Transcript from filmed iecture
November 14, 1953

Lectare #4 - Bacterial Genetice: Sexual Reproduction
by J. Lederberg

Studies on bacterial mutation, particularly those concerning the effects of
various chemicel reagents on the production of new genetic forms in bacteria, have
been of great value, and will continue to be in elucidating the structure of the genetic
materia. Nevertheless, there are serious limitations which are quite obvious in any
genetic analysis which is confined to studying those changes which arise in the course
of purely vegetative reproduction. If we are to conduct any experiments on the interac~
tion of different genetic components, we rust haye some mechanism available to us for
recombining genetic factors from more than one kind of cell.

Until about a dozen years ago this was not considered possible. Many investi-
gators had locked for sexual processes in bacteria, using principally morphological
analysis, searching through the microscope for evidence of unusual forme that might
betray the fusion of dissimilar cells. The futility of these oxperiments is demonstrated,
after the fact, by the calculation that in early experimeata on recombination in Eecheri-
chia coli the interesting events occur S0 a3 to involve only about one per miilion of the
bacteria in any given culture. In this situation there is an evident use for the principles
of selective isolation of specific genotypes, a8 has already been appiied for the measure~
ment and demonstration of occurrence ef genetic changes in the course of vegetative re-
production. Among the mere ugeful markere for atudies of this sort are the nutritional
defects such as have been discovered and elaborated by Beadle and Tatum in Neurospora.
Just as in Neurospora, genetic changes arising in Escherichia coli, either spontaneously
or under the influence of various mutagens, are capable of resulting in defects in a bie-
synthetic sequeace and thereby in establishing a requirement for 2 specific growth factor

which must be furnished in the medium if the bacteria are to be capeble of continued
~2-

growth. Using distinct blocks in different biosynthetic chains one can easily set up an
advantageous system for the detection of genetic interaction, as i can illustrate in the
following way. Let us suppose that starting from our wild-type strain, and, to speak
the facts here we might refer to strain K12 of Escherichia coli, that we start out from
the original wild-type, whose nutritional sufficiency can be indicated by the symbols mt
T*, and from this obtain two distinct mutants ~- one of these a mutant requiring say
methionine, which we will symbolize M~ T*, another requiring an alternative amino
acid, let us say threonine, we will symbolize M* T-. These are two distinct auxo-
trophic, nutritionally dependent, organisms, and neither of them by itself will be cap-
able of producing any cclonies when cultures are plated on basal minimal media which
contain the general carbon and nitrogen sources but are deprived of the specific supple-
ments needed by these strains.

Therefore, if in a mixed population containing some M- T+ celle and other
Mt T- cells, there should be any recombinational events, these can be detected with
great efficiency by the occurrence of a new form which will resemble the wild-type,
and is therefore called a prototropb, and which is capable of growing on the synthetic
medium. Since these will be the only celis capable of forming colonies on agar plates,
even though they be innoculated with many millions of celis of either parent, we have
indeed an efficient mechanism for the original detection of recombinational events.
This was indeed the case when experiments were done with the K12 strain beginning
about 1946.

The early experiments involved a mating type, F+, to which I will refer a little
bit later, and did give a very low frequency of recombination. Subsequently when work

of this kind was continued, in my own and in many other laboratories, new strains dis-
playing very much higher rates of recombination have been discovered, and these strains
which had a high frequency of recombination are therefore called Hfr strains. Continuing
the analysis of sexual reproduction in bacteria by means of these strains, Wollman and
Jacob have established in somewhat more detail some aspects of the system of feriliza-
tion. First, we consider that in microscopic examination of Hfr by F- crosses it has
been possible to demonstrate the occurrence of conjugal pairs of cells where a male and
a female cell having made a random contact with one another, a conjugation bridge is es-
tablished between them. And I'll take the opportunity of displaying a photograph taken
through the microscope of just such an event. In kinetic experiments where male and
female cells were mixed and then subjected at various times to the great shearing forces
which operate in a Waring blender, so as to separate what these Frenchmen eall the
‘happy couples" Wollman and Jacob were able to show that there was a scission of the
bridge separating the two cells, without disturbing their viability. And they showed
furthermore that there was not an immediate transfer of a block of genetic material of
the entire nucleus from one cell to another, but rather a progressive transfer of differ-
ent genetic elements.

