviral action. It is not known whether there is any
necessary relationship between these two properties. Possibly tests of
the sensitivity of a virus to interferon measure indirectly the probability
that a particle belonging to a particular virus population will induce
the production of interferon instead of virus, in both normal cells and
interferon-treated cells.
Differences found in the sensitivity to the antiviral action of inter-
feron may be quite considerable. Thus roughly 30 times more inter-
feron was required to cause !%&inhibition of plaque production by
Newcastle disease virus than by O’nyong-nyong virus grown in chick
embryo fibroblasts ( Ruiz-Gomez and Isaacs, 1963a). An early observa-
tion was that herpes simplex virus was much more resistant to the
action of interferon than vaccinia or cow pox viruses grown on the
chick chorion (Isaacs et al., 1958). Ho and Enders ( 1959b) found that
herpes simplex virus was much more resistant to interferon than vac-
cinia or Sindbis viruses grown in human amnion or human kidney cells.
Relative resistance of herpes simplex was also observed by Cantell
and Tommila (1960) in the rabbit cornea and by VilEek and Rada
(1962) in chick embryonic cells, and the closely related pseudorabies
virus was shown to behave similarly by VilEek (1962) and by Dinter
and Philipson ( 1962). Adenovirus type 7 was found to be very resistant
to the action of interferon in HeLa cells (Cantell, 1961a). Viruses that
have been shown to be relatively resistant to the action of interferon
include strains of fowl plague, Newcastle disease, herpes simplex, pseudo-
 INTERFERON 9

rabies, and adenovirus. However, Glasgow and Habel (1962) found

herpes simplex virus relatively sensitive to the action of interferon in
mouse cells; it is not known whether this is due to the use of a different
strain of virus, or different'cells from those used by other workers.
As a general rule, vaccinia virus and many arboviruses and rhino-
viruses seem to be relatively much more sensitive to the antiviral action
of interferon (Baron et al., 1961; Sutton and Tyrrell, 1961) although
differences in sensitivity among the arboviruses can be shown. Dif-
ferences in sensitivity among different viruses have been related to
differences in oxygen requirement, in optimal temperature for virus
growth, and in virus virulence. These points are discussed in the
following sections since they may throw some light on the mode of action
of interferon.
F . Factors Concerned with the Production of Interferon
Production of good yields of interferon was observed within 6 hours
of infection of chick chorioallantoic membrane with heated influenza
virus (Isaacs and Lindenmann, 1957) or infection of human leucocytes
with parainfluenza 1 virus (Gresser, 1961b), which does not multiply
in these cells. Incubation at about 37OC. was required, incubation at
2OC. giving no significant yield of interferon. Production of interferon
continued for 24 hours when it gradually ceased, but a second inoculation
of heated influenza virus at this time gave rise to a second crop of
interferon ( Lindenmann et d.,1957). Irradiated influenza virus gave
a more long-lasting stimulus, production of interferon being detectable
in small amount even on the third day after infection (Burke and
Isaacs, 195813). The fact that protein synthesis is required for the
production of interferon is indicated by the inhibition of interferon
formation produced by treating cells with p-fluorophenylalanine ( un-
published observations).
Within a short time of being detected within cells interferon was
rapidly liberated and was recovered in good yield from the suspending
medium (Isaacs and Lindenmann, 1957). Interferon produced in
chorionic cells could be shown to diffuse not only outward from the
chorionic surface but also inward through the mesoderm to the allantoic
cells (Isaacs et al., 1958).This rapid liberation from cells first suggested
that interferon might be capable of protecting not only the cells initially
infected but also neighboring cells.
Early production and rapid liberation of interferon is characteristic
of infection with inactivated or nonmultiplying virus. However, following
infection by live virus of cells able to support virus multiplication,
interferon is usually detectable only after some delay. When chick
chorioallantoic membranes were infected with a large dose of influenza
10 ALICK ISAACS

virus, multiplication of virus occurred within the first 24 hours but no

interferon was detected, During the next 24 hours virus multiplication
slowed down and interferon was then produced (Burke and Isaacs,
1958b). Wagner (1980) has shown growth curves of influenza virus
cultivated in the chick embryo in which interferon appeared in the
allantoic fluid 24 hours after the production of viral hemagglutinin.
Production of interferon occurring 24-48 hours after virus production
was also observed for infection of the mouse brain with Onyong-nyong
virus (Hitchcock and Porterfield, 1981) and infection of chick cells with
tick-borne encephalitis virus (VilEek, 1961). This delay following infec-
tion with multiplying viruses contrasts with the early production of
interferon when inactivated or nomultiplying virus is used. One
possibility is that the delay allows time for virus inactivation to occur
and that the inactivated virus then sets off the production of interferon.
Alternatively, it is possible that an individual cell actively supporting
virus multiplication produces no interferon until a late stage of virus
multiplication is reached, when interferon may then accumulate and
help to bring virus production to a halt.

111. PROPERTIES

A. Phystcochemfcal Properties
Interferon is nondialyzable and not sedimented on centrifugation
at 100,OOO g for 4 hours (Isaacs et al., 1957). Estimates of its molecular
weight have been based on its rate of diffusion and its behavior on
centrifugation. Porterfield et al. (1960) measured the size of the zones
of protection produced in virus-infected chick cells to which were
applied at various intervals of time after infection beads containing
either chick interferon or viral antibody. The diffusion coefficient of
interferon was found to be much higher than that of rabbit antibody
and the molecular weight was estimated as less than 80,OOO. Burke
(1981) studied the behavior of purified chick interferon in the analytic
ultracentrifuge. The interferon behaved in the ultracentrifuge as a
single component with a molecular weight of 83,000 and a sedimenta-
tion constant of 4.77 S. Little work has been reported on the molecular
weight of interferons of other animal species, although they resemble
chick interferon in being nondialyzable and not sedimented at 100,OOO g
for periods of 1-2 hours.
Note added in proof: Recently, new information has appeared on the
molecular weight of interferon. Lampson et d.(1963) studied a highly
purified preparation of chick interferon and estimated, by means of
 INTERFERON 11

high-speed centrifugation, that it had a molecular weight of 20,000-

34,OOO. Rotem and Charlwood (1963) carried out studies of the molec-
ular weight of chicken, mouse, and monkey interferons by means of
sedimentation in sucrose density gradients along with radioactive-
labeled markers of known molecular weight. By use of t h i s technique,
all three interferons were found each to have a molecular weight close
to that of lysozyme with limits of 13,000-25,000. It seems likely from
these findings that the preparation studied by Burke cannot have been
purified sufficiently.
The protein, glycoprotein, or polypeptide nature of interferon is
inferred primarily from the fact that its antiviral activity is greatly
reduced or abolished by treatment with proteolytic enzymes, e.g.,
trypsin (Lindenmann et al., 1957), pepsin (Burke and Isaacs, 1958a),
or chymotrypsin (Wagner, 1960).On the other hand, it was not affected
by treatment with ribonuclease, deoxyribonuclease, or neuraminidase.
Some of its other physicochemical properties are those that might be
expected of a protein. According to Lampson d al. ( 1963), one unit of
interferon activity in an assay in chick cells was 0.0042 pg. of protein.
Interferon is stable on storage at 2O, -loo, or -7OOC. However,
the reports of the stability of interferon on heating have been very
conflicting. Chick interferon was inactivated on boiling for 5 minutes.
In an early report it was found to be inactivated on heating at 6OOC.
for 1 hour (Isaacs et al., 1957) but this result may have been due to
the pH not having been controlled. On heating at pH 7.2 to 7.4 it
resisted heating at 60°C. for 1 hour (Isaacs, 1960b). Wagner (1960j
found interferon prepared from chick allantoic fluid to resist heating
at 70% for 1 hour, and it is possible that other proteins present in
the allantoic fluid may stabilize the interferon to heat. Human interferon
was found to have its activity reduced but not abolished by heating
at 56OC. for 30 minutes (Ho and Enders, 1959a) and to be completely
inactivated by heating for 1 hour at 60°C. at pH 7.8 (Gresser, 1961a),
a finding which corresponds to our experience with human interferon.
On the other hand, Chany (1981) found human interferon to be
completely inactivated at 56OC. for 30 minutes, whereas Mayer (1962)
found it to be stable on heating at 60°C. for 1-2 hours. Rabbit interferon
was found to resist heating at 56OC. for 30 minutes but to lose activity
on heating at 85OC. (Nagano and Kojima, 1958). Mouse interferon was
found to be more heat-labile than chick interferon, being inactivated
by 60°C. for 1 hour (Henle et al., 1959; Isaacs and Hitchcock, 1960),
whereas Glasgow and Habel (1962) reported mouse interferon to be
stable after heating at 60°C. for 1 hour. In view of the biological dif-
ferences in interferons from different animal species discussed below,
12 ALICK ISAACS

it would not be surprising if they differed in heat stability too, as occurs,

for example, in the case of ribonucleases from different sources. HOW-
ever, some of the conflicting reports raise the question of the influence
of other constituents present along with the test materials on the ap-
parent heat stability of interferon.
Interferon is stable over a wide pH range, from pH 1-10 (Linden-
mann et al., 1957). It is also very stable on irradiation with UV light
(Burke and Isaacs, 1958a; Nagano and Kojima, 1958; Zemla and VilEek,
1961b).* It can be precipitated by saturated ammonium sulfate (Linden-
mann et d.,1957) or by acetone or ethanol (Zemla and VilEek, 1961a,b).
Its reported behavior with ether seems to be variable.
Most of the reported investigations have been concerned with chick
interferon. More investigation is required to know whether interferons
from other animal species have similar physicochemical properties.

