Transcript from filmed lecture
November 15, 1959

Lecture #7 - Virus Genetics: Plant and Animal Viruses
by J. Lederberg

The viruses that I plan to discuss in this session have perhaps greater topical
interest than the ones attacking bacteria which were the subjects of previous lectures,
bacteriophage. The main utility of bacteriophage seems to be providing experimental
material for genetic analysis, previous attempts to use these particles for therapeutic
purposes having, for example, failed so far. There are, however, # number of types
of viruses which attack animals on the one hand and plants on the other, and these are
of great economic and medical significance. Unfortunately, in many respects, techni-
cal difficulties have hindered the collection of precise quantitative information along
directions that have been indicated in bacteriophage research.

One of the chief difficulties in dealing with the particles of viruses attacking
higher organisms, and especially those attacking higher plants, is the problem of ac-
curate quantitation of the number of infective units ina sample. For example, in or-
der to titrate the quantity of virus attacking the leaves of a higher plant, it's necessary
to rub a sample of the virus over the surface of the leaf, with the certain knowledge
that only a small fraction of the virus particles that might be present in that sample
will find susceptible cells in the leaf surface and be capable of penetrating them to give
what, in some cages, appears as a demonstrable lesion, Similarly, if one is dealing
with an animal virus, such as poliomyelitis, under circumstances requiring the use of
intact animale as the host for the propagation of this virus, the accurate determination
of the number of particles is likely to be an extremely expensive and time consuming

process. Some mitigation of this difficulty has been achieved in more recent years by
-Q-

the development of plating techniques for the production of plaques of animals viruses,
quite gimilar to the clearing plaques that were demonstrated for bacteriophage. For
example, it is possible to achieve an accurate titration of poliomyelitis virus by plat~
ing samples onto egar layers which have been seeded with cells from tissue culture
susceptible to this virus. And great progress has been made during the past very fow
years in the genetics of polio viras using techniques of this kind.

Unfortmately, for many other viruses on which one would like to get more pre- -
cise genetic information there is not yet a sufficient cytopathic effect of the virus on the
cells which are attacked by the virus to obtain detectable clearings on agar plates of
this sort. And for these one must still use much more laborious techniques of limit
dilution. For example, the influenza virus can be adapted to grow in the fluids, or
rather on the cells lining the fluid cavities of the chick embryo. It is possible to ob-
tain a titration of the number of influenza virus particles present in a given sample by
diluting that sample and innoculating small aliquots into each of a series of eggs, and
then after 48 hours, or some later time, harvesting the eggs and determining what
fraction of the eggs did contain a virus particle. Provided one is working at nearlimit
dilutions, that is to say at dilutions where the probability that any one sample will
contain a virus particle is low, it is possible to estimate the content of virus by the
fraction of eggs which yield detectable virus.

The same procedure can be used for the isolation of clones. That is to say,
under conditions of near-limit dilution of the sample, those eggs which do contain virus
are at least Hkely to contain a population of virus that had descended from a single particle,
and moat of the work on the genetics of the influenza virus, which has been the most ex-

tensively studied from a genetic point of view, bas been done on this basis.
-3-

Before taking up the specific questions of genetic recombination, I would like
to survey some of the general features of the life cycle of some animal, and to some
extent these are paralleled by plant, viruses, although one must admit in the first
place that there is a much wider variation in the characteristic behavior of the group
of viruses indicated here, and second that our detailed knowledge ig somewhat scant
in many respects.

One of the difficulties in dealing with the mammalian cell is that its boundaries
are to some extent less well defined than are the boundaries of a bacterial cell. The
bacterial cell in many cases has a very well defined rigid coat, which is responsible
for the characteristic shape that one observes in such cells. The mammalian cell is
a much more flexible object, its shape often determined by exterior forces acting on
it at the time, and one now knows that there is some ambiguity as to the precise mar-
gin as to the mammalian cell —- where it ends and where the medium in which the
cell is immersed begins. In fact, the outermost layer of the cell consists of a mucoid
coat which acts as a substrate for an enzyme found on the surface of the virus and
therefore constitutes a receptor for that virus. The structure of the influenza virus
can be. symbolized as an outer coat on which this viras-destroying and virus-reacting a
enzyme is found, and on inner matarial consisting of ribonnoleic acid which corresponds
to the genetic material of the bacteriophage.

