Transcript from filmed lecture
November 15, 1959

Lecture #46 - Virus Genetics: Bacteriophage
by J. Lederberg

So far our discussion has emphasized behavior of organisms of a considerable
degree of complexity. These organisms, starting with bacteria, including Drosophila
and mammals, have a sufficient content of information in their genetic material so as
to specify a large number of different enzymes. These enzymes, together with other
proteins which make up various structural features, suffice to give these organisms a
free-living character. A bacterium, for example, is capable of subsisting in a rela~
tively simple environment which it converts to its own use through a wide range of dif-
ferent synthetic pathways. For each step of these pathways we must consider the pro-
bable presence of a distinctive protein whose amino acid sequence has been specified
by its DNA content.

Probably the most significant measure of biological complexity of different or-~
ganisms is farniahed by a count of their nucleotide information expressed in the pairs
down the length of the postulated Watson-Crick helices in the DNA. And we see that a
bacterium, for ali of its emall size, already has 6 million micleotide units per nucleus
a8 compared with the 80 million in a Drosophila, and the 5 billion in a mouse nucleus.
if we assume, as now appears to be an approximately correct estimate, that it will
require something of the order of 2000 nucleotide units to specify a single protein of
average size, we see that there is sufficient DNA in a bacterium to specify something
like 3000 different enzymes. Now by no means no number of enzymes remotely ap-
proaching this number has actually been isolated from any single bacterium, but if one
considers the wide range of metabolic activity which free-living organisms are capable of

sustaining, this is by no means an unreasonable estimate. In fact, one might be sur-
prised that a bacterium could get along with so low a mumber.

When we come to the simpler organisms that will be the subject of the current
presentation, we see that their genetic simplicity already poses a definite limitation on
the possible complexity of biological processes that they are capeble of sustaining. The
phages that we have been discussing so far are particles whose nucleotide content is of
the order of one hundred thousand pairs, and this would allow, according to the same
estimates, for the production of something like 50 different proteins. Considering
the complexity of the enzyme systems that would be necessary for the production of
the amino acid constituents of these proteins and for their assembly and for the produc-
tion of the elements of the DNA and ita assembly in the bacterium, we would have to
conclude that any particle of these dimensions must rely very heavily on the available
metabolites, and not only metabolites, but the metabolic systems of the host in which
this particle resides.

Given this degree of simplicity, of a bacteriophage like T2 or lambda, it is
therefore not at all surprising that these organisms are incapable of a free-living ex-
istence, that they are unable to multiply and produce more of their kind except within
the confines of a metabolically active host cell. This host cell provides most of the
accessory metabolism which is necessary for the synthetic processes of the virus;
the virus adds those elements which are unique to its own structure, enabling it, on
the one hand, to compete with the metabolic systems of the cell that it infects, and, on
the other hand, to produce protective structures which will enclose the virus during
the interval when it must live outside the host cell.

When we come to the smaller viruses, for example, the animal viruses in the

encephalitic diseases, and an exceptional bacteriophage called phi X 174, we find that
we can have living particles of even smaller dimensions, of the order of 5000 nucleotide
units, which is hardly more than the information needed to, which is encompassed within,
one or two, something of that small number, of distinctive genes. It is of some interest
that these smaller viruses so far established are all one-stranded in their nucleic acid
composition, Most small viruses, and this ie true of all of the small viruses of plants,
and of animals es well, are composed of RNA which is typically one-stranded, or whose
structure at least is not so decisively determined in the two-strand configuration as is
that of DNA. The virus just mentioned, phi X 174, has created some excitement in the
world of biochemical genetics since it has recently been found that it ie a unique example
of a DNA configuration which is one-~stranded. In fact, some authors have suggested that
this virus nucleic acid really ia much more reminiscent of RNA than of DNA because of
its secondary structure, although it has all of the basic elements,of a deoxyribose
sugar, and of the characteristic bases,that we see in other DNA molecules.

We must consider that a virus of this sort does a very remarkable job since it is
capable by virtue of the production of perhaps two distinctive proteins, or something
of that order, of sufficiently diverting the normal metaboliam of the host cell which it
infects so as to result in the production of large quantities of this particular nucleic acid,
and of the protein associated with this nucleic acid, rather than the normal metabolic
products of its host.