B. Biological Properties
I. Antigenicity
Interferon is antigenically quite distinct from the virus that induced
its production (Isaacs et al., 1957). This is such a fundamental point
of distinction that it has been included in the definition given at the
beginning of this chapter.
Interferon appears to be a poor antigen. When inoculated into
rabbits or hens either alone or with oil adjuvants or after precipitation
with alum, chick interferon did not induce the production of neutralizing
antibody (Burke and Isaacs, 1960; Lindenmann, 1960) nor of precip-
itating antibody ( Belton, personal communication, 1960). Nagano and
Kojima (1960) found that a series of injections of rabbit interferon into
hens, guinea pigs, and two groups of rabbits produced no neutralizing
antibodies; however, a third group of rabbits developed neutralizing anti-
bodies as measured in the rabbit skin. Later Nagano and Kojima (1961)
confirmed this finding and also found neutralizing substances in the
serum of immunized fowls. Recently Paucker and Cantell (1962) have
found that after prolonged immunization of guinea pigs with mouse
interferon a very low-titered antibody was found. Antibody could be
demonstrated only by using very dilute preparations of interferon. As
far as the evidence goes, therefore, interferon appears to be a very
weak antigen.
The fact that interferon is quite distinct from virus serologically
allows the use of viral antibody to inactivate virus without affecting
* Lampson et al. (1963)do not find this to be true of highly purified chick inter-
feron.
 INTERFERON 13

interferon in materials containing both constituents. Other methods

that have been used for the same purpose are high-speed centrifugation,
treatment at low pH ( 1-2), heat, and UV irradiation. With each virus
that is tested it is necessary to be sure that the method of inactivation
employed is effective in removing all traces of infectivity, e.g., treatment
for 24 hours at pH 2 is effective with most myxoviruses but not with
poliovirus.
2. Species Specificity
The first observation of species specificity of interferon was that of
Tyrrell (1959), who found that calf and chicken interferons were much
more active when tested in cells of the homologous than the heterologous
animal species. Subsequently, species specificity has been found between
interferons in chick and rabbit cells (Isaacs and Westwood, 1959a),
chick and human cells (Ho and Enders, 1959b), and even chick and
duck cells (Wagner, 1961). The species specificity is not absolute, the
general finding being that interferon is less effective when tested in
heterologous cells. However, even this depends on the technique of
assay employed; an insensitive assay made it appear that monkey inter-
feron was much more active when tested against vaccinia virus in human
thyroid than in homologous cells (Isaacs et al., 1961b). As described
in Section II,B, many lines of tumor cells also produce interferons that
are more easily assayed in normal cells, even of heterologous species
(Chany, 1961), than in homologous cells.
Sutton and Tyrrell (1961) have tested interferons of a number of
animal species in cells of homologous and heterologous species. Rhesus
monkey interferon showed some antiviral activity when tested in calf,
human, rhesus, or cynomolgous monkey cells. Human interferon, on the
other hand, showed a greater species specificity. Monkey kidney inter-
feron was found to be active when tested in monkey or calf cells, but
calf kidney interferon was active in calf but not in monkey cells.
Curiously, Sellers and Fitzpatrick (1962) found just the reverse one-
way relationship between calf and monkey interferon. This raises the
question of whether the species specificity of interferon is an absolute
or a variable factor, which may depend on the preparation used, its
degree of purification, or other unknown factors. An absence of cross-
protection between chick and mammalian interferons was reported by
Pollikoff d al. (1962).
It is not known whether the species specificity depends on the uptake
of interferon by cells or on its behavior at an intracellular site. It is well
known that in vivo homologous antibodies are better taken up by cells
than antibodies from foreign animal species but it is not yet known
14 ALICK ISAACS

whether interferon behaves in a similar way. Recently, Gifford (1963a)

has found some preliminary evidence that would favor this interpretation.

3. Adsorption
In studies carried out in a tube assay, adsorption of chick interferon
to cells was found to be slow (Lindenmann d al., 1957) and similar
observations were made by Sellers and Fitzpatrick (1962) for monkey
interferon in a test-tube assay, However, Wagner (1961) found much
more efficient adsorption of chick interferon when it was applied in
very small volumes to monolayers of chick embryo cells. When 0.1 ml.
volumes were applied to sheets containing roughly 2 x lo7 cells, 75%
of the interferon activity was removed in 20 minutes, Thereafter ad-
sorption slowed and was not complete by 4 hours. The different results
found in these methods are probably due to more efficient absorption
occurring from a small volume of fluid.

IV. PURIFICATION
Although a number of investigations have been carried out in
different laboratories on various steps in the purification of interferon,
there was, until recently, only one published report on the purification of
interferon. This is a report by Burke (1961) describing stages in the
purification of chick interferon and some properties of the purified
material.
The starting material was crude chick interferon containing 150-
200 pg. of protein/ml., prepared by incubating chick chorioallantoic
membranes with UV-inactivated influenza virus in a buffered salt solu-
tion. This was first concentrated by precipitation with ammonium sulfate
followed by pressure dialysis. Dialysis against pH 2 buffer served as
a convenient sterilizing step and also caused precipitation of some
heavily pigmented material without loss of interferon activity. On testing
a number of different procedures it was found that after chromatography
on diethylaminoethyl (DEAE ) cellulose columns at pH 6.6 the eluate
could be dialyzed to pH 4.5 and run on to a previously equilibrated
column. Interferon was not retained by this column and the biological
activity was recovered quantitatively in the eluate. Three other compo-
nents which had been shown to be present by starch-gel electrophoresis
were now found to be retained by this column. The biologically active
eluate was next chromatographed at pH 5.8 when a single symmetrical
peak was obtained which gave a single band on starch-gel electrophoresis
at both pH 8.9 and pH 2.0. Examination in the ultracentrifuge also gave
a single component of molecular weight 63,OOO.
 INTERFERON 15

In this work, as well as in the study of certain physicochemical

properties mentioned in Section 111, it is not yet known how pure a
given preparation of interferon is, nor whether a particular property
under investigation might vary as a result of association of interferon
molecules with other substances present. With this reservation, the
preparation studied by Burke (1961) was found to be a protein con-
taining no nucleic acid and only small amounts of carbohydrate, i.e.,
1.6%;hexosamine 2.4%.Since it was retained by DEAE-cellulose columns
at pH 5.0 but not 4.5 its isoelectric point was between pH 4.5 and 5.0.
The degree of purification achieved was about eO-fold, materials con-
taining protein at a concentration of 6 pg./ml. showing good antiviral
activity. However, the total recovery of interferon was low. Hence it is
likely that the figure of 20-fold purification achieved may be an under-
estimate, since it was not possible to be sure how much inactivation
of the biological activity of interferon occurred during purification. At
the moment there is no way of knowing whether low recoveries of the
biological activity of interferon during purification are due to inactiva-
tion or to loss of material by coprecipitation with other substances
present.
Note added in proof: Recently, Lampson et al. (19f33) have published
a detailed account of the purification of chick interferon obtained from
embryonated eggs infected with influenza A virus. The technique in-
volved ( I ) precipitation with perchloric acid to remove virus and ex-
traneous proteins, ( 2 ) concentration and purification by precipitation
with zinc, (3) column chromatography on carboxymethylcellulose,
followed by ( 4 ) zone ionophoresis on pevikon. One unit of interferon
activity corresponded to 0.0042 pg. of protein, a considerably higher
degree of purification than had been achieved hitherto.

V. MODEOF ACTION

A. Site of Action in Virus Growth Cycle

At an early stage of the work it was clear that interferon acted by

rendering cells resistant to virus multiplication and did not act on
extracellular virus. This could be shown by the fact that in an assay
in chick membrane fragments interferon exerted its full effect only
when it was incubated with cells for some hours before infecting with
challenge virus (Lindenmann et al., 1957). Absence of direct interaction
of interferon and virus was observed by Ho and Enders (1959a,b) and
Vilbk (1960).
16 ALICK ISAACS

In support of this conclusion are the findings of Isaacs and Burke

(1958) that cells treated with sufficient interferon to induce 95%inhibi-
tion of virus growth were able to take up either live or inactivated virus
which gave rise to good yields of interferon. Further support comes from
the findings of Wagner (1960, 1961) that, when studied in a single
cycle of v i r u s growth, interferon showed an antiviral action even when
applied to cells 2 hours after infection had been initiated. Wagner’s
results also showed that interferon does not act by affecting adsorption
or uptake of virus. Experiments which revealed no effect of interferon
on virus adsorption when tested directly are described by Wagner
(1960, 1961) and by Isaacs (1960a). The studies of Grossberg and
Holland (1962) on poliovirus demonstrate that interferon does not act
by inhibiting release of virus from cells. It seems clear, therefore, that
interferon inhibits virus replication at an intracellular sitk.
Further investigation suggested that interferon acts ,at an early stage
of virus growth. Thus it inhibited not only the prodpction of mature
virus particles but it also inhibited to a correspondin4 degree the syn-
thesis of cell-associated virus antigens such as the yaccinial hemag-
glutinin (Isaacs et aZ., 1958), the influenza1 nucleoprotein soluble
antigen, and the viral hemagglutinin (Burke and Isaacs, 1960). More
precise evidence comes from observations that show that interferon
inhibits the synthesis of viral RNA. De Somer et qZ. (1962) found
that in chick cells treated with interferon and infected with Western
equine encephalitis virus there was inhibition of the synthesis of viral
RNA and of mature virus. These results were confirmed by Ho (1962c),
who noted, in addition, that the synthesis of RNA was inhibited slightly
less than the synthesis of mature virus particles. This last finding sug-
gested that interferon inhibited indirectly the synthesis of viral RNA,
or alternatively, that it inhibited the synthesis of another viral constituent
in addition to viral RNA.
Inhibition of the replication of infective viral RNA was first observed
by Grossberg and Holland (1961, 1962) and has been confirmed by
others. In these experiments it is essential that a single cycle of virus
growth should be studied. If this precaution is omitted it is possible
that a substance under test might not inhibit the replication of infective
RNA but inhibit the multiplication of mature virus particles formed after
the first cycle of virus growth. In order to be sure that a single cycle
of virus growth was observed, multiplication of poliovirus RNA was
studied in chick embryo fibroblasts (Grossberg and Holland, 1961; Ho,
1961) or in the chick embryo (De Somer et d.,1962). Interferon was
found to inhibit the replication of viral RNA and showed no direct
action on the extracellular RNA (De Somer d al., 1962). These findings
 INTERFERON 17

