At the Revolution with Fred Sherman

Abstract

Fred Sherman was a prominent yeast geneticist and my mentor in graduate school. Fred passed away in September 2013 at the age of 81. In this minireview, I describe what it was like to know Fred and be in his lab from 1977 to 1982, the extraordinarily exciting time when the recombinant DNA revolution hit yeast genetics.

INTRODUCTION

I was at the revolution during my graduate research in Fred Sherman's lab (Fig. 1). Fred passed away a few months ago after a long and successful career at the University of Rochester School of Medicine. He was a king of yeast genetics in the 1960s and 1970s, having developed forward and reverse screens and genetic selections to obtain mutants and revertants of the yeast CYC1 gene, encoding iso-1-cytochrome c. By collaborating with John Stewart, a biochemist who sequenced revertant proteins of cyc1 frameshift and initiator mutants, Fred confirmed the genetic code for proteins in eukaryotes (1). These and other fundamental studies over 20 years shed much light on eukaryotic mutagenesis, translation, protein structure, protein processing, and posttranslational modification. But the best was yet to come.

FIG 1.

Fred Sherman, center, having a good time at the author's Ph.D. thesis party in 1982. Members of Fred's laboratory shown are (left to right) Ken Zaret, Tom Cardillo, Robert Laiken, and Steve Baim.

Considering the amino acid sequences of frameshift revertant proteins, Fred was able to accurately deduce the DNA sequence of the 5′ end of the coding region of CYC1 (2). Fred then collaborated with Jack Szostak and Ray Wu at Cornell, who chemically synthesized an oligonucleotide complementary to the coding strand. The oligonucleotide approach allowed them to identify the RNA encoded by a protein-coding gene, a first for yeast (3), as well as allowing Michael Smith (University of British Columbia) and Ben Hall (University of Washington, Seattle) (4) and Fred (5) to clone the CYC1 gene itself. These initial cloning steps, enabled crucially by the prior development of recombinant bacteriophage vectors by Kevin Struhl and Ron Davis at Stanford (6), helped open up the world of recombinant DNA technology to yeast. Yeast began to replace Escherichia coli as the model organism of choice. I started graduate school and arrived in the Sherman lab in the fall of 1977, as these events were unfolding, and the feeling was electric.

Nearly 20 years' worth of a mutant collection, Fred's entire career in boxes containing live yeast strains, was a treasure trove awaiting molecular analysis. Along the way, Fred had collected a series of deletion mutations that spanned different segments across the CYC1 locus, allowing him to rapidly map the location of new alleles. Most readers of Molecular and Cellular Biology today may not appreciate that “map” in this context means to make a genetic cross, using yeast strains of opposite mating types, sporulating, and testing for recombinants that result in a wild-type protein and phenotype. “Map” did not mean using PCR to recover the entire locus in the morning, sending out the raw DNA fragment in the afternoon, and having it sequenced by a facility the next day.

Revolution indeed.

Every few weeks, Fred would get a phone call about a new recombinant DNA technique and pass it on to the lab. In 1978, Fred took the Cold Spring Harbor Course on Advanced Bacterial Genetics, taught by Ron Davis, John Roth (University of Utah), and David Botstein (Stanford), where the latest recombinant DNA techniques were brought together and disseminated among the new faithful. Fred gave us rough photocopies of the course manual, replete with imprints of chemical stains. It became a bible for how to handle phage lambda cloning vectors, grow and isolate plasmids, perform restriction enzyme digests, ligate, transform E. coli, make radioactive DNA probes, and perform the newly exalted “Southern blot” technique (7). The concept of a biotech company emerged, with some companies producing thin brochures listing a few enzymes for sale, while others aimed to make biologics and pharmaceuticals of the future. The reach of these companies in technical products and drugs now spans the globe.

The deletion mapping in Fred's lab had predicted that certain cyc1 mutations resided outside the coding region of the gene, in other words, in potential regulatory DNA sequences. Thus, we knew that there were gems in the mutant collection to unveil the basis for gene transcription. Yet my initial graduate project was to perform a classical genetic screen for tRNA mutants that could function as extragenic suppressors of cyc1 frameshift mutants. Since, in principle, many tRNA frameshift suppressors could restore a proper reading frame, prior to the occurrence of an improper stop codon, the screen could expand our identification (in the pregenomic era) of tRNA genes. While no extragenic tRNA suppressors were obtained in initial studies of E. coli and Salmonella frameshift mutants, John Roth (Utah), Donald Riddle, and John Carbon (University of California Santa Barbara) finally succeeded (8, 9). To enable a frameshift suppressor screen in yeast, I isolated a poorly reverting mutant of CYC7, encoding the minor iso-2-cytochrome c, thereby providing a cleaner genetic background with which to score iso-1-cytochrome c expression.