Now if we look a little bit more closely at the structure of a bacteriophage

particle, bacteriophage being a rather special class of viruses with a stracture unique
to these, these are the viruses attacking bacteria, we can get a schematic diagram,
something like this. We are dealing here with a particle whose characteristic dimension

is about one tenth of a micron, about a tenth the diameter of the bacterium that it in-
fects. It has a protein coat which has a somewhat crystalline form; within that protein
coat is a tightly packed unit of nucleic acid. In fact, one can calculate that, if properly
extended in a double-strand helix, this micieic acid will be some 34, long, and yet it is
packed in an efficient and orderly way in a box of these particular dimensions.

Accessory to this protective box around the DNA is a tail, which has a spiral
protein around it which may serve some contractile function in the penetration of a
host bacterium by a bacteriophage, and there are tail tip fibers which are believed to
play an important role in the specific adsorption of the bacteriophage particle to the
bacterium.

This structure which is characteristic of the virus in its free state, (in) outside
the bacterial cell, is however not found in the recently infected bacterium. And in
fact for many years there was considerable mystery concerning the basis of the so-
called eclipse period immediately after the infection of a host cell when a bacterium
that was destined again to yield intact bacteriophage did not demonstrably contain any
element of it, This mystery was eventually resolved by tracer labeling experiments
by Hershey and Chase, who found that they could differentially labild the meleic acid
component of the bacteriophage particle by incorporating radioactive phosphorus 32 into.
the DNA as against the protein component which contains sulphur amine acids and could
be labeled with radioactive 8-35. They found that after bacteriophage had adsorbed onto
sensitive bacteria, and electron micrographs support the contention that this adsorption
is first at the tail, that there was subsequently a separation of the P-32 labeled DNA
from the protein. The P-32 somehow enters the bacterium, leaving a shell of protein
outside. In fact, Hershey and Chase were able to demonstrate that after the bacteria

had been initially infected they could strip off the empty shells containing all or virtually
all of the protein of the bacteriophage by exposing the bacteria to high shearing forces,
as in a Waring blender. This, however, did not in any way alter the fate of the innoculated
bacterium which had received the genetic element, the DNA from the bacteriophage.
However, this is not, this DNA is not, a bacteriophage particle. This is the genetic
element of the bacteriophage, but before a particle is produced which has the capacity
for free existence outside the cell, and which is capable of reinfecting another cell, a
new eycle of protein synthesis and organization and assembly of the intact phage particle
must first take place. This does not begin for several minutes after the initial infection
of the bacterium. During this time there is considerable replication of the DNA com-
ponent, which is now called the vegetative phage to distinguish this from the intact
phage particle, which is the infective phage particle. This vegetative phage is con-
' gidered to form a pool of DNA units of the replicating DNA of the phage from which from
time to time components are sampled, in the sense that a fraction of the vegetative pool
will undergo a condensation of the DNA and then around that DNA a new skin, let's say
a new head and a new tail, will be produced, again to form an infective phage particle.
As a consequence of other materials produced by the bacterium which is infected
by the bacteriophsge, namely endolysins, which are enzymes that can lyse a cell wali,
after some 20 to 30 to 40 minutes the bacterium wili be lysed and the infective phage
particles contained within the bacterium will be liberated in the medium. These infec-
tive particles are stable in free suspension, and can persiat until the next occasion when
they encounter a suitably competent bacterial host, again to initiate a cycle of infection.
This is the typical cycle for a lytic bacteriophage, or to speak more precisely,
this is the lytic cycle of a bacteriophage, and it represents one of two alternatives which

is available to a bacteriophage like lambda. As we already know, lambda is capable of
-6-

penetrating a bacterium, and it does so through a separation of ite DNA and protein in
much the same fashion, but here the vegetative lambda can pursue either the course
of lytic multiplication resulting, that is, in the ultimate lysia of the bacterial cell, or
in establishing a new relationship with the chromosome of the bacterial host, in this
case giving rise not to a dead, lysed cell, but to a lyeogenic bacterium, which betrays
its peculiarity by conferring the property on its descendants of eventually producing
some of the same bacteriophage.

Now, when this bacteriophage is liberated into the medium it not only destroys
the cell from which it has been liberated, but it is also capable of now, these new particles
are capable of, infecting any bacteria of a susceptible type that may be their environment.
Consequently, we can readily detect the presence of a bacterial virus by using 2 sensitive
strain as an indicator host. This can be done in liquid medium, in a fairly obvious way,
by waiting for the lysis of the entire bacterial culture. However, for many purposes of
enumeration it is more convenient to conduct these assays on a solid medium. These
assays represent one instance of the great utility of looking at nothing to get a considerable
amount of information.