suggest that interferon acts after the virus particle has been adsorbed
to and penetrated cells and after its protein coat has been removed,
but before its viral nucleic acid has been replicated.
It is not yet possible to say very much about the site of action of
interferon within the cell, although this point is mentioned briefly in
the following sections.
B. Conditions Required for Action of Znterferon
For the full action of interferon, some hours of incubation at tempera-
tures around 37OC. are required. This was shown by experiments in
which cells were allowed to adsorb interferon for 3 hours at 37OC. when
they were washed and then incubated for 21 hours at either 2O or 37OC.
before virus challenge. Less viral inhibition was found in the cells kept
at 2OC., suggesting that a metabolic process in the cells requiring some
hours’ incubation at 37OC. was required before the action of interferon
was fully established (Lindenmann et d.,1957). An essentially similar
result was found by VilEek and Rada (1962). These findings are not
easy to reconcile with the observations of Wagner that interferon shows
some antiviral action even when given 2 hours after the initiation of
virus infection. It may be that when large doses of interferon are used
only a short period of time is required to observe some antiviral action,
but that a longer time is required when small amounts of interferon
are assayed.
The duration of action of interferon seems to be very variable and
to depend greatly on the metabolic state of the cells. Isaacs and West-
wood (1959b) studied its duration of action by exposing chick cells
suspended in a maintenance medium to interferon on a single occasion,
infecting them with West Nile encephalitis virus, and observing the yield
of virus at daily intervals. Under these conditions the cells showed no
cytopathic effect and released no virus during a period of 11 days. How-
ever, when the medium was enriched with calf serum and chick embryo
extract the cells degenerated rapidly and released virus into the medium.
The explanation that was suggested was that cells kept in a maintenance
medium were unable to divide and retained sufficient interferon to inhibit
virus growth. However, on addition of nutrients cell division commenced
and the intracellular concentration of interferon fell below that required
to inhibit virus replication. An alternative explanation would be that
interferon has a greater antiviral action in resting cells than in metaboli-
cally active cells.
The cells used are a second factor and Sutton and Tyrrell (1961)
found that calf kidney and rhesus monkey kidney cells treated with in-
terferon on a single occasion showed some resistance to virus infection
18 ALICK ISMCS

for 5 days. Wagner (1981)found that chick cells treated with interferon
developed resistance to infection 4 to 5 times more quickly than they re-
gained susceptibility on further incubation in interferon-free medium. A
rapid regaining of susceptibility in chick embryo cells treated with in-
terferon was reported by Bader (1962); however, the particular assay
used measured susceptibility to Rous sarcoma virus, an assay which takes
roughly 7 days to read, so that Bader’s results do not s&m to be in
disagreement with those of other workers.
It is implicit in these conclusions that interferon does not replicate
and this has been demonstrated experimentally. At the same time it was
observed that no interferon could be recovered by disrupting cells that
have taken up relatively large amounts (Isaacs et al., 1957; Wagner,
1W)The . significance of this finding is not at present clear.
C . Mechanism of Action
Chick cells treated with interferon and then infected with either live
or irradiated influenza virus were found to give good yields of inter-
feron although their ability to support virus multiplication was greatly
inhibited (Isaacs and Burke, 1958). This suggested that the action of
interferon might be described as a redirection of the pathway of in-
fecting virus from the synthesis of virus toward the synthesis of inter-
feron. VilEek and Rada (1962)found that chick cells treated with inter-
feron and infected with tick-borne encephalitis virus showed inhibition
of the production of virus and interferon, but Ruiz-Gomez et al. ( 1963),
in studies with Chikungunya virus in chick cells, found essentially similar
results to those of Isaacs and Burke. The reason for these variable
results is not yet known. It may depend on whether, under the experi-
mental conditions studied, some virus multiplication is necessary before
interferon is produced. Alternatively the results may reflect whether the
interferon preparation used contains some inactivated virus, as suggested
by Ho and Breinig (19s2). This would seem to be an unlikely explana-
tion for influenza virus inactivated by pH 2 treatment, since this has the
effect of abolishing the hemagglutinating activity of the virus, and
presumably, therefore, its ability to adsorb to cells. The suggestion that
interferon treatment of cells redirects the pathway of the infecting virus
from the synthesis of new virus toward the synthesis of interferon may
be an expression in biological terms of the biochemical findings, de-
scribed below, that cells treated with interferon are able to grow and
divide and can presumably synthesize normal cellular proteins and
nucleic acids at a normal rate, and yet are unable to support the replica-
tion of viral nucleic acid.
 INTERFERON 19

Numerous hypotheses have been put forward to account for the

mechanism of action of interferon. Wagner (1960) thought that inter-
feron behaved like a basic protein which could combine with viral
nucleic acid once it was released from viral protein in the cell. The find-
ings of Burke (1961) on its isoelectric point and the evidence of De
Somer (1962) that interferon does not inactivate viral nucleic acid
when mixed with it directly would not favor this hypothesis, although
De Somer suggests that interferon may stimulate cells to produce a basic
protein. The author has been guilty of propounding from time to time
hypotheses among which is the suggestion that interferon inhibits an
oxidative process that supplies energy for virus synthesis.
This hypothesis was first suggested by the observation that chick em-
bryo cells treated with large doses of interferon showed a slight increase
in oxygen uptake and a greatly increased glycolysis (Isaacs, 196Ob). It
was di5cult to be sure that it was the interferon which was responsible,
but in favor of this conclusion a "mock" preparation prepared in the
same way, but omitting the virus, showed no such activity and the active
factor shared with interferon a similar heat stability, stability on treat-
ment at pH 2, and absence of dialyzability. Stimulation of glycolysis in
chronically infected cultures shown to be producing interferon was
observed by Green et al. (1958) and has since been observed by many
workers. Increased glycolysis was reported by Allison (1961) in mouse
cells treated with mouse interferon and by Gresser (1961~) in human
cells treated with human interferon. Levy et d. (1962) commented on
repeated observation of increased glycolysis produced by chick and
mouse interferons but found that this effect lacked the species specificity
found in the antiviral actions of these preparations. Zemla and Schramek
(1962a) found increased glycolysis produced both by a preparation of
interferon and by a mock interferon preparation lacking antiviral activity
and concluded that it was not the interferon that was stimulating
glycolysis. This is not a necessary conclusion since it is possible that a
number of different substances might stimulate glycolysis. However, at
the moment it is not clear whether the stimulation of glycolysis observed
by a number of workers is due to interferon, to some closely related
substance, or to unrelated substances (cf. Lampson et aZ., 1963). The
same reservation is required with regard to other metabolic changes
observed in cells treated with interferon. It will not be possible to decide
on the significance of these findings until they can be repeated with
highly purified preparations.
Substances that uncouple oxidative phosphorylation stimulate cells to
increased glycolysis while at the same time the oxygen uptake may be
 increased. The observation that dinitrophenol, an uncoupler of oxida-
tive phosphorylation, did not inhibit the growth of poliovirus in
HeLa cells when given in doses that inhibited virus multiplication in
normal cells (Gifford and Blakey, 1959) was made at about the same
time as Ho and Enders (195913) found that interferon was produced in
HeLa cells but had to be assayed in normal human cells since HeLa
cells were relatively insensitive to its antiviral action. These resemblances
between the actions of interferon and agents that uncouple oxidative
phosphorylation were added to when it was found that a virus that was
particularly sensitive to interferon was also more sensitive to the action
of four different uncoupling actions, whereas a virus more resistant
to interferon was also more resistant to the uncoupling agents (Isaacs
et al., 1961a). This is indirect evidence that interferon may act by un-
coupling oxidation from phosphorylation, but at the moment there is no
direct evidence on this point; if such a mechanism were operating it
would be necessarily at a localized site within the cell, possibly at a
nuclear site. The significance of the evidence of Zemla and Schramek
(1982b) that interferon inhibits virus growth in chick cells under
anaerobic conditions depends on how sure one can be that complete
anaerobiosis was obtained.
Most workers have found that virus growth in normal cells is poor
under low oxygen tensions and Baron et al. (1961) observed that
different viruses have different oxygen requirements, as judged by the
depth to which virus will grow in chick cells kept in a culture tube filled
with agar. Those viruses with the highest oxygen requirements were
the most sensitive to the antiviral action of interferon and the converse
was equally true (Isaacs et al., 1961b). In addition it was found that in-
creased oxygenation of cultures tended to diminish the antiviral action of
interferon, whereas reduced oxygenation had the reverse effect. This
might suggest that interferon was inhibiting an oxidative process that
supplied energy for viral synthesis; apparently viruses differ in their
oxygen requirements, those with the highest oxygen requirements being
most readily inhibited by interferon. Also in favor of this interpretation
is the finding that tumor cells and the cells of young embryos (see Sec-
tion I1,B) are less sensitive to the antiviral action of interferon than
normal cells. Tumor cells and the cells of young embryos are generally
less dependent on oxidative processes as a source of energy than normal
cells.
Mosley and Enders ( 1962) found that polioviruses grown in monkey
kidney cells had a bicarbonate requirement which vaned with the virus
strain. In general, avirulent strains multiplied very poorly when grown
in tubes plugged with cotton wool, which allowed accumulated carbon
 INTFZFERON 21

