November 2, 1952 Dear Norton: This ie to answer yours of the 17th and 2lst. I'm sorry not to have been able to get in touch with your perhaps I would heve been able to interpret your letters more acourately if I could get in better touch with what you have been doing. The paternity of SW-534 seems now pretty well established as Edwards! 157, except that I could not recover Salmonella from the tabe you went of $6, but only some gren-positive contaalnant. I have written to Edwards for another sub. I don't think any detailed exposition of this vexing paternity question will ever bedneeded, if I can exclude, #6, as should be possible. #157 was apparently ‘iolated by Cherry 10 years ago as a second phase of a java "b:-" strain. { am trying to get the original. Its genetic behavior seems to be best explained by assuming that it derives {te 1,2 charesoter as 11 allele at Ay, the “epscific phase" locus. For example, in 157--x abony*(an well as the reoiprooal), goe gets l2:enx (sic!), rather than the b:12 that one finde from typhiauriua “ -x abony, and comparable to the L:anx of tymurt ex sbony,. 3o far, there does not seem to be any influence of the phase of the recipient celle,on the outcome of any tranaductions. Yor oxampla, a3 mentioned previously, abony! ~x typhisuriwua + Ow 2 gives b:1,2. I would dmigine that the ictivity of the aiieles at A) and A, are mitually exclusive, and that the confused eal] that becomes Ay" Ag” by transduction aust auky some choice, oniy one outcomes being immediately selected by the serum agar. One can drag the cytoplasm into the story by assuaing that it imposes the state on the locus in such a way tht the latter ia carried over into the new sell, and theks acts in turn on the cytoplasm. So far, I can fing nv nved to invoke atates other then local. Since the recipient cells are alaost always contaminated with alternative phases, the work will have to be done mre precisely to see whether the phase of the recipiant has any effect on the efficiency of tranaductions to the tro phases. Most of the experiments eo far have been defective by reason of too dense incculetions, whish overgrow and inhibit most of the transductions before the latter can swim out. With mach lighter inccula, che results ecems moh batter, and more susceptible to quantitative etudy. In fast, som of the buds from i -x b:~ on beagar appear to or flarea, which just possibly aight be b-segrecants from an intermediate bji state. I agree with your remarks on the scope of the ms. that Brace is now br over. There would be nc harm in casual mention of any point that is iumedilately pertinent, but there will be no end to it if we practise free association in organizing it. Concerning the relationship of phage to transduction, the PLT22 / sW-666 (543 al system may be very useful indeed. The host adaptation of PLT22 to 34666 is apparently a mtation which persists when the "PLT22B" 1s grown agadn on LT2. It is not difficult to make S666 lysogenic for 228, without decreasing its transimiuceability (likely increasing it 2-3). Anyhow, there seems to be a definite correlation between Gal+ transduction and lysogenisation. Papillae from 228(543) -x 666 were picked, and the purified Gal+ compared with the contaminating Gal-, and with interpapillary Gal-, for lysogenicity. The closest coaparison is between the Gal+ and the Gal- from the same streakings. There did not seem to be any correlation (the main point of the three point comparison) as would indicate a common descent of individual + and -. The results give the. following table: Ip” Lp 3 which gives a X“ , p between Gal+ 18 13 05 and .0O1. GQal- § 15 Any bias seems, however, to be against a difference, as the Gal” may well have been reinfected hy the Gal” in a few cases. If one adde the interpapillary Gal-, the difference is very marked, as the incidence of Lp” there is only about 10%. A better control is still needed, however, a comparison of added, marked Gal+ Lp® recovered from the dame plates as the transductions. The results already point to a distinct corralatiog af the induction of lysogenicity with transduction. Whateve: this is due to, and especially if it is a matter af insufficient phage to saturate the plates, 1t does point again to FA=phage. With most other systens, the effdelency of lysogenlsation 4s too high to do this experiment. See Po. @ points that I am wrking on now are a) the determination of phases, as meti- tioned; 5) the curiows geastics of #157; °) the PLT22zB story as apove. and d) the theme determlincvion of b71 ratio in the Linked transduction. The two possibilities that are still unsettled for the latter are that: a) fragments from an SW-543 derivative are clwaye larger than from LT-z, or b) the selection for Linkage in the first transduction (tymur -x 543)to give 1 has resulted in a tighter association of the linked lock in subsequent transductions, You wili recail tuat typhisurium -x 543 gives mostly b, while the(tymer =x 543)4 whose Fa's venaylor should settle thie gues ion, “§ phoer! has sons oup Fr Re3ee new; but they ‘aosorb poorly aft ap— parently Jo nit transduce. More are being tried. Zam glad you have cleared up the SW541-565 story; are you sure of it? I was fairly sure I had purified 541 befowe setting up the mutant Isclation; perhaps it is unstaole., Also, some transductions LT2 -x 565 are full +/:'1 am not sure & know what lgtic variant you are talking about. How do you define a lytic variant? We thought once thet lanbda was lytic for 123, but it turns out to have been "host—modi- fied". Have you answered the question of your 5$h paragraph, L1/17 ? Your respreading experiment is interesting. Did fou not get some similar results lagt Spring? How about a more detaiked account of this one: did you study Gal+ and 5S simaltaneously, or anly the latter? ‘he delay in Si action is expected, judging from killing curves, Newcombe's failure to find evidence of phenomic lag in spontansous mutation, and the behavior of mmgamggumk segregants from S"/3*® in E. coli. You can hope that phenomis lag does not influence the transinduction counts. It might show ap as an unexpectedly rapid increase in ST whan they do come through. it would be interesting to look for this using 5M under optimim eonditions and concentrations for dmmediate effects. Do you believe there are any phage-resistance markers in Saimnelia excepting lysogenicity and S-R mechanisms? don't quite see how replica plating can heip very much use. You can only transfer 1-10% at a time. Just what sort of application did you have in mind? The big question will be to prove that the delay in increase of transinductions is due to sagregation rather than an irregular inhibition of growth. If something of the order of 1/100 of your cells are transinduced for some character, and these are inhibited, 14 would be difficult to determine whether jus} these cells are, for som reasn, differehtially delayed.(1/100 1s guessed as 107° jeff. of tranad/trait] x 10% number of traits). P.S. An interesting sideline: A few of the SW-666 x- 543, Gal+, are apparently unstably genig. The colonies have a very rough appearance, rarely thyow wroo th-10o Lp’. This has been seen before, but rarely so clearly. I expect that early reports (e.g. Deakowlts) on unstable R-S variation may have bgen based on this sort of thing. It can be predicted too that the "rough" Lp /® will be hemolytic. Joe Bertani has notioed somthing sigilar in Shigella. I don't think these types are likely to be diploid (they do not segregate Gal concurrently in the tests so far), but may be chues to the intermediate states preceding lysogenic stability. ‘sincerely, * Joama Lederberg