April 350, 1953

Dear Bruce:

Your letter of April 9 and the two halves of the ms. all arrived
during the past week, the firstnamed marked "insufficient postage for air-
mail", which accounts for what may seem like an unconscionable delay in
my reply. Norton visited a couple of weeks ago, and we are in substantial
agreement in our views on the ms., in particular that while we do not wish
to delay you any further, the ms. can still etand considerable condensation.

My own view is that you have met every major issue I may have raised ( barring
any further strong comment below), and will now leave further questions that
may be raised to your own judgment. If the @itors of JGM are willing to

accept the paper as is, it may well be left ak that; if not (as I would etrongly
suspect) you may have a job of surgery etill on your hande. At any rate, there
ie every reason to get this in the press. We should look ahead to the problem
of reprints, as this may have to be decided rather euldenly when proof comes.
This may seem exaggerated, but I think we have to anticipate a demand on thie
side of the atlantic of at least 750 reprints. Of these, Norton should have

2003 I expect most of the stateside requests will he directed here, and it would
be rather silly to forward them to London, but if you would prefer to Kandle ali
responses to request cards, you could cut my figure from 750 to about 550. My
general mailing list takes about 350-450. Before making a definite commitment,
I had better have at least a rough estimate of the costs. If you can get this,
plus your own estimate of how Many you expect to get yourself it will help on
this recurrently vexatious business. (I might add that Norton and I shared 650
reprints of the Z&L, and the supply has long bince been exgh exhausted, except
for a contingency reserve (for future students, etc.))

To take up your letter first (let me parenthesize that I don't hope
to get an overall view of a letter like this, and must follow your own point
by point. Anything not meitioned has been noted and is, presumably ok.

"3" Don't pay much attention to the ones I remarked! they are essentially a
random sample. The address, acknowledgments seem ok. Treatzent of transduction
ve. K-12 story is ok, as is. Trandduction ie defined at p. 681 (2&L) aa
"genetically unilateral transfer in contrast to union of equivalent elements

in fertilization. The working hypothesks that FA is an agent of genetic
transduction provides....", and in Physiol. Rev. 32:413 as "restricted transfer
of genetic material to the cell". wI¢ you want to see why I emphasize thke ter-
minology, see Dobzhankky's comment on E-Taylor as "progress on the road towards
the induction of specific mutations in specific genes" (Amer. Natur. 87:123) which
propagates the error in his monograph. I fully accord with "pneunococcus transfor-
mation" as a designation for that particular case, for historical reasons, and
will have no complaint if you do not explicitly subeume this under transduction,
60 long as PnT is not contrasted with Sal. Traned. I pose the hypothetical
question: what happens if we succeed in extracting the genetic fragments

from the phage particles, and can inject them by some other means? The

concept of ¢ransduction as the overall mechaniem is the only one sufficiently
general to cover the whole situation. I regret my error in calling the

transfer of F state us a transduction. Perhaps I can wiggle out of it by

setting aside “genetic transduction" as the distinctive term, and leaving
traneduction to its dictionary meaning. There is ebyiously no common ground
betweem the F transfer and fertilization (from a genetic point of view) that
requires a contrasting terminology.
I was delighted to hearaabout the Glasgow strains. If anything, the agreement between
genotypy and lysotypy should be emphasized even further, aa belstering both. How about
speciffing NTC 3047 for Glasgow 0.

mportant) ** In view of the current hassle over Salmonella nomenclature, I think it would be
most hazardous to describe species (cf. Joan Taylor and the British Enter. Subcom. in
the Int. Bull. Bact. Nomeh....) Why not just serotype, type, or serological type?

"gene" 1s taken too seriously by some; "genetic factor" is lesa insistent as an
absolute unit, and to my eare just sounds better in the absence of a complete discussion
of the "gene theory?

"combinatorial" still sounds adequate. You have tried every combination of bacterium
with FA! there is mex no question of permutation with non-equivalents, to be fussy. IE.
a x- b is the same as b x- a. Dictionary uasge also gives, e.g. “combinatorial analysis=
mathe study of permutations and combinations."