dioxide to escape, whereas the growth of virulent viruses was little

affected. Evidence was produced that the decreased rate of growth in
cotton-plugged tubes corresponded to the low plating efficiency found
by Vogt et d. (1957) with avirulent polioviruses grown with an agar
overlay containing a low bicarbonate concentration (the d marker). The
avirulent and virulent viruses used correspond fairly well, as discussed
in Section VI, with strains showing greater or lesser sensitivity to the
antiviral action of interferon. It appears, therefore, that certain virus
strains that are sensitive to the antiviral action of interferon have rel-
atively high oxygen and high bicarbonate requirements compared with
strains that are less sensitive to interferon. Gifford (1963b) has studied
the oxygen and bicarbonate requirements of certain viruses and has
found that both increased oxygen and bicarbonate have the effect of
increasing the depth to which a virus grows in a tube culture filled with
agar. The effects of oxygen and bicarbonate were found to be additive,
so that increased bicarbonate could not replace oxygen. It is possible
that an oxidative process required for virus synthesis may be dependent
not only on oxygen but also on bicarbonate or carbon dioxide.
If interferon were blocking the supply of energy needed for viral
synthesis it might be anticipated that with small doses of interferon,
once the antiviral effect had worn off, the viral RNA would be able to
resume its replication, so that delayed virus growth would be observed.
This has in fact been found by Mayer et al. (1961, 1962), De Somer
et al. ( 1962), and Grossberg and Holland ( 1962). A similar observation
has been made by E. Heller (personal communication, 1962) with regard to
polyoma virus. With larger doses of interferon virus growth is suppressed
and in vaccinia1 infections of the rabbit skin, for example, interferon
does not merely delay but prevents the appearance of viral lesions
( Isaacs and Westwood, 1959a). Whether cellular nucleases or other
mechanisms are responsible for the suppression of infection is not known,
Also, in agreement with the suggestion that interferon inhibits the supply
of energy, is the fact that it does not block the uptake of virus and the
release of nucleic acid from protein but that it inhibits the replication of
viral nucleic acid. This would be one of the &st steps requiring an
energy supply from the host cells. Unfortunately, it is not yet possible
to provide more than indirect evidence that would support this or any
other hypothesis on the mode of action of interferon.
D. EflBcts of Interferon on Cells
Three types of effects of interferon on cells have been studied: mor-
phological changes, alterations in the growth rate of cells, and bio-
chemical changes.
22 ALICg ISAACS

1. Morphological Changes
With most tissue culture systems an absence of any significant mor-
phological change has been noted in cells treated with interferon.
Wagner and Levy (1960)studied chick embryo fibroblasts protected by
interferon and infected with Eastern equine encephalitis virus. Cells
stained with acridine orange showed normal architecture and a distribu-
tion of DNA and RNA that was indistinguishable from that of normal
cells. Despite the viral infection mitotic figures were readily seen in cells
pretreated with interferon, The picture was in striking contrast to the
rapid degeneration found in cells that were infected with virus but not
treated with interferon.
One interesting morphological change in cultures of human amnion
cells treated with interferon was described by Gresser (1961~).Two to
three days after the introduction of interferon many of the normally
polygonal cells became fusiform and resembled whorls of fibroblasts, so
that recognition of the original cell type was difficult. The changes were
readily reversible, the cells resuming their normal appearance within 24
hours of removing the interferon from the medium. Only preparations
with antiviral activity showed this effect and the active factor resembled
interferon in many of its physicochemical characters. Treatment of
primary cultures of human kidney cells or of a continuous cell line
derived from human amnion cells produced no morphological changes.
2. Growth Rate of Cells Treated with Znterfmon
Baron and Isaacs (1962) found that cultures of primary human
thyroid cells treated with about one hundred 504; inhibitory doses of
interferon were resistant to the multiplication of vaccinia virus but were
nevertheless able to grow and divide normally. The interferon was
present from 24 hours after the cells were first seeded in tubes and the
cells formed a complete sheet at about the same time as the untreated
control cells. Paucker et al. (1962) made a careful study of the growth
rate of L cells kept in suspension for periods of 25 days and treated
with different amounts of interferon. Cells treated with 10 units of
interferon showed a very slight inhibition of growth rate. Cells treated
with 100 units showed slight inhibition for the first 8 days with steadily
increasing inhibition thereafter. Cells treated by continuous exposure
to 700 units showed almost total cessation of cell growth. When the
interferon was removed, even after prolonged contact with L cells, there
was a gradual recovery with resumption of the normal rate of growth.
Treatment of cells by a single exposure to 2000 units followed by
removal of the interferon led to a short-lived depression of cellular
 INTERFERON 23

multiplication which lasted for about 3 days after which the normal
rate of growth was resumed.
3. Metabolic Changes in Cells Treated with Interferon
The increased glycolysis and uptake of oxygen in cells treated with
interferon were referred to in Section V,C. Few other changes were
described and Levy et al. (1962) reported a number of negative findings
in their attempts to find biochemical changes in cells treated with
interferon. Recently, however, Levy and Baron (1963) have made some
interesting new observations.
Exposure of chick embryo fibroblasts to actinomycin D inhibits RNA
metabolism by about 90-95%, the metabolism being measured by uptake
of H3-uridine into phenol-released RNA. Treatment with interferon
blocked about 50-75%of the remaining RNA metabolism. If actinomycin-
resistant RNA metabolism is non-DNA-dependent then interferon appears
to inhibit that small fraction of RNA metabolism in normal cells. What
function this RNA has in normal cells is not yet known.
In cells infected with Sindbis virus, at 34 and 4$ hours after infection,
in the presence of actinomycin, virus-infected cultures took up much
more H3-uridine into RNA than did uninfected cultures. Presumably
much of this RNA synthesis can be attributed to the formation of viral
RNA and it was completely inhibited by inclusion of interferon in the
medium.
These results suggest that a non-DNA-dependent RNA synthesis is the
target for the action of interferon. It is possible that interferon acts
either on the supply of energy for this particular RNA synthesis or that
it inhibits the formation or action of a polymerase or other enzyme
required for the formation of this RNA.
E . Production and Action of Interferon at the Cellzrlar Level
An early observation was that the degree of virus inhibition produced
by interferon was, within certain limits, independent of the dose of
challenge virus, and depended only on the amount of interferon used
(Lindenmann et al., 1957; Lindenmann and GBord, 1963). Thus, a
given dose of interferon produced the same degree of viral inhibition
when the dose of challenge virus was vaned over a 100-fold range. This
result resembles that found by Fazekas de St.Groth and Edney (1952)
for viral interference produced by heated influenza virus. Ho (1962a)
has found that when chick cells were treated with a small dose of inter-
feron and then infected with vesicular stomatitis virus and the values
for multiplicity of virus input plotted against the number of cells re-
quired to produce one infectious center the proportion of protected
24 ALICK ISAACS

cells was greater at low virus inputs. This suggests that the protective
effect of interferon in terms of prevention of cell infection may be
overcome by large virus inocula. These results offer a possible explana-
tion of some factors in the development of a vinis plaque in cells treated
with interferon. When a low virus multiplicity is used the first cell
infected will have received a single virus particle. However, when this
cell produces large numbers of virus particles which infect its neighbors
those secondarily infected cells will receive a much higher dosage of
virus. This may help to explain why many plaques found in cell sheets
treated with a low dose of interferon show a relatively normal plaque
size and appearance.
Bellett and Cooper (1959) studied an interfering component, prob-
ably interferon, that was produced by chick cells infected with vesicular
stomatitis virus. When the concentration of interferon was plotted against
the logarithm of virus yield relative to the maximal yield, a negative
exponential relationship was obtained between dose of interferon and
cells remaining uninterfered. This result was found with relatively low
virus doses, i.e., multiplicity of 2.5, and it suggests that one particle”
of interferon per cell was sufficient to induce cell protection. However,
this one-hit curve does not tell us how many molecules of interferon
per cell must be present before one effective particle will be found.
Cooper and Bellett (1959) found that the total virus yield from inter-
feron-treated cultures was reduced by about the same factor as the
number of cells able to release virus. This implies an all-or-none response
of cells which would either produce a normal yield of virus or show an
absence of virus production.
These two conclusions have been confirmed by some workers and
disputed by others. Bader (1962) measured the yield of Rous virus
from chick embryo cells treated with different dilutions of interferon
and concluded that one unit of interferon was sufficient to inhibit
replication of a single particle of Rous virus. Ho (1962a) studied the
number of plaques produced by vesicular stomatitis virus in cells treated
with different dilutions of interferon. He did not find a linear response
except possibly at low doses of interferon. At high doses of interferon
there was a relatively decreased inhibitory effect. Lindenmann and
Giflord (1963) plotted the dose-response curve of the logarithm of
interferon concentration against plaque count with vaccinia virus, They
found an S-shaped curve which was linear over a range of interferon
concentrations, but which became flatter at high concentrations of
interferon. Possibly the result of experiments of this kind may depend
on whether an assay measuring virus yield or an assay measuring plaque
production is used.
 INTERFERON 25

