An interview with Dr Henry Erlich
August 20, 2024
To celebrate International PCR Day, John caught up with Dr Henry Erlich and asked him some PCR questions.
It’s my pleasure to ask Dr Henry Erlich questions about the early days of PCR. While Kary Mullis was working on PCR, his boss also charged Henry and his group at Cetus with working separately on PCR as well – to help boost a diagnostic test they were working on for sickle cell anaemia. They would ultimately perform all the early published results on PCR, both in the original 1985 paper and the later 1988 paper using Taq polymerase. Yes, PCR started before Taq but more on that from Henry.
Do you remember your first thoughts when you first heard Kary talk about his idea?
I thought it was an interesting idea….and a surprisingly simple one. I wondered if the primer binding would be sufficiently specific so that the exponential reaction would work as conceived, that is, producing a discrete target DNA fragment of a predicted size, defined by the 5’end of the primers.. We decided to test the idea by attempting to amplify a 110 bp fragment of beta-globin from human genomic DNA. We used a cell line that had a homozygous deletion of the beta-globin gene as a control genomic DNA template.
There are reports of the experiments your team ran that essentially made the method work. Do you recall those first improvements that made you think ‘we’ve got something here’ ?
The critical element was trying to detect a “signal”, however, faint, amidst the “noise” and then varying the conditions of the reaction to enrich the signal. We initially used an E. coli DNA polymerase (the Klenow fragment). Randy Saiki tried varying the reaction conditions (temperature and time for primer annealing, extension and denaturation, primer and polymerase concentration, and Mg and KCl concentrations) but analysis of the reaction products by gel electrophoresis showed the same smear (no discrete band) with the genomic DNA and the cell line control….so, apparently, no specific amplification. But then, we used a radioactive probe specific to beta-globin and did a Southern blot and, under some conditions, were able to see a faint band on the autoradiogram. So, without the probe, we would have been unable to get the idea of PCR to work. Once we could detect a signal, we could vary the conditions to optimize the amplification. With the E. coli polymerase, however, only about 1% of the amplified DNA was the target, beta-globin. Sill, we achieved a huge enrichment (around 10e7) but around 99% of the amplified DNA was non-target. Of course, Taq polymerase changed everything and we then could detect a specific band without using a probe.
PCR with Taq polymerase: more specific, efficient & automatable
PCR has enabled so many disciplines as a highly-flexible method. Is there a development or utility of PCR you’re most proud (satisfied?) of having helped come to fruition ?
Probably, the applications in forensics genetics have been the most rewarding and satisfying. We did the first DNA criminal case and the first exoneration. One exoneration involved a man who was scheduled to be executed days after I reviewed the data and filed my report (Earl Washington Jr. case) . The other application I’ve found fascinating is the analysis of ancient DNA, like Neanderthal and Denisovan. These two apparently disparate areas are the topic of my new book Genetic Reconstruction of the Past: DNA analysis in forensics and evolution.
And we’ve got some copies of that book for prizes for International PCR Day. So can you describe the type of meeting that was ‘the PCR Club’ at Cetus ?
That group meet weekly to design experiments and discuss experimental results. One group was me, Randy Saiki, Norm Arnheim and Steve Scarf. We were pursuing the use of radioactive probes to analyse the results from genomic DNA. The other was Kary Mullis and Fred Faloona who were analyzing a plasmid containing the beta-globin gene without the use of a specific probe.
Many of the early PCR team from Cetus went across to Roche with its acquisition of PCR. What was the difference in team culture or attitude between Cetus and Roche ?
As one might expect, the culture in a large international corporation was more formal but Roche Molecular Systems, because it evolved from Cetus, was more open that one might imagine.
Which was the bigger leap for making PCR practical? Taq polymerase? Or the thermalcycler ?
The thermostable polymerase made the PCR much more specific and efficient and, of course, automatable. There would be no thermal cycle without a thermostable enzyme. But they were both critical in making PCR practical.
Did Taq really work first attempt after it was purified ?
No, Randy had to define conditions under which it would work
And finally, do you still keep in touch with any of the early PCR team from the late 80’s ?
Yes. We are bonded by our history and experience.