To signalize this progressive transfer which requires about 100 minutes for
completion, we conclude that there is a substantial unwinding of the DNA so that it is
represented no ionger by a coherent blob of material as we see it in the usual stain
preparations, and that one end of the DNA string is preferentially transmitted first,
followed by the remainder of the DNA, until finally all, or almost all, of it is transferred.
There is thus a Unear progression of the various markers which are represented by the
segments of DNA which are arranged in a linear sequence down this chromosome, from
which Jacob and Wollman have succeeded in making the most meticulous maps of the

genetic material that have been made available.
Expressed in somewhat different terms, the rate of transfer of the DNA cor-
responds to approximately one thousand nucleotide pairs, or to coin a new unit, n’its,
per second, which is about half the extent of a typical gene -~ a gene being here defined
as one of the cistrons responsible for the production of a characteristic protein. By
the way, Levinthal's analysis of the genetic control of the enzyme alkaline phosphatase,
using procedures of this sort, and applying them to a variety of mutants in which this
enzyme was defective, verified that there was approximately the four to one coding
ratio, which is demanded by the elementary considerations of correspondence between
nucleotides and amino acids, as other speakers have summarized.

After the genetic material has been transferred, and perhaps beginning before
the transfer is complete, there must be some process of synapsis, of orientation of the
newly incoming material with some of the DNA of a pre-existent nucleus. And following
this, by mechanisms which are still quite imperfectly understood, an exchange between |
the two ~- this resulting in a recombinant which may have some markers from the fe-
male parent which it has retained, some markers from the male parent, whose DNA has
been transmitted into the female cell. It is by no means certain at the present time
whether this recombination of the parts of the chromosome represents an actual physi-
cal breakage and reunion as envisaged in some of the canonical models of crossingover
in higher organisms, or whether there is a mechanism of choice of alternative templates,
that is a copy-choice, during the replication of daughter DNA, using as alternative
models either the one or the other parental material.

As a consequence of all these processes, first the initial injection of the DNA
during fertilization, its synapsis with the genetic material of the female parent, and

the integration or crossingover involving information from the two DNA components,
5

we have the formation of the recombinants which are in the first instance detected by
the selective procedure that was outlined before. However, this represents a cross in
which only two markers are involved, and this would not give us a great deal of detailed
information. Therefore, typically in crosses involving Escherichia coli we use six,
eight, or even a dozen differential markers, scoring different biochemical, physiologi-
cal, drug-resistance characteristics, and I have here on the light box a plate which il-
lustrates for more or less typical recombinants, and how we score them. On the first
line we have a lactose-negative clone, which on this medium scores as a colorless, or
very lightly colored colony. This has been brushed across 4 line of a loopful of strep-
fomycin which was placed on the plate before these streaks were made. And the strep-
tomycin is in the vertical line, represented by this pencil. The first type, which might
be one parental form, was 2 lactose-negative streptomycin-sensitive type. Another
parental form might have been a lactose-positive streptomycin-resistant form, and
then in addition to these two parents, two recombinants which might issue from such a
crosa would be a lactose-positive streptomycin-sensitive, and a lactose-negative strep-
tomycin-resistant.

This type of marker can be extended by the use of other sugars for which we
have fermentation mutants, by the use of other antibiotica for which resistance may de-
velop by quite independent mutations, by the use of bacteriophages for which again we
can develop resistant mutants, and in other instances by markers such as specific
antigenic quality, ability to move in semi-viscose media, and so forth.

We vow have to consider the initial basis of sexual differentiation in this ape-
cies. I've referred loosely in the diagram to a male and a female type, and we might

add that these designations are partly, although not entirely, arbitrary, and are based
on the fact that the male cell contributes a migratory gamete, whereas the female cell
remains entirely sessile. Having contributed a migratory gamete, the male and femaie
in the ordinary course of events, or in a blender experiment such as Jacob and Wollman
have done, can be separated and will continue to propagate. The male cell, however,
has suffered no genetic alterations, it has no new genetic information as a result of

this experience, whereas the female cell will produce a clone which will be a mixture

of unaltered markers that it originally had, representing those nuclei that had not been
fertilized by this event, and will sleo segregate a number of recombinants resulting from
the interaction of ita newly received segment of DNA with the DNA of the existing form.

Now, what then determines the sexual distinction between these two types 7
We first have to mention that the original wild-type strain was self-compatible, and
for some time it was not realized that there was a differentiation into two mating types.
However, in subsequent work mutant strains have been derived which have apparently
lost the capacityto function as male, and these we call female strains, since this is the
only function that they can entertain in the mating process. For reasons which will soon
become apparent, we also call these female strains F-. The typical male strain, in the
sense that this was the type of the original culture, we call F+, and these are our first
examples of female and male strains, respectively.