Results similar to those of Cooper and Bellett (1959) on the all-or-

none response of cells to interferon were described by Ho (1961) for
inhibition of the development of poliovirus RNA in chick embryo fibro-
blasts. Later Ho (1962a), in studies with chick cells infected with
vesicular stomatitis virus, observed that the reduction in total virus
yield was greater than the reduction in the number of infective centers.
He also repeated his earlier experiments with poliovirus RNA and now
found that the ratio of virus yield to infective centers was usually lower
in interferon-treated than in control cultures. These findings suggested
that in addition to suppressing virus development in cells, interferon can
also lead to a reduced output of virus in certain cells. A similar conclu-
sion was reached by Wagner ( 1961), who observed a prolonged latent
period in cells treated with interferon. Wagner suggested that if a
reduced virus yield were to be attributed to a normal yield of virus
from a minority of the cells the release of virus should have occurred
at the same time as in the control cells. However, a delay in the appear-
ance of virus in cells treated with interferon was discussed in Section
V,C, and this was advanced to support the view that interferon acts
by inhibiting the synthesis of viral RNA, which can be resumed in
certain cases once the effect of interferon has passed off. This delay is
stressed in the study by Gifford et al. (1983) of the times at which
plaques appear in cells treated with interferon. Thus, the apparently
greater reduction in the yield of virus in proportion to the reduction
in the number of cell yielders described by Ho (1962a) may simply be
an expression of a delay in virus synthesis in a minority of cells and a
complete suppression of virus synthesis in the majority. Investigation
of samples taken at Werent times could settle t h i s question.
Cantell et al. (1962) found that L cells treated with interferon and
infected with vesicular stomatitis virus frequently showed the develop-
ment of viral antigen within the cells in the absence of production of
infective virus. However, Cantell et al. point out that vesicular stomatitis
virus has a toxic property for these cells and that it would be unwise
to generalize from the results with t h i s virus. There is evidence from
the work of Gresser (1961a) and Gresser and Enders (1962) that
interferon which protected against virus multiplication gave only very
weak protection of human amnion cells against the toxic effect, which is
not associated with virus multiplication, and is produced by Sendai or
Sindbis viruses. This is a very similar finding to that of Cantell et d.
(1962) and it recalls an earlier observation of Isaacs and Fulton (1953)
that viral interference induced by irradiated influenza virus was much
more effective in the allantoic than in the chorionic cells of the chick
chorioallantoic membrane. In the allantoic cells a complete cycle of
26 ALlCK ISAACS

influenza virus growth was found and this was readily inhibited, whereas
in the chorionic cells only an incomplete cycle of virus growth occurs
and this is much more resistant to viral interference. It seems important
in studies of the effects of interferon at the cellular level to distinguish
complete cycles of virus growth from incomplete cycles.
The results of a virus-cell interaction in a cell that has taken up
interferon will presumably depend on the amount of interferon taken
up, the stage in the virus cycle at which it is able to exert its action,
whether it has had sufficient time to exert its full antiviral action, or
whether its action has been reversed. It will depend, too, on the multi-
plicity of the infecting virus and possibly on the metabolic state or the
stage of division of the cell. More precise information will require studies
of individual cells.

VI. INTERFFBON
AND RECOVERYFROM Vmus INFECTION

Recovery from virus infection has been investigated in uitro, in the

chick embryo, and in mature animals. In the first two cases, cellular
factors in the recovery process can be studied free from any complications
of antibody production or delayed hypersensitivity, but study in mature
animals is a more complex problem, The factors involved in this last
case are considered in more detail in the review by Baron (this volume).

A. Recovery from Virus Infection in Vitro

Bang and Gey (1952) studied the growth of Eastern equine en-
cephalomyelitis virus in 13 established strains of rat cells. There was
great variation in susceptibility with extremes formed by a normal cell
strain and its specific malignant cell derivative. The normal cells were
resistant to virus growth, whereas the malignant cells were rapidly
destroyed with the production of large yields of virus. Between these
two extremes cell strains continued to support virus growth for periods
of several months. Destruction of cells in these cultures seemed to be
focal in nature and Bang and Gey (1952) suggested that inhibiting
factors might be present in the media used.
One possible explanation for cellular resistance found in chronically
infected cultures is that virus acts selectively, destroying the most
susceptible cells and favoring the growth of resistant cells, Bang et al.
(1957) found that chronically infected cultures could be cured of virus
infection by raising the temperature to 37OC., whereas lowering the
temperature to 31OC. favored maintenance of the chronic infection.
However, cultures that showed recovery from viral infection were con-
sistently susceptible to 3 successive reinfections, implying that the rela-
 INTERFERON 27

tive growth of resistant cells was not an explanation for the recovery
in this case. An alternative suggestion, discussed below, that interferon
might be playing a role in the behavior of these cultures, is in line with
Bang and Gey’s finding that tumor cells were more susceptible than
normal cells, since tumor cells are usually less sensitive to the antiviral
action of interferon than normal cells (see Section I1,B). Also, the poor
growth of virus at higher temperatures would fit with the observation,
discussed further below, that a rise in temperature may favor the pro-
duction of interferon.
Ho and Enders (1959a,b), Henle et a2. (lS59), and Mayer (1962)
studied chronic virus infections in vitro and observed the accumulation
of interferon in the media of these cultures. This raised the possibility
that interferon might be responsible for maintaining the state of chronic
infection. This question was posed by Glasgow and Habel (1962) in a
series of elegant experiments. They studied a mouse cell line that was
chronically infected with vaccinia virus and showed that it was possible
to stabilize the chronic infection, or to produce either complete cure of
the infection or total cell destruction. In order to make the infection take
one or other course all that was required was to raise or lower the con-
centration of interferon in the medium, or keep it relatively constant,
by changing the medium at different intervals of time. Treatment
of the culture with small doses of trypsin, which inactivates interferon,
had the same effect as frequent changing of the medium. Chany (1961)
had also found that in chronic infection of KB cells with parainfluenza 3
virus treating the cells with trypsin led to rapid cell destruction. In
passing, there would seem to be some contradiction in the suggestion
that chronic infections of certain tumor cells could be mediated by
interferon, since many tumor cell lines are relatively insensitive to the
antiviral action of interferon added to the medium. However, the low
sensitivity of tumor cells to exogenous interferon might be compatible
with a greater sensitivity to endogenous interferon produced at a site
close to its site of action.
An interesting observation from studies of cultures chronically in-
fected with a myxovirus was the finding that less than 10% of the cells
contained any evidence of the presence of virus, yet the whole culture
was resistant to challenge with vesicular stomatitis virus (Henle et al.,
1959). The significance of these results is emphasized by the finding of
Gresser and Enders (1962) that when two different kinds of human
amnion cells were mixed, one of these being sensitive to the destructive
action of Sindbis virus and the other being resistant, the mixed culture
showed resistance to virus infection. Even the presence of a minority of
resistant cells was able to protect the sensitive majority of cells. It was
28 ALICK ISAACS

found that on infection of the resistant cells large amounts of interferon

were produced which could then protect the susceptible cells.
It seems clear from this evidence that interferon plays an important
role in a number of examples of cellular resistance to virus infection
in vitro, as shown by recovery from infection or the maintenance of a
chronic state of virus infection. We can return here to the question
mentioned in Section I1,D-in a culture infected with ‘live” virus and
producing both new virus and interferon, is it likely that single cells
are producing both virus and interferon at the same time or is it more
likely that some cells are producing virus only at a given time, while
others are producing interferon only? The mixed cultures of Gresser and
Enders (1962) provide a model of a situation in which a population of
cells reacts to infection by one virus in two different ways. In this ex-
ample, production of interferon by a minority of cells has the effect
of protecting the whole culture. In the example provided by Henle et aZ.
(1959), the minority of cells were producing virus and it is possible
that the majority were producing interferon. In most virus infections, it
is not known how much variation there is in the response of individual
cells. However, study of the cellular changes induced in bovine cell
cultures by infection with high multiplicities of influenza virus shows
that only a proportion of the cells in infected cultures show pathological
changes detectable by staining with acridine orange or fluorescent
antibody or the presence of viral antigen as revealed by hemadsorption
(Niven et d.,1962). This recalls the finding of Magee and Sagik (1959)
that when chick embryo cells were infected with Newcastle disease
virus at a multiplicity of 5 or 10 only about 6O!Z of the cells could be
shown to yield virus. If it should prove to be correct that in a normal
virus infection, cells infected with virus particles that are unable to
multiply respond by producing interferon, whereas cells do not produce
interferon while they are supporting virus multiplication, such findings
would fit well with the observations discussed in Section II,F of the
rapid liberation of interferon from cells treated with inactivated virus,
in contrast with the delayed production by cells which are actively
supporting virus multiplication.
B. Recovery from Virus Infection in the Chick Embryo
It has been known for some time that the results of infecting chick
embryos with many different viruses depend to a large extent on the
age of the embryo at the time of infection (see Beveridge and Burnet,
1946). In general, susceptibility to the lethal action of many different
viruses decreases as the embryo ages. When it was found that both
sensitivity to the antiviral action of interferon and the ability to produce
 INTERFERON 29