The first step of virus adsorption to the mammalian cell is presumably the at~
tachment by means of the superficial enzyme particles to the mucoid coat, followed
eventnally by the penetration of the cell by the virus particle. ‘The details of this pene-
tration are not well understood, it has been suggested that it may consist simply of the

enguifment of the virus particle by the normal peendopodial activity of the cell to which
4
it is attached. It is a necessary condition for attachment of the virus that the mucoid coat
be intact, as can be demonstrated by stripping the coat by means of specific enzymes or
reacting substances such as periodate. Once inside the cell the virus particle enters an
eclipse phase similar to that already deseribed for the bacteriophage, although less is
known at the present time concerning the physical and chemical basis of this alteration
from an infective particle to the vegetative stage. During the subsequent several hours
of growth of the vegetative virus inside othe cell there is an increase of the ribonucleic
acid component, and this is followed by the gradual liberation from the cell, over a
period of considerable time, of intact particles. In the case of the influenza virus
there is some evidence that the virus particle does not receive its final coat except in
the act of emergence from the cell from which it is being Uberated -- in other words, |
that the coat of the influenza virus particle consists in part of cell-specific materials
already present in the host cell, in part of the materials elaborated in the host, as
a joint action of host genetic materials and the genetic material of the virus itself.
Whether the details of this picture can be applied equally well to other viruses is not
at all clear, and certain it is that there are considerable variations in the details with
which these must be applied.

We are now in a position to discuss some explicit experiments that have been
conducted originally by Burnet in Melbourne, and subsequently by investigators in this
country and in Great Britain, on the possibility of genetic recombination among strains
of the influenza virus. In order to accomplish this it was first necessary to secure a
number of useful markers, and thie has not yet been done in any very efficient way
through procedures of spontaneous mutation, although this has perhaps not yet been

fully exploited. Instead, Burnet has used a pair of strains designated MEL for Mel-
-5-

bourne and WSE, an egg-adapted strain, which pair of strains differ in respect of a
mumber of markers. We may indicate here 2 serological difference, the marker A
as against W; sensitivity to ovomucin, that is the capacity of the mucin found in egg
to inactivate the virus; and thirdly the path ‘ogenicity of the virus for the egg when
deposited on the chorio-allantoic membrane. In this case the small e refers to the
pathogenicity for the egg, the large E the non-pathogenicity characteristic of the
MEL strain.

When cells of the egg membranes sre infected at high multiplicity of infection
with mixtures of these two viruses, so as to insure that many cells will have received
an input of at least one virus particle of each kind, a number of interesting events
have been detected in the output from such mixedly infected cells.

Before proceeding to the question of genetic recombination it's necessary to
stress 3 somewhat different process which has been found to take place rather inde-
pendently of exchange of genetic material. This is phenotypic mixing, which refers
to the fact that although two quite distinct genomes are represented in the individual
particles which iesue from the doubly infected cell, that these particles show a common
neutrelizability by sntiserum. tn fact, in this case we have an anomdlous neutralization
where it proves to be the case, that the progeny from a mixedly infected cell can be
neutralized almost perfectly efficiently either by the anti-WSE, or by the and-MEL
serum. This, interpreted a6 representing the presence on the coats of both types of
virus particles, on the coats of virus particles containing either WSE or MEL genomes,
of substances which will react with the antibody which reaction results in the neutrali-

zation of the virus particle. However, most of the virus particles which issue from
such crosses retain their original individuality, or at least the intact individuality
of a WSE or of a MEL particle, and this phenomenon should not be confused with
genetio recombination.

A second finding which Burnet has reported and which is. based on the procedure
for isolation of clones of virus by the method of limit-dilution, is the occurrence of
particles that appear to be heterozygous. That is to say, what are nominally single
clones are found to contain mixtures of two genetic types, and this phenomenon is in
principle at least quite independent of the mixing of their phenotypes. The exact
genetic basis of heterozygosis is debatable. One could argue thet it merely repre-
gents the close adhesion of different particles to one another, which would then behave
as a unit on limit-dilution. Or possibly within a single particle there are two separate
genomes which can then be distributed to different progeny in further replication of the
virus RNA. At any rate, this again does not yet represent a complete phenomenon of
genetic recombination. There has been no exchange of parts of the genetic material
of either one or the other in the formation of any stable clone of virus. And the same
results as we attribute to heterozygosis would pertain, if in fact the infected fluids
which we consider to be single clones had in fact derived from more than one virus
particle.

