duly 27, 1949. Dear Cavalli: - So Firat, let me thank you for your kindness 4n sending the Hfr strains, which arrived about a week ago in excellent condition. For convenience, I will refer to your "$8161 Ny” [Hfr]" as W-1033. As you may imagine, I have hastened to try to confirm your findings, and to a considerable extent have done so. My earlier "kinetical" experiments were conducted in deep minimal agar. I now comared 58-141 with W-1033 at various cell densities spread with a constant excess (ca. 3 x 10°) of 4-677 on minimal agar + B,. I was surprised to find that under these conditions, even 58-161 gives aordinarily high yields of prototrophs: ‘ | 58-161 W-1033 { X ¥-677] mlLog dilution lm-= 10? 0 1000+ “peeve , 3 nL 1000+ 4 2 200 5 he 6 - 4 58-161 does not seem to give a Linear cutoff, but one can adjudge that We1033 is roughly 100x as efficacious. These quantitative experiments are,of course, largely vitiated by the formation of microcolonies whose size probably varies inversely with the density of incoulation. However, this may depend a great deal on the quality of agar used, and on the chemical purity of the — reagents. Howbelt, I still have not appreached an efficacy of prototroph for~ mation of the order of 10%-100% as you achieved. From these experiments, also, “ I could not conclude that the difference between 58-161 and W-1033 related specifically to sexal potency. W1033 might merely give larger microcolohies as a result of syntrophic efficacy, or might be mre motile. However, stimlaged by your findings, I have tried another approach which has given astonishing — results. Earlier, I had tried to see whether complementary segregants could not be identified by conducting “crdésses" not a selective minimal, but on am indi~ cator complete medium. For example, I grew Lacj— and Laca-,(each of which is lactose negative, but which recombine readily to give Lac+, )tokethser 4n a complet medium, then plated out large numbers of cells on FMB lactose to look for any sectors of Lact, I did not find any with previous stocks, but decided to see whet. recombination with Hfr was so frequent that growth-selection might be dispensed with. The material first available was W-1033 itself, which 1s Lac+, so a slight! different approach was tried. W-1033 was inceulated heavily with W-677 into | Pemnassay broth, and then after 24-48 h. the mixture was diluted and plated on ~ various EB agars. The colonies were then tested on other sugars to look for ossible recombinations of fermentative character. One such experiment is tabulat W-1033 + W677] plated on Xyl EMB. 44 Xyl+ colonies tested on Mal EMB. 39 were Mal+ (parental); 5 were Mal- (recomb.) Similarjy, of 35 Xyl~, 34 Male (par.) and 1 Mal+ (rec.) The she apparent recombinants were then thoroughly diagnosed: | Expe 5830 Culture Lac Mal Xyl -Gal MAL Vy. Nutr. W377 ao - - ~ - r aes : i W1033 + + + + + 8 M(B] 8B not scored throughout ZL > - + - _- r L 2 > ~ + - + 8 ur 5 - - + + + r L 6 - + - ~~ - # TLE, - Thus there were 6 recombinants with respect to Xyl and Mal out of 79 tests! And what a wealth of material! ; If any of these selections are Hfr (and I have prelininary indications that some are: they will provide axcellent material for another technique cf selection, namely the use of ENB-—tvo sugars. For example, if one parent is Lac+iale and the other Lac=ifal+ then both parents will give a pure + reac$ion on EMB-iLac-+ifalj. However, recombinants may be Laceifal- and give a = reaction. I am especially interested tc use this technique to find sectored coloniss which may represent tha segregation of single aygote cells. To this end, I have also irradiated 11033 (and it appears as expected to be somevhat resistant to uv) and have picked up six new lactose= negative mtants. Gne is Lac.-, and Ismay hope that ona cr more of the others is also distinct from Lac,-, which would allow plating on TMB Lac to Lcok for Lact recombinants as mentioned abova. I have yet to re-study 58161 from the viewpoint of Exp. 5820 abive, but oyspast experiente does not readily adnit of the possibility that it will behave in a sinhlar manner. To my mind, this kind of evidence would cliach the comelusion that in Hfr you truly have isolated a more sexnally active strain. The hizh frequency of recombinants also strongly heightens the possibility that the sexual fusion may be cytclogically damenstrabla, for which project you certainly carry my best wishes. I have also just crossed 1033 into Het stocks and isolated some Lacy diploids. From this material also, 1b should be possible to select, good Hfr strains to pursue our investigations. I will send you whatever matarial