I don't dislike your suggestion about Fla,--H,+ I am suspicious of it only because
it is too obvious. To talk about predictions, would would think that H.--H., would be
even more likely linked, but there is no sign of this. The present version fs sufficiently
cageye Would you prefer to quote the more general discussions of pseudo- and para-alleles
(References: latex,s—when-i-get—beek-te-labyTaku-Komai, Amer, Nat. 84:381; Laughnan, more
recentiy in the same; several papers in CSH 51: Benner, Lewis, Stephens, Giles, Pentee--
the firet is probably sufficient). Frankly I am not yet entirely convinced that this ie
more than a coincidence, even considering SW-553

I have a record of motilizing: SW970--x SW545, and SW972--xSW541. We should dewelop
other markere in all the atendard testers. May I suggest you do this for guy TM's , and
I'll keep the others in mind. I've gotten only 4 phases from pullorum, gallinarum --x
#01. This doesn't mean much. Altogether, using 543Fla¢ --x SL13, I've gotten just one
or two ewarms (both a), nothing with Fla,- --x SL13. However, in the course of some
track isolations, I picked up some derive. of SLI$ whieh may show a higher frequency
of transduction. If se will send you these (and repeat linkage teets). These experiments
were designed to see whether tracks were transductions initially abortive, or crossovers
with a residue, o.g. distinguashing SW666 and 3W553 as Fla, and Flaj,, tracks were picked

from 1- H,> —x la- H,8P 40 verify whether all the tracks were ctill la- H,5P, or some

possibly l- or la- H,» (temporarily motile by a residual crossover fragment ). In quite
a large experiment, involving all the feasible combinations, no such crossovers were
found. However, some of the tracks fram TM2--x SL13 seem to be more amenable ( possibly
simple selection for better transinducible cells), to subseugent transduction when they
were tested. The experiment was motivated by Morse! result, who has found that some of
the unstable Gal+¢ from Gal,- --x Bal)- aplit eff occasional Gal,© as well as Galy-.

your p. 4 con'd: Yes. M2 --x SW971 gave gn. (Culture unrelated to 970, 972). 970 and 972
may possibly be the same, an tracing history. 972 come from Kauffmann from Floyd from
fresh eggs in Cairo. Both are gm.

I would prefer TM2, just as a strain label. Norton agrees. I would indeed like to
hear details about origin of Ol can't find it in print, and have been meaning to
ask Felix. (Met Anderson at Urbana last week: he didn't know wither. Anderson will be
here in about two weeks. I might interpolate that we bought a tape recorder from him
which he picked up in NY, and later found could not use on train electric circuits).

Wunderbar! on track cell.

We have abandoned hope of going to Europe this summer. Have no preference or
objection to whatever you might like to present, joint or separate (if former is hased
on this paper). I am going to ask Cavalaigsince he hae sought my advice) to invite you
to give a longer paper af Rome. As to MGB, however, I dissent (but will not insist).

I just don't see any point to a preview which is going to come out in full detail. Please
don't quote my ow past sina, but I have become rather sour about MGB, which is now
neither fish nor foul (private circular vs. publication).
ne

I agree about leaving out S¥553, 970, 972. You have to atop somewhwre.!

SW35 = S. stanley, Edwards #15. I think you or I had done M2 --x SW535 at
Madison, with the same result.

Before this letter is buried in ms. details, may I ask whether you ever streaked
out the SW684 (unstable Gal+ g£ansduction) which I believe I did send Fou some time
ago. Our own culture seems to have gone to pot (mixture of pure + and -, no Galy), and
I would appreciate 2% getting it back, if you have it. Also, as mentioned further,
I have to give up TM2 for phase variation studies. Have started with SL46 as the most
stable in your 1949 series with approx « back and forward, but will eventually want
to compare different strains. Could you send ms a batch of those for whichyou had measur
rates? Finally, have you ever looked at the S. enteritidis NTO 3045, mentioned by Schftz
I would appreciate the strain and ite history, if available. Are there any more 0 forms
floating about in NTC? The Army evidently threw out a bunchbthat Bruner collected durin
the war. Also, LeMinor recently published ene in Ann. Inst. Pasteur (typhi)-- have you
got hold of it? Which reminds me, did you ever perfect BMEX a technique for distingudsh
O and H on agar, without excessive overgrowth? We don't seem to be able to hit the right
ager concentration (plate to plate variance very high), and methoce! did not work.