Finslly, among the yields of such mixedly infected cells one does find a nuam-
ber of particles which can yield pure clones which can be attributed to be recombinants.
These are clones which will have stably the characters, say large A and small c, or
small a -- I used the symbol W for that, for the alternative serology ~- and large C.
These recombinants are stable by the criterion of repeated single clone isolation

through limited dilution and also by the virfiie of the fact that although virtually all
-7~

of the virus population of this type of recombinant can be wiped out by the use of anti-W
serum there is not then found a residue of the alternative type. This is just one of a
number of lines of evidence for the purity of these clones, although we see already

the technical difficulties which beset this type of investigation as compared to systems
where we can isolate clones by simple methods of plating, as of bacteria or of plaguing
or plaque formation es for bacteriophages.

Similar phenomena of genetic recombination have been found in the rather more
complex virus of the vaccinia group, which Fenner has investigated, and for which 2
number of markers, for example the formation of hemagglutinins, resistance to heat,
and specific virulence for rabbit and mouse hosts, ag well as the characteristic forma-
tion of poieon the chorio-allantaic membrane have been used as markers. At the pre-
sent time vaccinia appears to be perhaps the best system in which we now have the
technical possiblity of investigating genetic recombination among viruses, but the re-
sults so far reported in this system have been very recent indeed.

At the present time, the virus in whose genetics there is perhaps the most
practical interest is undoubtedly the polio virus, or the group of polio viruses. Here
there are active efforts contimially being pursued and tested on a very large scale in-
deed to obtain mutants of the polio virus which would be capable of acting as efficient
immunizing agents and yet free from the hazard of disease and paralysis associated
with typical polio. It must be admitted that the methods so far used for the selection
of such mutants have been rather empirical. They have often involved the growth of
the virus in unveual hosts, or in tissue culture of cells of types not associated with
fhe nervous system, and while there have been apparent successes in obtaining mutants

better adapted, say to epithelial celle than to the nervous system, we are not yet in a
position to precisely define the genetic changes which were the basis of these changes
in virulence.

Fortunately the development of methods of isolation of single plaques has at
least made it possible to obtain pure clones of those mutants which have been secured,
and thereby to minimize the possibility that theac clones subject to test had been con-
taminated by admixture with particles of another genotype. Unfortumately attempts
to demonstrate genetic recombination among strains of the polio virus have until now
been unsuccessful. If they should succeed this would be perhaps the most productive
method of obtaining recombinant genotypes which would be capahle of showing the im-
munizing activity that we are seeking, and from which we have deleted, by recombina-
tion, the specific adaptation to the human host characteristic of these viruses.

Although genetic work with the polio virus, and likewise with other encephali-
tic viruses and with the tebacco mosaic virus, has thus been limited to mutational
studies, they have been the basis of other equally remarkable findings in the field of
genetic chemistry. Several workers have found, with each of the viruses that I have
just enumerated, that it is possible to separate the protein from the nucleic acid com-
ponent of these viruses. These can be done, as Schramm did, for example, by ex~-
posing the virus to aqueous phenol solution which will destroy the protein and leave
the ribonucleic acid intact. Conversely, the protein,of the plant tobacco mosaic
viruses at least, can be removed in reasonably native condition by exposing the virus
to moderate alkalinity.

Two phenomena of great interest have been found. First, that in the case of
these viruses, an extract of the virus containing only the ribonucleic acid portion and

lacking any demonstrable protein does retain substantial infectivity. We might con-
-9-
sider that if we were dealing with a hypothetical virus particle, which we do believe in
general consists of an outer protein coat and an inner genetic component of the nucleic
acid, that in fact only the enclosed RNA is essential for the maintainance of the state
of infection of a cell into which it is introduced. On this basis, the primary function
of the protein coat is to act as the nose cone, so to speak, for the preservation of the
virus particle in transit from cell to celi, and for the reentry of this particle into a
new cell which it might infect. Given this, and this is merely 2 restatement now of
the observations that have been made, the isolated nucleic acid may be expected to be
infective provided it.is kept under rather gentle conditions, so as to minimize the pos-
sibility of its chemical inactivation while it is in solution, and provided there exists
some mechanism for the reimportation of the virus nucleic acid into the infected cell.
We do not know what that mechanism is in the case either of the tobacco mosaic virus
or of the mammalian cell, but this does appear to happen, that isolated nucleic acid
is capable of penetrating the mammalian cell. Once having penetrated the cell, this
forms part of a vegetative pool of virus which is, in fact, identical to the normal fate
of intact virus that might enter the cell, since we have reason to believe, as I will
elaborate further, that the protein plays no important part in the further replication of
the genetic material, or in the production of new coats for that matter.

Once inside the cell this vegetative pool enlarges, particles will then mature,
and will develop whatever protein coats are specified to be around them by virtue of
the information contained in the nucleic acid inside the cell, and they will then be foi-
lowed by the liberation of intact virus of the usual sort at some later interval. At the
present time different preparations of purified nucleic acids have infectivities ranging

from a very small proportion, perhaps one part in several thousand, to (a) several
-10-

percent of the infectivity of a eproparable amount of the intact virus.