Now in many organisms sexual differentiation is determined either by an entire
chromosome, as is true, for example, in man, or by the balance between one chromo-
some and the rest, as is true in Drosophila, or, as is true in many fungi, by a single
segregating gene. And we might expect, ab initio, that the same would pertain for this
microorganism. It was therefore quite startling to discover that sex, or rather the

determination of male sexuality in Escherichia coli was an infectious phenomenon.
That is to say, the introduction,of just one male F* cell into a population of F- female
celis, would rapidly result in the spread of the male character to many and eventually
to all of the female cells in that culture. And this male character, newly acquired,
was hereditarily atable, in the same sense as any other marker; all, or virtually all,
of the progeny of the newly converted F+ celis would continue to be male.

We are dealing here then with the transmission of a factor which is highly con-
tagious and which,from the dynamics of the spread through the population, must be cap-
able of multiplying at a rate at least twice that of the typical cell. However, all at-
tempts to isolate a cell-free virus particle, which one might have postulated in view
of the high infectivity of the male character, have failed. And we are therefore con-
streined to conclude that while there is an extra chromosomal factor which distinguishes
an F~ from an F+ cell, we can symbolize the genome of an F- cell by a single unencum-
bered line. This represents one of the nuclei stretched out. Although not stretched out
to its fall extent, since a consideration of the number of n'its in a whole bacterium would
tell you that there would be something like 22 » in total length of the DNA, and that
would be of the order of 5 to 10 times the length of the bacterium. So that we can see
the important role of close packing of the DNA strand in maintaining a properly function-
ing organism.

However, we do represent the normal genome of an F- cell as an unencumbered
line, symbolically, so. Then the F+ cell must have something else, and by a variety of
lines of evidence, chief among which is the high infectivity and the high reproductivity
of this particle, we conclude that we are dealing with an extra chromosomal F particle.
This particle is evidently transmitted from male to female cells when they are planted

into the same culture medium by means of a transient mating similar to that which operates
~8-

in the transfer of a chromosome. However, these matings which involve the transfer
of only an F particle are evidently much more unstable and do occur rather more fre-
quently than those involving chromosome transfer, and therefore account for the rapid
spread of the F factor among the female cells in a population, far exceeding the rate
at which chromosomes as a whole are transmitted.

Another feature which distinguishes the F particle from the chromosomal fac-
tora is the possibility of eliminating it differentially by means of certain dyes, for ex-
ample, acridine orange. The exposure of an F* cell to these dyes results in its recon-
version, back to the F~ state, apparently owing to the differential inhibition of repro-
duction of the F particle. At the same time the other genes in the bacterium which
have been exposed to the dye are apparently unaffected. The F particle confers several
interesting attributes on the cell which it inhabits. In the first place, it must modify
the cell wall of the male cell so that it 1s capable of recognizing, (with) and reacting
with, female cells with which it may come into contact. It must then form a bridge be-
tween the two cells, and finally must confer motility on the chromosome so as to enable
it to be transferred from the male to the female cell, aa previously described. That it
may do this by first attaching, at least temporarily, to the chromosome is indicated
by the occurrence ofa third mating type, designated Hfr. These mating types have
been studied and the factor determining thelr maleness analyzed by the methods ap-
plicable to other markers. And it has been found that in these mating types we are
dealing with a. mating type determinant which is located on the chromosome itself.
Quite characteristic of the Hfr types is the location of the F factor at the very end of
the chromosome; consequently it is the very last marker to be transmitted in the

course of a typical mating -- this, in distinction to the transfer of the extra chromo-
somal factors in an F* mating type, which are easily transmitted even in the absence of
any chromosomal transfer.

How do we know that we are dealing with an F factor, as the determinant of
maleness in an Hfr strain? This assertion is based on two findings. First, that all
Bfr strains so far described, and a great many of them have, are derived from Ft
cultures. Well, F- cultures have been stably female and never give rise to any male
strains, unless they are first contaminated by exposure to male cells, permitting the
transfer of an F particle.

Secondly, some, in fact many, Hfr cultures are relatively unstable, and will
now and again give rise to a reveraion from the chromosomal state of the F particle
to an extra chromosomal state, at the same time losing the very high fertility which
is characteristic of a male strain. And finally, while the male character of an Hfr
strain is not susceptible to inhibition by the drug acridine orange, the character of an
F* strain, as already mentioned, is. The reversibility, then, of the position and mani~
festation of the F particle is indicated by the occurrence of Hfr mutants which are not
disinfectable by acridine orange, which will not transmit F particles contagiously,
but by their capacity to revert back to F* strains which will.