I am beginning to believe that TM2 goes through 3 distinct phases: 4, 1+1,2, 1,2.
The last is rather unusual. Alec, the i+l,2 phase seems to be distinctly more motile.
than the 1. Edwards quotes it as a fairly common occurrence that one phase is much ~
leas motile than the ether, and I think I can confirm this for several cases, especially
with artificial phases like z,,. The phases huxs may yet have dietinet adaptive values
not directly concerned with their antigenicity. “~. Let me add that abortus-equi
~-x TM2 has given an itenx from which I have been unable to ‘solate anybhing else. The
i agglutination 1e usually delayed (even with cultures passed through enx serum), but
both the i and enx agglutinations seem complete. I have some microschpic studies under
way to check on this phase confusion.

~ae ae Now the paper,

Z note your difficulty with species vs. serotype. I see no reason not to use the
binomials, but to refer to them as serotypes (without making any point of it).

There are two things, generally, which dilute the*paper (aside from a prolixity of
style which is entirely a matter ef taste). First is the adoption of a duplicate terminc
legy, one bacteriological, one genetic, with the terms repeatedly apposed. I think the
latter can be dropped, or oned defined used to the exclusion of the former. Second is
a repétition of general statements about the transduction of individual factors, the
divorcemenét of phage from FA, the traile as abortive traneductions, @hater alli. There
are often gocd rhetorical reasons for such repiti¢ken, but the writing here reaches
such length that redundanckes should be excised. These may be mentioned particularly
below. Some of the experiments are fiven in excessive detail, e.g. the method of
preparing phage. But as indicated before, theese are items most of which can be correcéé
(as the editors may well insist) on the advice of the referee.

Specific items are cited by page and om. from top of page.

43:20 customary for easy. 5:1 serotypes/eptes specios.ceses.

5:19 “extent of flegellation"— what hae been measured is usually H-agglutinabi lity,
or motility. Can you document these as mutative? (exeeps the clow spreader ty
is explicitly given as having normal flagellation). Do you have in mind your
masked H?

6211 my own findings support phase varietion as a sort of reversible differentiatic

I may be embarrassed iater at thie phrase, although it is compathble with the loosest

denotation of "mutation". Would you be willing to delete "the process consists of

mutation and reverse-mutation", and-sub which does not tell your bacteriological reade
very much, and substitute variation/mutation in statement about rate? Alternatively,
you night have to qualify your meaning of mutation, which would be awkward for what
worth.
6sbottom TM2/L#2... 7:23 a miner example of redundancy! "excessive dose phenomonen" is
superfluous, reference sufficient.
fliagelliated
20 8:10 this phage attacks many Salmonella strains, regardless of their serotype, but
only when they have flagella.

10:16 -agree-reughnese-dees-net "subculture" is not quite clear enough. Exphasize
nugerous single colony isolations. Ex: "“Gentrery-te-expeetation— Extensive
single colony isolations from flimres crowded with microcolonies alwaye fave
stable, motile subcultures similar te those obtained...."

10:23 I agree that roughness does not explain flares. However, since flares are found
when rough motile bacteria are inoculated, the flares have no definite connection
with transduction, and therefore do not need to be elaborated on here.

26: has/have
those-—of-
15:17 typical of/the species (or serotype) is perfectly correct, and lees clumsy.

16:32 and elsewhere. How about ¢ for diphasiv variation (etter than ; which I see
fe in your table).

1831@ flagellation provides a valid method, which may be of practical value, for
determination—ef-species typing stable O strains....
Does not have to be keyed to differences in pathogenicity, which are not
entirely reliable anynow. If typing is of practival value, so is this and
no apeclal justif. required.

18:26 etff. How about the subjunctive! lacked etc.

