Synchronous Cells
My search for synchronously dividing E. coli began in late 1958.
Unfortunately, it took 4 years to come up with an acceptable technique,
another 4 years to finally figure out an optimal way to
apply it, and two additional years to generate and correctly interpret
key data on the division cycle. What an unexpectedly long
and difficult experience that was. Of course there were personal
milestones along the way, including the births of sons Charlie and
Michael, but without sheer stubbornness in the face of endless
setbacks, my research career would have collapsed on more than
one occasion.
The first step was to choose a strain of E. coli. I don’t recall
how I came to the decision to choose E. coli B/r, but it was likely
because Aaron Novick had been running a bank of chemostats
containing strain B in an adjacent laboratory (Novick and Szilard,
1950). I was also aware of Evelyn Witkin’s radiation resistant
mutant B/r (Witkin, 1946) and felt comparing a radiation resistant
mutant with the parental strain might be useful. Also, strain
B/r appeared to form fewer filaments during exponential growth,
a seemingly advantageous property. That decision turned out to
be the first of two incredible strokes of pure good luck during this
investigation. If the far more popular strain K-12 had been chosen,
most of the crucial experiments I will mention later would
likely have failed and the baby machine technique would not have
been developed, at least by me1
.
In the late 1950s there were basically three different
approaches reported for bacterial synchronization: single or
multiple temperature shifts, single or multiple nutritional