To the cycle of facultative participation as an extra chromosomal or as a chro-
‘mosomal element, the term episomal cycle has been given, and F itself can be called
an episome. The episomic character of F which is revealed by the reversible altera-
tion between F* states and Hfr states may algo play some role in the normal low residual
fertility of F+ by F- crosses. What may be proposed here is that in order for an F* cell
to be fertile there must first be a transient association of an F particle, originally from

the cytoplasm, with the chromosome. This association will then confer upon the chro-
~10~

mosome of that cell the same capacity for movement into an F- cell, as is stably repre-
sented in an Hfr clone. However, owing to the infrequency of the fertile cells in popu-
lations of F* clones it is difficult to make categorical statements about the genetic basis
of those rare events.

| F is but one of a number of particles having episomic characteristics that have
been identified in Escherichia coli, K12. Another of extraordinary interest for some as~-
pects of genetic analysis is a bacterial virus called lambda. When originaily isolated,
strain K12 of Escherichia coli showed no overt evidence of carrying a bacteriophage,
and it was handled in the laboratory by many investigators for many years before it was
realized that this was so. A necessary condition for the detection of this symbiotic virus
was the natural unexpected occurrence of a sensitive mutant strain which had, for no
evident reason, lost the bacteriophage. Having once been established and recognized
as such, these lambda-sensitive, lambda not carrying mutants, act as indicators for a
bacteriophage, which is now known to be produced with some regularity by cells of E.
coli, strain Ki2. This association between a bacterium and a virus,which it can potentially
produce as an infectious particle,is called lysogenicity. Lysogenicity has been recognized |
for many years a8 a property of a number of bacteria. In fact, its recognition dates al-
most to the original discovery of bacteriophages by d'Herelle. Having once established
the necessary cultures to determine whether or not a given clone was or was not carry-
ing a particle of lambda, it became of interest to study the way in which this trait itself,
the trait of lysogenicity as contrasted with sensitivity, was inherited. And accordingly,
crosses were set up, again using the basic designs that have already been mentioned,
to determine the inheritance of this particular quality. I might add that the probable ex-

pectation for the results of such experiments was that lambda would be found io be an
“au
extra chromosomal factor, in the hope, or the expectation, that we would here have a
system where we have living in the cytoplasm of the cell some unusual symbiont which
was capable of occasionally being activated to kill the cell and then be released.

It was, therefore, somewhat surprising when the results of mating experiments
between lysogenic and gensitive strains showed that this attribute also was inherited as
2 chromosomal marker. Not only could lambda be assigned a definite place on the chro-
mosome, but it was found to be clearly linked to another marker, which is responsible
for the fermentation of galactose. And to the locus for the positioning of lambda on the
chromosome the symbol Lp was given. A definite demonstration of the chromosomal
nature of lambda as it exists in the lysogenic cell comes from the isolation of certain
clones which are non-disjunctional for this chromosome which are heterozygous for
these traits.

These exceptional heterozygotes can be symbolized by the formula Galt Lpt
representing one chromosome, Gal~ Lp® representing the other. These diploids are
susceptible to fairly frequent segregation, in the course of which one can observe the
close linkage between the Galt marker and the capacity to produce bacteriopnage which
we call lysogenicity, and the Gal~ marker and the sensitivity to the same bacteriophage
as is produced by the heterozygote iteelf, and by the other class of segregants.

The completion of the episomic cycle for lambda is only manifest when this
chromosomal factor, Lp, undergoes a stage of maturation to prodnee the intact virus,
before which it goes through a cycle of vegetative multiplication in the cytoplasm. This
event takes place quite frequently when cells carrying the Lp*t gene are subject to ultra-
violet light, as was shown by Lwoff. We do not know in any detail the events which
transpire so as to result in the release of the genic factor into the cytoplasm, but, after

an interval following ultraviolet treatment, there appear in the cytoplasm of lysogenic
~12-

cells a number of particles of the virus. This, however, ia a lethal eomplex, and the
cell must lyse and liberate intact free virus particles. As we will have occasion to
analyze in somewhat further detail, these free particles are capable of entering other
cells and there to follow one of two aternative pathways -- either to continue their
multiplication as a parasite, or as a virus lambda, or to reenter the chromosome and
thereby give again the lysogenic status. The fact that two of the most important ele-
ments of E. coli have proven to be episomic has naturally raised questions as to the
generality of this phenomenon, of genic factors being capable under some circurnstances
of entering the cytoplasm and thereby mimicking cytoplasmic factors as are known in

other organisms.