p o d yields of interferon on vinis infection increase as the embryo

ages (Isaacs and Baron, 1960), an opportunity was presented to see
whether resistance to virus infection and development of the interferon
mechanism were related. Since the chick embryo does not produce
antibody and does not show delayed hypersensitivity, any relationship
of interferon to recovery from virus infection could be studied in
relative isolation.
When this comparison was made it was found that resistance to
infection with four different viruses and sensitivity to the antiviral
action of interferon in uitro were related to aging of the chick embryo
in a very similar manner (Baron and Isaacs, 1961) , Both of these factors
were low in embryos of under 7 days and both showed a sharp increase
between 7 and 10 days, followed by a much more gradual increase. An
increased resistance to the lethal action of mumps virus (Cantell, 1961b)
and a considerable reduction in the ability of chick embryos to support
the growth of poliovirus RNA (Denys and Prinzie, 1962) were also
found to begin at about the seventh or eighth day, These findings favor
the conclusion that the interferon mechanism plays an important role
in the ability of the developing chick embryo to recover from virus
infection.
C . Recotiery from Virus Infection in Adult Vertebrates
This subject is dealt with in detail in the article by Baron (this
volume), who has reviewed the evidence that recovery from virus infec-
tion cannot be accounted for solely in terms of production of antibody.
In this section a short summary of some points of interest with regard
to interferon will be given.
Production of interferon in the course of virus infections in tiitio was
observed by Nagano and Kojima (1958) in vaccinia1 infection of the
rabbit skin, by Isaacs and Hitchcock (1960) and by Link and Raus
(1961) in the course of influenza virus infection of the mouse lung,
by Hitchcock and Porterfield (1961) during infection of the mouse
brain with an arbovirus, and by Friedman et (11. (1962) during vac-
cinial infection of the skin of the guinea pig. The peak of interferon
production was found to occur early in infection, either at the time
of the peak of virus production or very shortly afterward, whereas the
peak of antibody production was much later.
In trying to assess the importance of the production of interferon on
the course of a virus infection in duo it was interesting to see what
effect inhibiting interferon production would have on the course of
the infection. One such inhibitor is increased oxygenation, which was
found to inhibit the antiviral action of interferon in uitro (Isaacs et aZ.,
30 ALICK ISAACS

1961b), a finding which is in line with the suggestion that interferon

acts by inhibiting an oxidative process (see Section V,C). Increased
oxygenation was found to show an adverse effect on the course of
influenza viral infection of mice in uiuo, as shown by a higher mortality
and a shortened incubation period of lethal pneumonia (Sawicki et al.,
1961). A second inhibitor of interferon production is cortisone, which
inhibited both the production and action of interferon in chick cells
(Kilbourne et al., 1961) but only its production in rat cells ( D e Maeyer
and De Maeyer, 1963). Cortisone has a detrimental effect on the course
of many viral infections, but since it may also affect antibody production
and delayed hypersensitivity, it is more difficult to assess how much
its detrimental action in virus infections is due to its effect on the
interferon response.
In many virus infections of animals resistance to a number of virus
infections increases with age. This was observed to be the case in
infection of mice with parainfluenza 1 (Sendai) virus. Sawicki (1961)
has shown that in the course of aging mice develop an increased ability
to eliminate this virus from their lungs and this is accompanied by
increased production of interferon.

D. Virus Virulence
Enders (1960) commented on the higher yield of interferon from
cells infected with an avirulent strain of measles than from cells infected
with a virulent strain. He suggested that this relationship might be a
more general one which could yield an interesting clue to the nature
of virus virulence. Since this first report, many examples of a similar
nature have been observed.
De Maeyer and Enders (1963) found that 5 strains of poliovirus of
low virulence induced production of interferon, whereas with 4 virulent
strains no interferon could be detected. Ruiz-Gomez and Isaacs (1963a)
studied production of interferon in chick embryo cells infected with a
variety of different viruses. In general it was found that viruses that
were most virulent for the chick embryo produced less interferon than
viruses of lesser virulence. If cellular susceptibility to virus infection
can be thought of as the mirror image of virus virulence it is of interest
that Glasgow and Habel (1962) observed that mouse cells that showed
lesser susceptibility to vaccinia virus produced more interferon than
cells that were more susceptible to the same virus. Again, Ruiz-Gomez
and Isaacs (1963b) noted that Newcastle disease virus grew well and
produced plaques in chick embryo cells in which, however, very low
yields of interferon were found. The same virus grew poorly in human
thyroid cells but produced large yields of interferon. Virus virulence
 INTERFERON 31

and interferon production thus share a characteristic in common, in

that they do not describe an isolated property of a virus, but only a
property of a virus in relation to a particular population of cells.
It is rather difficult to compare production of interferon by different
viruses since the yield of interferon depends on the dose of virus inoc-
ulated, the temperature of incubation, the pH, and other conditions of
culture. A measurement which seems to be less affected by these
variables is the sensitivity of a virus to the antiviral action of interferon.
It has been shown that there is quite a good correspondence between
the sensitivity to interferon of a number of viruses grown in chick cells
and virulence for the chick embryo, those viruses that are least sensitive
to interferon being the most virulent ( Ruiz-Gomez and Isaacs, 1963a).
Virulence therefore seems to be related to an ability of a virus either
to avoid stimulating the production of interferon or to be relatively
insensitive to the action of the interferon produced. One apparent
exception found was Kumba virus ( Ruiz-Gomez and Isaacs, 1963a),
which produced good yields of interferon and was sensitive to its
antiviral action in uitro, yet was highly virulent for the chick embryo.
However, recent investigations suggest that the virulence of Kumba
virus can be accounted for within the same theoretical framework by
postulating that this virus multiplies so rapidly that it is able to outstrip
its production of interferon. Indirect evidence favoring this interpretation
is provided by the fact that whereas most avirulent viruses, when inoc-
ulated on to chick cell monolayers at high virus doses show a prozone,
Kumba behaves like a virulent virus, such as Newcastle disease virus,
and does not show a prozone. The prozone seems to be due to the fact
that with large virus doses, sufficient interferon is produced early, to
inhibit the cytolytic action of the virus in these cells. With Kumba
virus, although good yields of interferon are produced, this presumably
occurs too late to inhibit the cytolytic action of the virus.
More convincing evidence of the relationship between virus virulence
and the interferon mechanism comes from studies of virus mutants of
differing virulence. Wagner (1962) compared the behavior in L cells
of mutants of vesicular stomatitis virus of differing virulence for mice.
The more virulent virus produced less interferon and was less sensitive
to the antiviral action of interferon in uitro than the less virulent virus.
A similar result was obtained by Finter (1962) with variants of Semliki
Forest virus a t different stages of adaptation to grow in calf kidney cells.
A most significant study is that of Thiry (1962) who prepared a
number of “ r e d mutants of Newcastle disease virus by treating virus
particles with nitrous acid. Red mutants give rise to “plaques” that are
intensely colored by neutral red and they show less virulence for chick
embryos and mice than the parent virus. Treatment with ethyl ethane
32 ALICK ISAACS

sulfonate had just the reverse of the effect of the treatment with nitrous
acid. These mutants could be arranged in increasing order of virulence
as measured by lethality after intracerebral inoculation of mice and
inoculation into the allantoic cavity of chick embryos. Thiry found
that the lower the virulence the greater the yield of interferon induced,
the two properties showing a very close correspondence.*
It is obviously not possible to account for virus virulence solely in
terms of the interferon mechanism. However, the above results would
suggest that in many examples of virus virulence that have been studied,
interferon production and action appear to play a significant role. A
question of interest that arose is whether a virulent virus is able to
avoid stimulating the production of interferon or is able to block
actively its production or action. Lindenmann (1960) found that produc-
tion of interferon by chick chorioallantoic membrane fragments, which
is readily induced by infection with heated or UV-irradiated influenza
virus, could be blocked by simultaneous infection with live virus. The
live virus could be given before, along with, or even 1 hour after the
inactivated virus and was still able to induce what Lindenmann called
inverse interference. In investigations of inverse interference in chick
embryo cells infected with an arbovirus, Ruiz-Gomez et al. (1963)
observed that viruses virulent for the chick embryo showed inverse
interference, whereas less virulent viruses did not. On the basis of
these and other findings the hypothesis was put forward that when a virus
particle enters a cell it either stimulates the production of interferon
and fails to multiply, or alternatively it inhibits the production of
interferon and proceeds to multiply. A number of cultural conditions
were mentioned which tended to favor one or other course. Apart from
the virus strain and the cells, raising the temperature, lowering the pH
or possibly the bicarbonate content (De Maeyer and De Somer, 1962),
lowering the oxygen tension, or pretreating the cells with interferon
all seeemd to favor the production of interferon relative to the produc-
tion of virus (Ruiz-Gomez et al., 1963). As discussed below, Heller
(1963) has found that minute doses of actinomycin D have just the
reverse effect.
The interpretation implied in these findings is that one aspect of
virus virulence is the ability of a virus to grow despite the normal
cellular defense mechanism, i.e., the production of an antiviral sub-
stance. Other interpretations of these findings are possible but the inter-
pretation proposed has the advantage of fitting logically with present
ideas of virus virulence.
* Sellers (1963) has now found that foot-and-mouth disease vinises of differing
virulence also show a corresponding variation in sensitivity to, and production of,
interferon.
 INTERFERON 33