What we set up is a Petri dish with an ample supply of nutrient agar, and on top
of this dish we add a layer which is heavily seeded with sensitive host bacteria. Each
of these bacterial cells will form a small clone, as you already know, but the number is
go large that they tend to merge into one another, and we have what we may very appro-
priately call a lawn of bacterial growth. If, at the same time as the bacteria are added
to this layer of agar, we have added a few phage particies, they will find themselves in
the vicinity of some susceptible bacterium, the phage particle after one cycle of lysis

will result in the liberation of some hundred, or several hundred, daughter particles,
which will then diffuse to the neighboring bacteria in the vicinity of the original burst.
This will then result in a progressively increasing zone of lysis, or of clearing of the
bacterial growth, at the point where a bacteriophage particle has alighted. I have an
example of that kind of a plate here, and this plate, which will now be projected on the
screen, represents a plating of the bacteriophage lambda, of which I've already spoken.
1 would just ask you to disregard that white dise spot in the middle, which is a bacterial
contaminant that has overgrown in the middle of the plate. In these clearing} which you
can see in the plate, are each of them so to speak, a colony of lambda which has grown
at the expense of the bacteria in its immediate environment. At the same time as the
lambda has grown, the other bacteria that have not been exposed to lambda, have algo
grown, and they give rise to the opacity against which the colony of lambda, the plaque,
appears as a clearing. We therefore have a ready way of counting the number of bac~
teriophage particles that were present in a given saraple added to the bacteria on this
lawn.

In addition there are characteristic features of the plaque, which we cannot al-
ways trace to any very simple attribute of the bactertophage-host system, but which
do give us the basis of a morphological differentiation of different kinds of plaques,
and hence of the genotypes of the phages which produce those particular plaques.

Now we may very appropriately inquire what there is about the structure of the
DNA of the vegetative bacteriophage which has infected a bacterium, and which we know
to be multiplying very rapidly, that distinguishes it from the existent DNA already pre-
sent in the host cell. Now we must assume that since we do have bacteriophages which
are quite virulent, which are capable of immediately stopping the synthesis of other

proteins, and indeed of the synthesis of bacterial DNA, that there is some remarkable

element in the chemical structure of the virus DNA to account for this. In at least
-8-

one case we have a clue as to the basis of thie differentiation, because the phages T2,
T4, and T6 do indeed have a characteristic base, hydroxymethyl cytosine, which ap-
pears in the DNA of these phages in place of the cytosine, which is characteristic of
most other DNA. And it can hardly be a coincidence that hydroxymethyl cytosine, as
far as we know, is found only in these particular bacteriophages. And it appears likely
that the hydroxymethyl cytosine DNA has features of resistance to the degradative en-
zymes which will be present in the bacterial eytoplasm which will account, at least

in part, for the unique reproductive capacity, which is fo say the virulence, of this
DNA as an intracellular virus.

However, we cannot appeal to this type of peculiarity in chemical composition
to account for the virulence of other phages like T1, or lambda, where we can find no
detectable differences in base composition. This dilemma is, in fact, especially
pointed for temperate phages like lambda, which can function as part of the bacterial
genome in the lysogenic cycle, and then can be liberated, for example by ultraviolet
induction, to function as lytic bacteriophages. We are left here with a mystery which
will require considerable further investigation of the biochemistry of virus infection to
elucidate.

Having established the basic features of the life cycle of a virus, we may now
inquire of some aspects of its genetic behavior. In order to demonstrate genetic inter-
action between virus particles, it is, of course, necessary to doubly infect a singie
host ceil with two such particles. This can be done by simply mixing sufficient bacterio-
phage with a suspension of bacteria to insure that each bacteriophage will have simul-
taneously adsorbed two or more bacteriophage particles -- I said bacteriophage, I

meant each bacterium adsorbed two or more bacteriophage particles. If these are of
different type we then have the opportunity to study the interaction between vegetative
DNA derived from bacteriophage particles inhabiting the same bacterium.

What has been found, as one might anticipate from the theme of all of these lec-
tures, is that, quite regularly, these mixed infections result in an exchange of genetic
material. This can be demonstrated, for example, by the use of such markers as the
host range quality, and the plaque type quality, which are designated by the symbols
hr. If one bacteriophage is a double mutant, another the wild-type (and the plaque
type with respect to) -~ wild-type with respect to plaque and host range -- we then can
find in the issue from such mixedly infected bacteria not only the two parental types,
but algo the two recombinant types, ht r and h rt.