The experiments which show the relative unimportance of the outer coat of the
virus particle have been done by Fraenkel-Conrat of Berkeley who has succeeded in
reconstituting tobacco mosaic virus by mixing preparations, on the one hand, of the
mucleic acid, on the other hand, of the protein -- that these will, under certain condi-
tions, again coalesce to form particles sessentially similar to the original virus par-
ticle. In the case of tobacco mosaic virus this is a long linear particle, and if we
drew a cross section of it, would look something like this. Let this be a longitudinal
section with a spiral of virus material inside a protein coat. This protein coat would
be built up of many monomeric building blocks, stacked on one another in a regular
array.

Fraenkel-Conrat also showed that one could make such preparations, on the
one hand of the RNA, on the other the protein, from two distinct viruses, which could
be distinguished from one another either serologically or by virtue of the specific
symptoms on the plant material that they infected. One could, for instance, obtain
preparations from a strain of tobacco mosaic virus called HR, Holmes ribgrass to-
gether with the standard tobacco mossio virus strain. And one could now make mixed
preparations which would have the protein of Holmes ribgrass and the nucleic aoid of
tobaceo mosaic virus, or vice versa. The characteristic behavior of such particles
is quite reminiacent of those which one obtains on phenotypic mixing from the issue
of cells which are mixedly infected.

One should interject that this phenomenon has not been directly observed in
the case of plant viruses, since in fact we have no assurance, owing to the low infec-

tivity of these particles, that we have ever mixedly infected a single plant cell with
~ii-
two viruses. This has been seen with many animal viruses.

The attributes of these mixed particles in this experiment, which is called more
properly a reconstitution than a recombination experiment -- the attributes of these re~-
constituted particles are this -- we are dealing, let us say, with a tobacco mosaic virus
RNA and a Holmes riberass coat. Well, such particles, unlike normal TMV particles
will not be inactivated by anti-TMV anti-serum. They will, however, be inactivated
by Holmes ribgrass anti-serum, since the point of attack of antibodies is the surface
protein coat and has to do with the initial entry of the virus particle into the cell. How-
ever, the progeny of these particles, although the initial infecting particle was a Holmes
ribgrass with respect fo its protein, are all typical TMV particles. The protein manu-
factured by cells infected with such a particle is determined exclusively by the nucleic
acid, and corresponding results were obtained on the other side with the opposite com-
bination. From this we can conclude with reasonable assurance that the nucleic acid
is the primary determinant, in fact perhaps the sole determinant, of the specificity of
the proteins produced in an infected cell.

There are other factors which distinguish the stability of intact virus particles
from RNA. The isolated mecleic acida are very much more susceptible to such enzymes
as ribormclease, to thermal degradation, to changes of pH. On the other hand, they
prove to be even more resistant than the isolated, than the intact virus particle, to
detergents and other materiale whose primary effect is on the protein.

Since there are now available a number of systems where it is possible to utilize
mucleio acids ea primers for the synthesis of increased amounts of these nucleic acids,
systems of this sort will eventually play an extremely important role in the replication

of biologically active nucleic acids. At the prezent time we have strong evidence for
~12-!

specific replication in an enzymatic system of DNA materials, but not yet for RNA
materials. To this end it is therefore encouraging that there are some recent findings
which support the possibility of the infectivity of the isolated nucleic acid of bacterio-
phages ~~ in this cage this is DNA. These preparations consist of suspensions of bac-
teriophage which have been treated in somewhat the same fashion as the particles of
the RNA viruses that I’ve indicated here. In other experiments they have been exposed
to high concentrations of urea, which seems to have the effect of at least opening up
the virus particles.

These preparations require to be resolved considerably further before they
ean be available for the same kind of reconstitution experiment as has been described
for the tobacco mosaic virus. At the present time the specific replication of virus
genetic material from anima! and plants viruses constitutes the best explicit evidence
for the exact replication of RNA, evidence which is already more than ampie for DNA.
Experiments on the recombination in influenza show that RNA is also susceptible to
the same fundamental processes of crossingover and genetic recombination. They
therefore furnish a strong motivation for searching for comparable mechanisms of
in vitro replication as have been described in DNA.

Throughout this discussion the common thread has been the continuity of genetic
information through the agency of the nucleic acids. However, it ahould be stressed
that DNA and RNA are not unique materials. They have the same relationship to the

information contained in them as carbon black does to the words in a dictionary.