19:30 Lederberg et al 1951 or mich better Lederberg, Geneties in the 20th Centary,
for the several Lac loci. Not Lab.

20:20 I'd rather not expresa a judgment of propriety. ‘Three strains were determined
to have the Vi B(V in the Kauffmann-White scheme) antigen.
a gain
20: I find it more a strain te postulate the double coincidence of 2e68 of V and
~¥eeurrenese infection with A2 than the recurrence of Fla~. Your next to last
sentence is fine. I would delete the last. (Can you document the variation in
V, independent of IV XII? It should be demonstrated experimentally in this
strain te support the hypothesis.

23:7 the same/a singleas+7/ 14 not

25:27-8 suggest that strain Glasgow could.... (past tense made me think you were
citing Schutze, at first reading)as

28:12 a vaccine is usually understood (in US: as a modified "virus" used to elicit
protective antibody. how about “as agelutinogene (or antigens) for the
production of diagnostic reagents.*-- this includes contingency of use as
es a diagnostic antigen as well.

50; I thought SW-553 was out. 2-~31 rep.
31: 13 occurred only excepthonally 32:P2 document.

531 the concept of fragment-transduction should be made exep—-3i explicit at the
very beginning; otherwise reader gets to think ef transfer of"single genes" and
the linkage gets to be a shock. Norton and I had some disagreement about this, but
I have never believed in the reality of “single genes"as physical unite, least of
all in transduction. See 2&L 695 (Also see New Yorker, 29(10):102, 4.25.53).
35-bottom. Lac), / Ladys

B3 et ff. linearity hs overemphabized. For the first-ebté evidence of it we should
have to show that in a 3~-factor ous enly one of three possible arrangement’.
48 consistent with the results. there is some hope of this in SW553--X SW666,
but it is not very strong. The geometry of the genotype dees not have te
be specified now. You have said all there ie to say whem "in all organiems
+eethe genes behave as if arranged in linear groups, which corresvond to
chromosomes (in every thepoughly studied organism)", and a similar orgeaniza-
tion may be predicated for Salmonella.

Z 34 Divorcerent. (Cf. p. 11). It was also noted in —x S. typhi but not mentioned
by ZL
Trails cf. 12
Double transduction Of. 26.
Why not summarize in one line, and refer back, as major bearing on
transduction per 46.

B5:12 "gene structure” ie awkward( do you mean internal structure?)-— genetic
structure, or more explicitly, shromosone

116 «"physico-chemical" 48 pretty fancy. Why not "The vector of transduction in
Salmohella is (evidently) a phage particle", and not misaéad anyone into trinkin
the phage is the active agent, rather than passive carrier.

36:P1 Actually Griffith claimed that Rigéxijponoxaremierntedxigxx S)--x R,

gave Ss. as well as S_. This dees net seer to have been followed up.

y
Actually, not a great many different R straine have been used In Pn. and these have
always been selected for absolute stability, which may have sorething to do with &t.

(Norton said something that suggests Hotchkiss may be running into this again).

P2 E-Taylor's euggeetion of a linear arrangewent does not bear discussion.
It does not strengthen a case to quote insubstantial evidence. Dele "The data sugges
ted.e..s" I would aleo dele the nest sentence unlegs you want to discuss this
casulstry.

37:7 the genetic fragyents /zenes. 37:11 Natdonel Institutes. Dele “Agr. Exp. Stes

That's all for now. Forgive the ferocity, which 4s only a conditioned reflex to
mee. these days.

_ Sincerely,
' ,

i ae & te
4 Yoshua. Lederberg

P.S. Dr. Dixie Lee Ray (U. of Wash.) who spent a few weeks at Madison has been
doing some interesting things en the agglutination of motile and paralyzed, but
net of non motile T™ by an amoeba (Hartmanella). She'll be in London latter part
of July and I have suggested she look yeunup.
sgen, Spicer lately? Ha's been quite £11
vo Sold arene aeden ee Soe Sean eho Totconttan”) 28 9
. * He} Ww y
awkward, magnetic tapes. cule be more fun. Or, if necessary but more