Connecticut Valley ¥ramch , Society of American 4acteriologists i¢ {5

Yade University Schoold of Medicine, December 7, 1946.

Genetic evidence for sex in bacteria. Joshua +ederberg, Jane Coffin Childs

Fellow in Medical Nesearch, and E+L+Tatug, Professor of Microbiology, Yale U.

In the short time allotted, it will be impossible to discuss in any detail
af the experiments that have led us to the conclusion that there is a sexual
phase in the life histary of a strain of 4scherichia coli with which we have
beem working. It seems best to present a single expaiment in detail, which
most forcefully illustrates the kind of evidence that has been accummlated.
The general thesis of these experiments wax is the same as that which was pro-
posed by Sherman and Wing 10 years ago, that if there is a sexual phase in
bacteria, it should be possible to show that when mixed cultures containing
organisms differéng in various cmracters are studied, new types will be found
which are characterized by recombinations of those characters. The strains used
by these authors were too variable for them to draw any conclusions as to the
occurrence of sexual fusions, and other authors have had similar difficulties.

Several types of characters have been used in our investiagtions.
Reokiompogperk They habe all been obtained as mtations in a single
strain of E, coli (strain K~12) .The spontaneous variability of eachbof these .
characters has been studied intensively and it can be anticipated categorically
that none of the phenomena which I am bout to describe can be accounted for
on this basis.

Most of the characters used are nutritional requirements, i.e. the need
for a specific growth <factor, including such compounds as biotin, thiamine,
threonine, leucine, and other vitamins and amino acids. ‘hey are sharply de-
terminable, sihce optimal growth can be obtained in the presence of the

spe€ific reqhirement, whereas in its absence growth does not take place.

wre ob
The requirements are—feund—te-originete- by mutation, one at each mutational

step. By ultravéolet irradiation of a stwmin which is already a biochamical

Miter ? multiple mutants can be obtained, and theese have been very useful
hn bles
in these studies. Other characters which Wave been ‘used are resistance to a

specific bacteripphage, and the inability to ferment lactose (as indicated on
Eosin-Methylene Blue~Lactose plates such as these.)
Character recombination in E. coli can be detected by the appearance of
tehld types in mixed cultures of biochemical mutants. Such wild types have never
been found in the individual cultures. Thextitimuudpodmaduonumomicdeom wild types
can be detected readily since they form colonies in minimal agar mediun, whereasy
the original mutants are unable to proliferate in the absence of their specific
requirements. Thus although only about 1 cell in 10 million in these mixed cultures
is a wild type, they can be readily detected by plating about 500 million cells
after careful washing, into a minimal agar plate.
If these wild types are the result of recombination of characters, wer tree
colonies should also show recombination of other character differences present
in the ‘parental strains+' This has been found to be the case. R
Herts Tab L-Ay- bac -T.> BOM -T+ Lt B,t Lact I,
Waemx By plating Y-53, which is.. and Y-10 into agar containing thiamin,
but lacking biotin, methiaine,threonine or leucine, the colonies which appear
should represnt those cells which have recombined with respect to B.M, T and L, as
indicated on the boardg. One should also expect to find that the other 3
charecters should recombine also, namely B=; Lac= and ™ . There are & possible
recombinations of these charatters, as indicated, and the ones which are part of
the papental complexes are encircled. All 8 have been recovered, as can be demonstrated
by this plate, and these tubes. The figures on the board reprent the relative
frequencies with which they havebeen obtaiend,The disparity in theg¢. nukbers
is of coupée,of great interest; some evidance has been obtatned that this disparity
may begxplicable oh the basés of the same kind of factorial linkage that characterizes
echt organi sms.
Egidence has been obtained that in mixed cultures coridftining more tgan 2 types,
recombination takes place between only 2 types of aell at a time, strengthening the
hypothesis that these data are t be explained by a cell fusion and the orderly

segregation of genes.