January 12, 1953

Dear Bruce:

I shall send the strains requested as soon as I can get around to
it. I believe I did not send SW-926, but substituted SW-938 instead.
SW-925 is abony —x 39546; SW-938 is SW-546 —x abony. As the mother
strain in the latter was already diphgsic, it seemed to me a better
exemplification of the anomalous Hy *“ and more likely to have a

straightforward subsequent behavior. The origin of the 1,2 phase of
SW-546 is still somewhat of a mystery. I have not been able to repeat
the isolation of an enx:1,2 from abony —x 546, perhaps because the

546 would have to be ina sort of cryptic second phase for the substitu-
tion to be effective.

Are we still at cross—purposes about S¥--674? (Or do you have some
other reference for S. dublin(? ) 0 in your lettar?) Si-674 ia tyohimurium
to which gp has been transduced in place of 1; I shall be very surprised
if it phage—types as dublin. You should find the othsr charectersa cf
g§W-435 (nutrition, Xyl- Gal- S") still intact also.

I didn't intend to stir up such a hassle about the authorship of the
paper-- and anticipated no personal difficulties at all. I won't rastate
my own views, and will accede entirely to your own decisions on the matter.
It was to avoid the tiresome necessities of divided leadership and respon-
sibility that I raised the point in the first place, but I suppose it will
make very little difference in the end. I hope yon will go ahead 2s if this
is your @wn paper, though of course you will tend to the superb advice you
get later from your colleagues. At any rate, you ought not to delay the
actual writing on any account, as it is impossible to suggest revisions
until one can see the whole. I agree with you that this paper mist be addressed
to the microbiologists, and the genetics confined to the least necessary
(which in this case is also the most) to make it intelligible. I think one
could go so far as to say that the suggestion of linkaze has been partly
confirmed (or at least the st&ted alternative disqualified) by subsequent
backcross tests, I doubt if the facts yet justify a more sophisticated dis-
cussion at any level! Norton's objection on suppressors is at least partly
semantic, as to what one means by suppressors vs. Fla loci. It ie true that
the different Fla— mutants have not been tested in a proven uniform background,
and that they might show different interactions in other combinations. This
would in no case disqualify them as worthy of being called "Fla—", and
the reservation of interactions with the residual genotype should be
implicit in any discussions of gene action. In any given test, whatever
locus is altered to restore motility was, by this defihition, a Fla-.

I should rather not, just yet, send out any heterozygous diploids.
Except tossomeone who is prepared to wallow in the whole morass, their
individual behavior is so complex as to lead to serious misinterpreta-
tions. I am not concerned about your falling into such a trap, but fear
that you would be embarrassed by requests from other quarters. However,
I have been meaning anyhow to send you SW 684, Gal+/— from PLT22—x666.
This is, I think, also undoubtddly a segregation, and rather sesembles
the behavior of the E. coli diploids (except, of course, that just one
character is involved). I don't quite see what you're doing with the
iodo-eekin analogues. I'll be happy to send you some of our eosin,
leek naukenant that wenn aeimnly anita arnannd for aammnlias till vou find a
batch that suits you. If my eeein doesn't work with your Metiylene Blue, that'll
bs your next step. You may have to vary the eesin: MB proportions slightly from the
indicated ratio of 6:1, by empirical tests. This sounds like a lot of trouble, but
the result (for years to come) is still worth it.
offices

I hope we can avoid a delicate sityvation on the publication of the serology.
I have been cultivating Edwatds good/WEMM for some time, and am deeply indebted
to him for innumerable favors. We have had, I think, an understanding that he
would collaborates on the detailed serological work. I suggestbthat you get
Joan Taykor's critical date as.a sort of addendum (perhaps under her authorship)
to the present paper. In th@\téxt it might also be indicated that Edwards has
confirmed the initial diagnoses. I will have to wait to see Edwards about further
developments, but expect we will collaborate closely on further generation of
serotypes: I am hoping, in fact, that this part of the work itself will be continued
largely at Chamblee, except for my own studies which will be more intensive than
extensive, at least anent phase-variation. I do not think that Edwards' propietary
interest will ta infringed, however, by a publication in the form just suggested.
Can you think cf any easier solution? I would hesitate to suggest to two other
people that the: collaborate with rach other until I have a much batter personal
insight.

Thaks for the Garman raferences: I'11 look them uo and send you “nglish ab-—
stracts. hile we're at it, I noticed Andrewes’ comment (1925) thet vertetion
was more frequent in broth than én agar. If he hee any real observations on this,
it is hard to sce how they could be frllactous. By the by. in his 1922cnaper he
refers to his owr. infection with typhimugrium, which was substantially pure
group phase (anent p. 411 your paper). In considering such questions, the purity
of the antibody response might be even more critical than necessarily limited
isolations. But snyhow, do you think Andrewes' comment on broth vs. agar could
be right? I have always been a little sfkeptical of your explanation of binns;
perhaps these are strains which ars unusually stable on agar. At any rate, this
seems like the most tangible lead on the control of phase variation. On the
activation-shift hypothesis, experimental control should be possible.