The simplicity of the DNA structure of bacteriophage compared to other organisms
has encouraged the hope that it might be possible to approach the ultimate complexity of
the DNA in genetical experiments. We would, after all, in principle, in order to com~
pletely map phage T2, have only 100 thousand different mutants at different sites at our
disposal. Now, we have not approached anything quite like that level of precision in the
analysis of the entire genome, but Benzer has set out to do this in phage T4, with respect |
to a definite restricted part of the total genoroe of that phage.

In phate T4 there is evidence for a single linkage group of the extent of 100
thousand nucleotide units, within which there is a segment whose length approximates
1% of the total map length of the bacteriophage, and this segment is labeled r. A large
variety of r type mutants have been found in T4 by Benzer and classified by attempts to
cross them with one another. Fortunately an easy selective system was available,
whereby r mutants can be distinguished from the wild-type, since the wild-type being
capable of growing both on cells of strain B and of strain K12 of E. coli, and the mu-

tant being capable of B, but not of growing on strain K12, it was an easy task to isolate
~10-

rare particles issuing from mixed infections of r mutante which had reacquired the
capacity to grow on strain Ki2.

In this way it becomes a simple task for Benzer to study the recombination
occurring between bacteriophage particles in mixed infections at rates of as low as
10-6 or 10-8. By studying a large collection of r mutants Benzer has found it possible
to classify about 1500 of them into some one hundred different sites within this r seg-
ment of the bacteriophage. The implication of this is that there have been many re~-
currences of mutations at certain sites, and indeed it does appear that there are
gome so-called "hot spots" which are much more liable to undergo spontaneous muta-
tion than other sites within this particular locus.

Now, considering that 1% of the bacterial map should give us approximately
a thousand nucleotide pairs, a thousand units susceptible of mutation in the locus, it
is apparent that Benzer has not yet quite exhausted all the mutational possibilities in
this region. However, he has achieved a reagonable step in that direction, and he does
indeed find that the distance of, that the frequency of least recombination which has been
found between, any single pair of mutants is of the order of .01%, which is in close
agreement with that predicted from the dividing the total map length of about 10 morgans
by the length of the segment that we are dealing with here. He therefore concludes, with
some degree of plausibility, that the minimum distance of recombination, observed
genetically, corresponds to the least unit for which we oan attribute any chemical basis,
namely the interval between two successive nucleotides in a DNA string.

Benzer can also study the phenotypic interaction between different mutant phages
growing in the game bacterium. A single bectertum may be mixedly infected with a

mixture of two distinct r mutants. This affords an opportunity to determine what possi-
-Li-

bility, if any, there is of phenotypic cooperation between the vegetative DNA of these
two genotypes. Benzer found that many combinations of r mmtants were capable of
cooperating so as to result in the growth of each of them and the lysis of the cell that
contained them, although by definition the individual particles of r, or Tg would be in-
capable of effecting lysis of the K12 cell. The yields from such mixedly infected cells
in these cases would consist of particles, for the most part, either of the parental rj
or of the parental rg type, and the phenotypic cooperation should be carefully distin-
guished from recombination which gives rise to genetically wild-type particles.

A complete classification of Benzer’s routants resulted in the establishment of
two groups within the r region. These groups may be considered cistrons, since the
test for cooperation of the two genotypes within a single bacterial cell is equivalent to
the cis-trans test of cooperation in a heterozygote. It was found that all of the mutants
could be classified either into cistron A or into cistron B, 50 that the mixed infection
of a single bacterium with one member of A and one member of B resulted in effective
cooperation. From this Benzer concludes that there are two separate protein elements
synthesized under the governance of this region, both of which are required for the ef-
fective growth of a bacteriophage in this strain.

Further evidence as to the fine structure of the genetic material is furnished by
analyzing a collection of mutants with a pomewhat different property from the point
mutants now given. These behave as if they are deletions covering various parts of
the total mila, indicated by these segments over the point mutants on the diagram.
Benzer found that one could uniquely assign the position of the point mutants in one and
only one order so as to give a compatible superimposition of the various deletion types.
Consequently we can conclude that even in the utmost details of its fine structure the

genetic material of phage T4 is organized in a linear fashion.