At the time Roger left, the phenol business looked wonderful - we got
reconstruction results like this:
But then we discovered that the differential was due the fact that we happened to be counting mutant survival on complete
agar, and wild - type
on minimal. There, wild type E. coli gives the following:
It seems a "phenol-killed" bug is dead only insofar as colony-formation on minimal is concerned - something in complete
agar "reactive" there. (There is a small % of the population that is inevitable killed)
On rechecking mutants for killing in phenol-minimal with and without growth-factor, we found just the reverse of what Hobby,
Meyer and Chaffe state (They claimed phenol acted like penicillin.) In our hands, an E. coli mutant survived better with
its growth factor than without.
We thought the difference might be due to their using G+ organisms, so we tried a B. subtlis mutant. The trust run, we got
no differential at 0.2%, growth in 0.17., but the desired differential (better survival without growth-factor) at 0.15%!
Then we couldn't repeat it. We have dropped it there, tho I feel that there may be something we're missing. You're
welcome to take over if you can figure out what's wrong.
We just heard with much excitement, about the bacteriological velvet," and in are converting our factory here to your
rubber-stamp method immediately. Did you have to have rings made, or are ready-made ones available?
We have just worked out something of which we are
[END PAGE ONE]
[BEGIN PAGE TWO]
quite proud, and in which I'm sure you'll be interested. We are using penicillin to select for auxotropha without
scattering the clones arising during intermediate culturation; i.e. we get each original mutation as a single colony. Here's
1) Grow wild-type E. coli (9637) in minimal, dilute and irradiate in minimal killing from 108 ->107/ml. (use log-phase
2) Immediately dilute in min, and plate out in minimal between protective layers, at about 2x103 survivors per plate.
3) Incubate at 32 degrees for 7-9 hours. (Each wild type survivor becomes a micro-colony of about 100 cells, while each mutant
(roughly 50 per plate)
divides until auxotrophy is expressed. Probably at least 10 cells are found per mutant colony.
4) We now layer with penicillin-minimal solution, and incubate 24 hrs.
5) Layer with pennicillinase solution (schenlay, 1 unit /100 units penicillin, or about 100 units/plate) Place at 5 degrees
C for 12 hours
to allow diffusion.
6) Incubate at 37 degrees for 48 hours. Mark wild-type colonies (we get about 30-40 /plate.)
7) Layer with complete agar. 24 hours later mutant colonies are up, at perfectly reproducible 2 to 3 % of survivors of irradiating
i.e. about 50 /plate if you plate 2 x 103 u.v.[?] survivors. wild-types are completely absent from the 48-72 hour crop.
We have picked and tested, and are getting a variety of types. the nice thing is that each colony is an original mutation.
If you don't layer with complete. There is a slow development of mutant colonies anyway, due to cross-feedings or leaking
from pen-killed cell.
But if you encircle with single growth factor, you could pick fairly efficiently. We are working on minimizing the number
of "undesired" mutants that come up in minimal, with the goal of having to pick only a desired class of mutants.
We'll be publishing shortly, I hope, and will be interested to hear the results in case you decide to try it. What is
Please send your 1957 reprints, and keep up the wonderful work - hope to hear from you soon,
KW: Ryan; phenol killing of E. coli; P-27; penicillin method; P-15;
differential survival of some strains on complete vs minimal medium;
not analyzed. heard about replica-plating