I can't be sure whether you'd gotten mine of the 25th before you sent yours
of the 29th. To avoid such uncertainties, it might be a good idea to include the
formality of vaference to previous letters rec'd (or is this too stuffy?) For
reasons similar to yours, there has been a temporary decrescendo, and + don't know
whether we shall pick up much befors packing off for Chamblee (ETD; Jan. 25).
Some minor nuggets: The host~afaptation of PLT22 is confuded. The P8lative e.o.p.
(typhiinurium;LT2: paratyphi B SW6646) goes as follows: PLT22 itself ca. 7, adapted

to SW666 ca. -2, readapted to LI2, 4. About the same holds for the lytic variant,22V.
Several, but not many separate lines were tested for the above, with consistent
results. Ad the adaptation is not completely reversible, there may be both an induced
phenotypic effect and a sekection of spontaneous mutants. W666 itssif carries ano-
ther phage, and it is an amusing possibility that the adaptation is partly a pheno-
typic blending with this phage. Efficiency of transduction is qualitatively similar
to the r.eop., but not quantitatively. E.G., PLT22 has ca. 1% FA on 666 as on LT2;
most of the transductions here are not lysogenic, Ad.22V does not appear to

induce lysogenicity in SW666 (unfortunately, as some interesting substitution expts.
would have been possible with such a markkr), although its plaques here are quite
muddy. I have had some i's from LT-2152 x SW666; have not yet followed them through
A fairly clean preliminary expt. using 22V to identify infected bacteria was com-
pleted (see last letter). It agrees quite well with the proposition that all (or

at least almost all) transductions occur to bacteria infected with temperate phage.
As the nulber of transductions is limited by the amount of phage, it seems unlikely
that phage-infection is simply an auxiliary condition. (Some more expts. on this
point, viz. the yield of transducticns at multiplicities below saturation
with various mixtures of Gal+ and Gal- FA, may be needed. They may add up to
rather substantial proof that FA is carried by the same particles as carry
lysogenizing «ctivity, 1l.e., that PA = phage not only qua skins, but also qua
contents.

I've done just a few experiments with SL-13, and can confirm getting a and i
—-x PLT22. FA from paratyphii B has so far had no effect, and the yields in all
gases have been very small. I thought this might be due to roughness, but the
overall somatic antigen seems well developed. I should do some adsorption expts.,
and will. Para A was an important type tc decide on the role of xII,. Do you have

any explicit information on the presence of this component} in SL-13° So far, its
susceptibility to transduction has been so low as to discourage any extensive work.
We may really have to buckle down to isolating a good many new mutants (Fla-) from
somes standard strain. I've thought of LT~l (as the only LT [ found susceptible to
Chi phage,. L'il seud gou anything promising that may come of this.

Spicer's still up to attempting somatic antigen transductions. There are many
enough technica: difficultics; it stili hasn't worked. dave you done any more
With technulyues sor classifying 0 coionies¢ 1 should think one could add aaounijans
MAREXWEKAX Somstiing like wetaylcellulose to increase the viscosity of motility
ager, aid waxe it an ladicator rather tnan a selective medium. 30 far, have had
very sloppy resuits trying to grow colonies at low temperature and let them migrate
briefly at ndgher,

I sent som reprints off to you; you will recognize the semicolons in some as
your own. Your critical judgment was invaluable, and + did not want to leave this
unsaid. Page proof nas only just come in for the L&L lysugenicity opus of which
you have a preprint. In case anyy references are needed, 1t turns out Genetics
38:51-64 (dan. 1953).

Sincerely,

Joshua Lederberg

P.S, Sageblel grey 3. typhi in chloromaycetin broth. In concentrations over

1 ug/ml the cells were immotile and H~inagglutinable. The Vi response also
dropped with more than about 2 mg/ml, while O-agglutinability increased.

He did not reinoculate into plain broth! so your question's not answered.

it would be easy enough to do.He made up his stock chloromycetin in propylene

glycol, I don't know why. .
the other paper wou quoted ig in ths bindery just now.

I hope [8ve given you the proper gunealogy of SW~534. SW-+703 is Edwards! #3,
and cahhot be the ancestor, if only sn grounds of inozitol and rhamnose fermen-
tation, as well as diphasicity, Edwards #157 was represented as SW-546, and
certainly is the ancestor of SW-534 ete. I will send you SW-703, but the allegagion
that it is the parent is ingcrrect. Perhaps thas 1s what you meant, and hope to
have Felix exclude 703, as * am sure he will.

JL