comes to hand that might be useful. Just new, your letter has bean received. Thank you for enclosing the USS. May I request, first of all, that any referance to the discovery of recombination include Professor E. L. Tatum, whose role in the initiation of the work ks not adequately indicatad&izeven by his joint authorship on some of the papers. © Tour letter gewe some detailed data on %1033 x 8-677 which is not in good agreement with mine. E.G., 2.5 x LOY W583 + 2 x 167 71032 gove §10 prototroohs in your ez perinant, bet only 2 (calenlated) # in mine. The thiamin agar plates I used were not mors than 2 day old, and quite mist. I don't believes that tha difference can pe a matter of strain, but 14 is conceivable. I do not notice in your MS any referenc 4o a comparison of $3-161 with Hfr like that on the first page of thei letter, I would be axtremely Interested to learn whetherbyou have found as high an efficLency as I did with 58-16). Finally, it may be sugcested that all present 53-161 cultures are heterogeneous for Hfr. It might be worthwhile to establish a number of single colony subcultures and examine their variance in protatrcph production, I've done this once without interesting result (Cf. pe 519 Genetics 1947), but it might well be repeated. It was extremely gratifying to note the close coincidence in many details between our data on various segregations.One always has an uneasy feeling about the reproducibility of results in another's hands. . . i do not think that the discrepancies in my obssrvations on Hfr are especially in- portant, but they mizht be worth some further ingpection. Your comment on the effect of xhxmumiuz loucine (fig. 1) is readily exclained: most samples are fairly heavily contatkhnated, especially with methionine. This is usually quite apparent on the plate because the background growth is och heavier, I have never used frequency of proto~ trophs as an index of crossing-over, except qualitatively with B,. No less interesting is the discovery of a coli strain which can cross with K-12, I think that itbwould be especially interesting to ascertain the nature of the ~ genic differerges between such strains, and in particular the genic deterdination of the sorolcgical distinctions between them. Kauffmann, in Copenhagen, has rather thoroughly worked out the serology of the coli group, and I strongly recommend that you send him the parental and a mumber of recombinant cultures for serological @omparison. I have been looking for some time for other crossable strains, both in coli and in Salmonella, and even resorted tc the use of "naturel" nutritional deficiencies in the latter. However, I hava come to the conclusion that with the seloctiva metheds ncw available, especially the penfcillin methed. it is more economical in the long run to start with wall defined, easily cultivated, strains and to induce mutants in them. If a short cut is needed, I think that the use of drug-resiatanca (using several agents) as a selector is available. It has worked with K-12, In Salmonella, using bicchetical mrtants, some meagre evidence of recombination in t9phimrium has been picked up, but the same peculiarity, the persistance en bloc of parental eharabters, has plagued us. Alsc, most Salmonella are lysogenic, and mutual lysis, or suicide, has complicated the results con~ _ @iderabl.a. It showld be sentioned that a third cr morse cf ecli strains also carry latent phages (K-12 does,for exampla) and that this mst be taken into account . in interpreting "recombinations". However, unless coli 123 is unstable Lact, the recovery pf Lac- protctrophs may be taken as nearly conclusive. Concerning the oppesitionol xwharacter of the Afr effect, T am not clear whetier the cross of the dorived Ar- x By-, both of which are presumably HEr, gave the same high rate of recombination. I.7., how mech is "rather high" (p. 10). As te starting from seratch, we have already done this, and developed several hev systems of mutents from KelL2. To map them all completely, however, seans rather hopeless, hevever. The main question that I hore to answer is whether the segregatic peculiarities af K-12 stocks derend on structural heterosyoosity, or séuxiomn seme intrinsic vettern of meiotic instability, That alterations in nattern do cccur is quite clear from the unique searresation patierns tmt have lately shown up. For example, I have a TLE,- Lec— Male segregent which when crossed with 58-161 gives over 95% Lac+ and MSl+, inatead of the exectad axcess of = secombinants. You will hear fren os again on the controls of the comparison of 41033. May you enjoy your vacation, Sincerely, a Joshua Lederberz.