9/30/58
Dear Josh,
It is 5 15 a.m. - -  just the time of the solid log. phase of this damned bug.  I am trying to do your lactose exp. and on top of that a penicillin exp.  Probably two sleepless nights - -  and people will probably be bored anyhow to know that galactose protects 3096 against penicillin (if it does?!).  Mike Yarmolinsky is with me - -  but to tonight (at 'midnite' we both went to our Baltimore rooms with alarm clocks and he probably just knock his down from the table).  I hope that I can get him and a graduate student interested in K12. Your question about galactose in soma or[?] cell wall can best be answered with respect to 4-epimerase-less 3099.  I ought to send you a photograph of a pretty chromatogram sprayed with silver (not Danish silver). This the glucose-gal spot rechrom. in a pyridine mix which separates gluc. and gal wall.  There are probably 100 gamma glucose in this nice spot and not a trace of gal (0.5 gamma would have
[END PAGE ONE]
[BEGIN PAGE TWO]
been just detectable.  Hence unlike many B strains which acc. to Westphal and al [sic] contain rhamnose (main part) and gluc and gal 10 to 3 (i.e. acc. to Lelois's[?] equil.) K12 3099 has no detectable galactose.  This corresponds well with the fact that '99' grows well in a pure glycerol amm. min. med.  The other strain analyzed is 3096 (transf.-less) and it has neither detect. amounts of galactose.  I did not amount as much on the paper as on 99 and the lack of galactose is therefore not that 'watertight.'  We must do 3100 of course but I need some coworkers.  It is a lot of work, espec. the fractionations.  The triply leaky "Gal 3" interests me very much.  Phil Hartman put his bet on an inversion.  This interests me somewhat - -  because I find that the constit. epimerase moderately lowered and the induced greatly lowered (only 1% of 3100, 3096 and 3092).  Could it be that the location of a repressor gene (or inducer gene) affects the amount of induced protein synthesized? I have still frozen the 'lysates' of a 300 Liter run of '3092'.  But the agony of getting things running here has prevented me from working the batch up.  The fingerprinting is far off.  But remember that Yeast epimerase has built in DPN and is probably only 30 times 10^3 big (pr. incon. DPN).  This applies probably also to
[END PAGE TWO]
[BEGIN PAGE THREE]
K12 epimerase.  Bernie Horcetur[?] has gotten some of your strains through a "semi phony" character at NIH.  I sent him 3096 upon his request.  I don't know what he is doing with it and I don't care; I suppose you don't either.  I am looking forward to see Jacques Monal in Boston next month.  Are you coming?  We missed you at the Genetic Symposium in Chicago.  It is 5 40 - -  next aliquots.
Hope to see you soon
As ever yours
H.