VII. THE ROLE OF INTERFERON IN NORMALC E m

In trying to speculate on the possible role of interferon in normal
cells the assumption has been made that its antiviral action is incidental
to another, more basic action. On the basis of this assumption some
interesting results have emerged which would tend to support this point
of view. However, any understanding of the role of interferon in cells
in the absence of virus infection is still at a rudimentary stage.
One question which was of interest is what is the stimulus that
induces cells to make interferon. Since the production of interferon can
be initiated by RNA and DNA viruses, and in the absence of virus
multiplication, it seemed that virus protein or virus nucleic acid must
be the stimulus (since production of interferon can be induced by entero-
viruses containing only protein and RNA). The findings of Paucker
and Henle (1958) suggested that only virus containing nucleic acid
could induce interference and the results of Burke and Isaacs (1958a)
indicated that treatment with UV light, which damaged influenza viral
nucleic acid without significantly affecting its antigenic or neuraminidase
activities, abolished its ability to produce interferon. These results
focused attention on viral nucleic acid as the essential stimulus to make
interferon.
However, this still left unresolved the finding that both RNA and
DNA viruses were able to stimulate cells to make interferon. An
hypothesis was put forward that the essential stimulus to make interferon
might be a nucleic acid that was “foreign” to the cell (Isaacs, 1981).
This hypothesis was tested by treating chick and mouse cells with chick
and mouse RNA and infecting them with vaccinia virus. It was found
that the heterologous RNA showed a pronounced inhibition of virus
growth, whereas homologous RNA showed much less inhibition or an
absence of detectable inhibition. In addition, when 100 pg. of mouse
RNA was incubated with chick cells, the cells washed, and then incu-
bated with maintenance medium, the cells produced very small amounts
of an antiviral substance that differed from the mouse RNA but resembled
chick interferon in its properties (Rotem et d.,1963).* These results
favor the hypothesis that production of interferon represents a response
of cells to a foreign nucleic acid. One could speculate whether the
interferon mechanism might play a role in other situations in which
cells are exposed to a foreign nucleic acid. This could conceivably
occur in the reaction of the body to skin grafts and possibly to tumor
cells, which might both have nucleic acids that are foreign to the host
Cells.
Isaacs et aZ. ( 1963 ) have now produced additional evidence that production of in-
terferon occurs as a response of cells to a number of non-viral “foreign” nucleic acids,
34 ALICK ISAACS

Heller (1983) found that actinomycin D enhanced plaque produc-

tion by a number of RNA viruses that multiplied in the cell cytoplasm.
Such a result could be due to inhibition of the production or the action
of interferon. This was investigated and actinomycin was found to have
no measurable effect on the action of interferon. However, in doses
similar to those that enhanced virus growth actinomycin strongly in-
hibited production of interferon. Besides accounting for the enhancing
action of actinomycin, these results suggest that interferon production
is under the control of DNA-dependent RNA synthesis. This implies
that the genetic information for the production of interferon is normally
present in cells and that the effect of virus infection is to release a
mechanism that is normally suppressed.
The findings of Levy and Baron (1963) were referred to already in
Section V,D. Their results suggested that the target for interferon
action is non-DNA-dependent RNA synthesis. What function this
particular RNA synthesis might have is unknown, but its role in normal
cells, in embryonic cells, and in tumor cells should be a rewarding
subject for future research.
VIII. INTERFERON AS A POSSIBLE THERAPEUTIC AGENT
In addition to its antiviral action in vitro, administration of interferon
in uivo has been found to induce varying degrees of protection in animals
experimentally infected with certain viruses. Inhibition of the growth
of vaccinia virus in the rabbit skin was reported by Lindenmann et uZ.
(1957), Nagano and Kojima (1958), Isaacs and Westwood (1959a), and
Andrews (1961), and in the rabbit cornea by Cantell and Tommila
(1960). Some preliminary results on the effect of interferon on polyoma
virus infection in hamsters were reported by Atanasiu and Chany
( 1980). Protection of mice against infection with Bunyamwera virus
was found by Hitchcock and Isaacs (1960); this is of interest since it
represents protection against a systemic virus infection. Kaplan et ul.
(1982) found suggestive evidence that guinea pig interferon gave
slight protection of guinea pigs against infection with rabies virus.
The fact that interferon was active against a wide range of viruses,
showed low toxicity and low antigenicity, and appeared to play a role
in natural recovery from virus infections encouraged the idea that it
might be developed as a possible therapeutic agent in man. With this
end in view, a collaboration was set up in Great Britain between the
Medical Research Council and three pharmaceutical firms-Glaxo,
Wellcome, and I.C.I. Laboratories. This has resulted in the preparation
of monkey interferon which was used in a first clinical trial.
The trial carried out was an attempt to see whether interferon would
influence the effects of primary vaccination in man. Volunteers were
 INTERFERON 35

inoculated intradennally with interferon or control material given in

coded form. The following day the volunteers were vaccinated at both
sites and the vaccinia takes read “blind,” i.e., with no knowledge as to
which site had received interferon or control material. Interferon was
found to produce a highly significant degree of protection (Scientific
Committee on Interferon, 1962). Some interferon was also used to treat
a small number of patients with primary vaccinia1 keratitis and although
this was not a controlled trial encouraging results were reported (Jones
et al., 1962).
Further trials of interferon in local v i r u s infections, i.e., virus infec-
tions of the eye and in common colds, are now in progress. Whether
interferon will be found to benefit these conditions in a practical way
remains to be determined. Possibly a sounder approach may eventually
be to try to understand the nature of the stimulus to make interferon, so
that it might prove possible to stimulate the natural resistance of
individuals to virus infections.

REFERENCES
Allison, A. C. (1961). Virology 15, 47.
Andrews, H. D. (1961). Brit. Med. J . I, 1728.
Atanasiu, P., and Chany, C. (1960). Compt. Rend. Acad. Sci. 251, 1687.
Bader, J. P. ( 1962). Virology 16, 436.
Bang, F. B., and Gey, G. 0. (1952). Bull. Johns Hopkins Hosp. 91, 427.
Bang, F. B., Gey, G. O., Foard, M., and Minnegan, D. (1957). Virology 4, 404.
Baron, S., and Isaacs, A. (1961). Nature 191, 97.
Baron, S., and Isaacs, A. (1962). Brit. Med. J . I, 18.
Baron, S., Porterfield, J. S., and Isaacs, A. (1961). Virology 14, 444.
Bellett, A. J. D., and Cooper, P. D. (1959). J. Gen. Microbiol. 21, 498.
Beveridge, W. I. B., and Bumet, F. M. (1946). Med. Res. Council Spec. Relit. Ser.
256.
Burke, D. C. (1961). Biochetti. J. 78, 556.
Burke, D. C., and Isaacs, A. (1958a). Brit. J. Expll. Pathol. 39, 78.
Burke, D. C., and Isaacs, A. (1958b). Brit. J . Exptl. Pathol. 39, 452.
Burke, D. C., and Isaacs, A. (1960). Acta Virol. (Prague) 4, 215.
Cantell, K. (1961a).Arch. Ces. Virusforsch. 10, 510.
Cantell, K. (1961b). Adoan. Virus Res. 8, 123.
Cantell, K., and Tommila, V. (1960). Lancet ii, 688.
Cantell, K., Skurska, Z., Paucker, K., and Henle, W. (1962). Virology 17, 312.
Chany, C. ( 1961). Virology 13, 485.
Cooper, P. D., and Bellett, A. J. D. (1959). J . Cen. Microbiol. 21, 485.
De Maeyer, E., and De Maeyer, J. (1963). Nature 194,724.
De Maeyer, E., and De Somer, P. (1962). Nature 194, 1252.
De Maeyer, E., andEnders, J. F. (1963).In press.
Denys, P., Jr., and Prinzie, A. (1962). Virology 17, 216.
De Somer, P. (1962). Proc. Roy. SOC. Med. 55, 726.
De Somer, P., Prinzie, A., Denys, P., and Schonne, E. (1982). Virology 16, 63.
Dint-, Z. (1960). Acta Pathol. Micmbbl. S c u d . 49, 270.
36 ALICK ISAACS

