Thesis defense: 50th anniversary

Today is the 50th anniversary of my Ph.D. oral defense. The event took place in the Department of Biochemical Sciences at Princeton University back in 1974. It began with a departmental seminar. When the seminar was over I retired with my committee to a small classroom for the oral exam.

I don't remember everyone who was on my committee. My Ph.D. supervisor (Bruce Alberts) was there, as was my second reader, Abe Worcel. I know Uli Laemmli was there and so was Arnie Levine. I'm pretty sure the external member of the committee was Nancy Nossal from NIH in Bethesda, MD (USA). It's a bit of a blur after all these years.

I remember being fairly confident about the exam. After five and a half years I was pretty sure that everyone on my committee wanted to get rid of me and the easiest way to do that was to let me pass. Bruce stood to gain $3000 per year of research money and Uli was going to get back the basement of his house where I had been living for the past month after getting kicked out of the married graduate students housing project for taking too long to complete my thesis.

The toughest questions were from Uli Laemmli, which should not come as a surprise to anyone who knows him. He has this annoying habit of expecting people to understand the basic physics and chemistry behind the biochemical sciences. Fortunately, my inability to answer most of his questions didn't deter him from voting to pass me.

This photograph was taken at a party that evening. I look pretty calm at that point but this may have had a lot to do with the various refreshments that were being served.

The amazing thing about the photograph—as I'm sure you all agree—is how little I've changed since then—apart from a haircut. :-)

Back in those days we didn't spend a lot of time writing a thesis. I started in the middle of January and the entire process of writing and defending took nine weeks. My thesis was bound and delivered to the library about one week after the Ph.D. oral. Committees didn't ask for minor revisions and modifications because it wasn't important.

The second page of my thesis has only three words on it. It says, "To Leslie Jane." This is my wife and partner—we've been married for 56 years. She really should have her name on the cover 'cause I couldn't have graduated without her. Typing my thesis was only one of her many contributions. There are 257 pages in my thesis and she typed every one. As a matter of fact, she typed them twice, one draft and then the final version. Keep in mind that when I say "typed" I don't mean on a word processor. She used a machine called an IBM Selectric but it wasn't a computer. It was an electric typewriter. (Google "typewriter" to see what it is.)

The figures in my thesis were all hand drawn. Back then we did not have to spend inordinate amounts of time learning how to be artists using complicated drawing programs. I've included one of the figures (below) to illustrate what I was doing during those five and a half years.

The Alberts lab was interested in DNA replication during bacteriophage T4 infections of E. coli. We knew that replication was carried out by a complex protein machine that assembled at a replication fork but we didn't know all the players or what they did.

The T4 proteins required for DNA replication were known from genetic studies. The most important genes were genes 30 (ligase), 32 (single-stand DNA binding protein), 41, 43 (DNA polymerase), 44, 45, and 62. The products of the unknown genes were called 41P, 44P, 45P and 62P.

We wanted to purify and characterize those proteins; my target was the product of gene 41, or 41P. I identified the protein on an SDS polyacrylamide gel using the technique developed by Uli Laemmli who just happened to be in the lab across the hall. I labeled the proteins made after infection of E. coli cells by wild-type T4 and a strain carrying an amber mutation in gene 41. As you can see (right) there's a protein missing in the mutant and that's the product of gene 41.

We had a cool assay for enzyme activity, developed mostly by a postdoc in the lab named Jack Berry. What we did was to prepare a cell lysate from cells that had been infected by bacteriophage carrying an amber mutation in one of the genes. This lysate could not support DNA synthesis, as measured by incorporation of ³²P nucleotides, unless we added back the missing component. This is the basis of an in vitro complementation assay that worked for each of the unknown proteins.

In my case, I used traditional protein purification methods to isolate fractions of proteins and then tested them for activity in the complementation assay. The figure below shows the elution profile of proteins bound to a hydroxylapatite column. The peak centered on fraction 61 is the activity of the complementation assay. It indicates that 41P elutes early as a sharp peak in the elution profile.

The complementation assay doesn't tell us anything about the function of 41-protein, only that it complements an extract that's deficient in 41P. Strictly speaking, it doesn't even tell us that the activity is due to the product of gene 41 since it could be something else that complements in vitro.

Fortunately we had another way of identifying 41P. I started my purification with extracts from 17 liters of infected cells. To this I added extracts from cells that had been labeled with radiaoctive amino acids. One batch was from a wild-type infection where all T4 proteins are labeled with ¹⁴C amino acids. The other batch is from an infection with an amber mutation in gene 41. In this case every protein except 41P is labeled with ³H amino acids.

You can adjust the settings on a scintillation counter so they distinguish between ¹⁴C and ³H but there's some overlap. The equations for calculating the contribution of each isotope in each window are relatively simple. All you need are good standards to get the distribution. One of the most fun things I did as a graduate student was to write a computer program (in Fortran) that did these calculations automatically and plotted them on a plotter. This was back in the time when computers were housed in large separate buildings and required dozens of people to look after them.

If you look of the elution profile in the figure you'll see there's an excess of ¹⁴C over ³H in the same fractions where the complementation activity is located. What this means is that the wild-type extract has a protein at that position that's not found in the am41 extract. It's another way of identifying the product of gene 41.

I nailed down the size of 41P by running some purified protein on an SDS gel then slicing the gel and counting the radioactivity in each slice. In the figure on the right you can see that there's a big peak of ¹⁴C in slice 16 and hardly any ³H. This identifies 41P and it migrates at a size of 57,000 daltons. (We had to make up our own protein standards to calibrate size, imagine that!)

The double label technique was useful 50 years ago but nobody does it anymore. It was fun while it lasted.

I started my postdoc in Switzerland about one month after my thesis defense.

(I never did figure out what 41P did during DNA replication but a few years after I left, a postdoc identified 41P as a helicase—an enzyme that unwinds DNA ahead of the replication fork. The enzyme is now called gp41 for "gene product.")

---