Dinter, Z., and Philipson, L. (1962). Proc. SOC. Exptl. Bfol. Med. 109, 893.
Enders, J. F. (1960). Trans. Studies Coll. Phystcians Phtlu. 28, 68.
Fazekas de St.Groth, S., and Edney, M. (1952). J. lmmunol. 69, 160.
Finter, N. (1962). Personal communication.
Friedman, R. M., Baron, S., Buckler, C. E., and Steinmuller, R. I. (1962).J. Exptl.
Med. 118, 347.
Frilton, F., and Armitage, P. (1951). J. Hyg. 49, 247.
Card, S . (1944). Acta Med. Scand. 119, 27.
Gifford, G. E. (1903a). In preparation.
Gifford, G. E. (1963b). To be published.
Gifford, G. E., and Blakey, B. R. (1959). Proc. SOC. Exptl. B b l . Med. 102, 268.
Gifford, G. E., and Heller, E. (lQ63).To be published.
Gifford, G. E., Toy, S. T., and Lindenmann, J. (1963). Virology 19, 294-301.
Glasgow, L. A., and Habel, K. (1902). J. Exptl. Med. 115, 503.
Green, M., Henle, G., and Deinhardt, F. (1958). Virology I, 206.
Gresser, I. (1961a). Proc. SOC.Exptl. Blol. Med. 108, 303.
Gresser, I. ( l w l b ) . Proc. SOC. Exptl. Biol. Med. 108, 799.
Gresser, I. (1961~).Proc. Natl. Acad. Sci. U.S 47, 1817
Gresser, I., and Enders, J. F. (1962). Vtrology 18, 428.
Grossberg, S. E., and Holland, J. J. (1961). Federation Proc. 20, 443.
Grossberg, S. E., and Holland, J. J. (1962). J. Immunol. 88, 708.
Heller, E. (1903). Bbchem. J. 87, 18P.
Henderson, J. R., and Taylor, R. M. (1961). Vtrology 13, 477.
Henle, W. (1950). J. lmmunol. 84, 203.
Henle, W., Henle, G., Deinhardt, F., and Bergs, V. V. (1959). J. Exptl. Med. 110,
525.
Hitchcock, G., and Isaacs, A. (1960). Brit. Med. J. II, 1268.
Hitchcock, G., and Porterfield, J. S. (1961). Virobgy 13, 363.
Ho, M . (1961). Proc. SOC.Exptl. B b l . Med. 107, 039.
Ho, M. (1902a). Vlrobgy 17, 262.
Ho, M. (1962b). New Engl. J. Med. 266, 1258.
Ho, M. (1062~).Intern. Congr. Mfcrobbl., Bth, Montreal, Abstrs., p. 101.
Ho, M.,and Breinig, M. K. (1962). J . Zmmunol. 89, 177.
Ho, M.,and Enders, J. F. (1Q59a).Proc. Natl. Acad. Sd. U.S. 45, 385.
Ho, M., and Enders, J. F. ( 1959b). Virology 9, 446.
Isaacs, A. (1959). Symp. SOC. Gen. bkrobiol. 9,102.
Isaacs, A. (l960a). Symp. Sect. M4cpbioZ. N.Y.Acad. Med. 10, 182.
Isaacs, A. (1960b). Virology 10, 144.
Isaacs, A. (1961). Nature 192, 1247.
Isaacs, A., and Baron, S . (1960). Lahcet ii, 946.
Isaacs, A., and Burke, D. C . (1958). Nature 182, 1073.
Isaacs, A,, and Fulton, F. (1953). J. Gen. Mlcrobbl. Q, 132.
Isaacs, A,, and Hitchcock, G. (1960). Lancet ii, 69.
Isaacs, A., and Lindenmann, J. (1957). Proc. Roy. SOC. Ser. B 147,258.
Isaacs, A., and Westwood, M. A. (1959a). Lancet ii, 324.
Isaacs, A,, and Westwood, M. A. (195913). Nature 184, 1232.
Isaacs, A., Lindenmann, J., and Valentine, R. C. (1957). Proc. Roy. SOC. Ser. B 147,
268.
Isaacs, A., Burke, D. C., and Fadeeva, L. (1958). Brit. J. Erptl. Pathol. 39, 447.
Isaacs, A,, Klemperer, H. G., and Hitchcock, G. (1961a). Virology 13, 191.
 INTERFERON 37
Isaacs, A., Porterfield, J. S., and Baron, S. (1961b). Virology 14, 450.
Isaacs, A., Rotem, Z., and Cox, R. A. (1963). Luncet ii, 113.
Jacob, F. (1959).Hnruey Lectures Ser. 54, 1.
Jones, B. R., Galbraith, J. E. K., and Al-Hussaini, M. K. (1962). Lancet i, 875.
Kaplan, M. M., Cohen, D., Koprowski, H., Dean, D., and Ferrigan, L. (1962).
Bull. World Heulth Organ. 26, 765.
Kilbourne, E. D., Smart, K. M., and Pokomy, B. A. (1961). Nature 190, 650.
Lampson, G. P., Tyttell, A. A., Nemes, M. M., and Hilleman, M. R. (1963). Proc.
SOC. Exptl. Biol. Med. 112, 468.
Lennette, E. H., and Koprowski, H. (1946). J. Exptl. Med. 83, 195.
Levy, H. B., and Baron, S. (1963). To be published.
Levy, H. B., Glasgow, L. A., and Baron, S. (1962). Intern. Congr. Microbiol., 8th,
Montreal, 1962, Abstrs., p. 83.
Lindenmann, J. (1960). Z. Hyg. Infektionskrunkh. 146, 287.
Lindenmann, J., and Gifford, G. E. (1963). Virology 19,302.
Lindenmann, J., Burke, D. C., and Isaacs, A. (1957). Brit. J . Exptl. Puthol. 38, 551.
Link, F., and Raus, J. ( 1961). Nature 192,478.
Magee, W. E., and Sagik, B. P. (1959). Arch. Biochem. Biophys. 82, 340.
Matumoto, M., Nishi, I., and Saburi, Y. (1959). Compt. Rend. SOC. B i d . 153, 1845.
Mayer, V. (1962). Actu Virol. (Prague) 6, 317.
Mayer, V., Sokol, F., and VilEek, J. (1961). Actu Virol. (Prugue) 5, 284.
Mayer, V., Sokol, F., and Vilcek, J. (1962). Virology 16, 359.
Mercer, C. K., and Mills, R. F. N. (1960). J. Gen. Microbiol. 23, 253.
Mosley, J. W., and Enders, J. F. (1962). Virology 17, 252.
Nagano, Y., and Kojima, Y. (1954). Compt. Rend. SOC.Biol. 148, 1700.
Nagano, Y., and Kojima, Y. (1958). Compt. Rend. SOC. Biol. 152, 1627.
Nagano, Y., and Kojima, Y. (1960). Compt. Rend. SOC. BWZ. 154, 2172.
Nagano, Y., and Kojima, Y. (1961). Compt. Rend. SOC. Biol. 155, 1183.
Niven, J. S. F., Armstrong, J. A., Balfour, B. M., Klemperer, H. G., and Tyrrell,
D. A. J. (1962). J . Puthol. Bucteriol. 84, 1.
Qrskov, J., and Andersen, E. K. (1938).Acta Puthol. Microbiol. Scund. Suppl. 37, 621.
Paucker, K.,and Cantell, K. ( 1962). Vtrology 18,145.
Paucker, K., and Henle, W. (1958). Virology 6, 198.
Paucker, K., Cantell, K., and Henle, W. (1962). Virology 17, 324.
Pollikoff, R., Donikian, M. A., Padron, A., and Liu, 0. C. (1962). Proc. SOC. Exptl.
Biol. Med. 110, 232.
Porterfield, J. S. ( 1959). Lancet ii, 326.
Porterfield, J. S. ( 1963). In “Techniques in Experimental Virology” (R. J. C. Harris,
ed.), Academic Press, New York. In press.
Porterfield, J. S., Burke, D. C., and Allison, A. C. (1960). Virology 12, 197.
Postlethwaite, R. (1960). Virology 10, 466.
Rotem, Z., and Charlwood, P. A. (1963). Nature 198, 1086.
Rotem, Z., Cox, R. A., and Isaacs, A. (1963). Nature 197,584.
Ruiz-Gomez, J., and Isaacs, A. (1963a). Virology 19, 1.
Ruiz-Gomez, J., and Isaacs, A. (196313). Virology 19, 8.
Huiz-Gomez, J., Rotem, Z., and Isaacs, A. (1963). To be published.
Sawicki, L. ( 1961). Nature 192, 1258.
Sawicki, L., Baron, S., and Isaacs: A. (1961). Lancet ii, 680.
Schlesinger, R. W. ( 1959). In “Viral and Rickettsia1Infections of Man” (T. M. Rivers
and F. L. Horsfall, eds. ), pp. 145-155. Pitman Med. Publ., London.
38 ALICK ISAACS

Scientific Committee on Interferon. ( 1962). Lancet i, 873.

Sellers, R. F. (1983). Nature 198, 1228.
Sellers, R. F., and Fitzpatrick, M. (1962). Brit. J . Erptl. Pathol. 43, 674.
Sueltenfuss, E. A., and Pollard, M. (1963). Science 139,595.
Sutton, R. N. P., and Tyrrell, D. A. J. (1961). Brit. J. Exptl. Pathol. 42, 99.
Thiry, L. (1962). Intern. Congr. Microbiol., 8th, Montreal, 1962. Abstrs., p. 85 and
personal communication.
Tyrrell, D. A. J. (1959). Nature 184, 452.
VilEek, J. (1960). Nature 187, 73.
VilEek, J. ( 1961). Acta V i d . (Prague) 5, 278.
VilEek, J. ( 1962). Acta Virol. (Prague) 6, 144.
Vilcek, J. ( 1963). Acta Virol. (Prague) 7, 107.
VilEek, J., and Rada, B. (1962). Acta V i d . (Prague) 6, 9.
Vogt, M.,Dulbecco, R., and Wenner, H. A. (1957). Virology 4, 141.
von Magnus, P. (1954). Advan. Virus Res. 2,59.
Waddell, G. H. ( 1962). Ph.D. Thesis, Univ. of Miami.
Wagner, R. R. (1960). Bacteriol. Rev. 24, 151.
Wagner, R. R. (1961). Virology 13, 323.
Wagner, R. R. (1962). Cold Spring Harbor Symp.Quant. Biol. 27, 349.
Wagner, R. R., and Levy, A. H. ( 1960). Ann. N.Y. Acad. Scl. 88, 1308.
Zemla, J., and Schramek, S. (1962a). Acta Virol. (Prague) 8, 275.
Zemla, J., and Schramek, S . (1962b). Virology 16, 204.
Zemla, J., and VilEek, J. ( 1961a). Acta Virol. (Prague) 5, 129.
Zemla, J., and VilEek, J. (1961b). Acta Virol. (Prague) 5, 367.