While I was doing yeast genetics, the surprises from the recombinant DNA world came rolling in. Beverly Errede, a postdoc then in Fred's lab (now at the University of North Carolina at Chapel Hill), contemporaneously with Philip Farabaugh in Gerry Fink's lab at Cornell (now at the Massachusetts Institute of Technology [MIT]), used cloning and DNA sequencing technologies to discover CYC1 and HIS4 alleles whose gene expression was affected by transposable element insertion (10, 11). Albert Hinnen, Jim Hicks, and Gerry Fink developed yeast transformation, where recombinant DNA molecules could be put back into yeast to test function (12). Terry Orr-Weaver (Harvard, now at MIT), Jack Szostak (Harvard), and Rod Rothstein (New Jersey Medical School, now at Columbia) expanded this by developing gapped repair techniques, whereby genes could be modified at will by homologous recombination at their normal location in the yeast genome (13). Gary McKnight, a fellow graduate student in Fred's lab (now at Intellectual Ventures, LLC), cloned an allele of CYC7 and discovered that the unlinked CYC1 and CYC7 loci were related by a circular permutation of their flanking genes, suggesting a circle excision mechanism for gene evolution by duplication and transposition (14). Maynard Olson and others in Ben Hall's lab discovered eight tryosyl-tRNA genes in one experiment, using the new recombinant DNA techniques (15). Yet the CYC7 allele that I isolated was my only useful product in Fred's lab so far; the frameshift suppressor screen had not yet succeeded.

One day Fred called me into his office and asked if I wanted to work on cyc1-512, one of the rare mutations that mapped outside the CYC1 protein coding region. We had assumed that cyc1-512 was a promoter mutation. I was in heaven. I closed my lab books on the frameshift project and started on cyc1-512 and recombinant techniques the very next day. Soon I discovered, by Southern blotting, that the cyc1 promoter region of the mutant was fine but the 3′ end of the gene harbored a small deletion. I then made phage packaging extracts, cloned the 3′ end out of a phage library that I made with cyc1-512 mutant yeast DNA, and sequenced the region “by hand,” using the dideoxy method. In those days of molecular biology, your character was defined by whether you sequenced DNA by chemistry, via the Maxam and Gilbert method (16), or by biology, via the Sanger dideoxy method (with DNA polymerase) (17).

With trepidation, I went to Fred and told him that the presumed promoter mutant was, after all these years, a 38-bp deletion at the 3′ end of the gene, spanning the presumed polyadenylation signal. He gave me the classic advice, “If you get a lemon, make lemonade.” He further asked if the deletion occurred at a repeat sequence, since from his years of sequencing cytochrome c revertant proteins, he had deduced that small deletions often occur at DNA repeats. I was stunned that he could predict that the 38-bp deletion in cyc1-512 indeed occurred at a 7-bp repeat and realized that not all of the insights from cloning and sequencing were going to be novel.

Things began to turn for my thesis. Beverly Errede helped me learn Northern blotting, and shortly thereafter, I discovered that the cyc1-512 mutant had small amounts of aberrantly long cyc1 mRNAs, indicating a termination or processing defect and instability. Yet the complex RNA patterns, using double-stranded probes, were hard to fully decipher, and methods to detect RNAs in a strand-specific manner were in their infancy. I came up with purifying double-stranded DNA fragments, performing nick translation, ligating in dilute solution to seal the nicks, and separating the two strands of each fragment on a native gel. After purification and use in Northern hybridizations, the single-stranded, high-specific-activity probes enabled me to learn that the next gene downstream is transcribed toward cyc1. Furthermore, the cyc1-512 deletion affected the terminator of that gene as well, resulting in the two genes exhibiting convergent, overlapping transcription in the cyc1-512 mutant (18).

Using poly(U) Sepharose beads, as oligo(dT) had not yet gained favor, I purified polyadenylated RNA and found that all of the much longer CYC1 RNAs in the mutant were polyadenylated. This surprising finding suggested that termination and polyadenylation were fundamentally coupled processes (18). Much later, Nick Proudfoot (Oxford) proved the coupling model with cyc1-512 when run-on transcription assays became feasible with yeast nuclei (19). A role for polyadenylation in transcription termination was later seen with mammalian genes (20), and thus it appears universal in eukaryotes.

The discovery of very closely spaced genes in yeast, with terminators presumably standing guard between them, was an emerging theme, as also seen with initial transcript mapping studies from Ron Davis's lab on the GAL and HIS3 loci (21). These early glimpses of yeast genome organization foretold the high gene density that would be revealed when the genome was finally sequenced (22), compared to the wider spacing of genes in multicellular eukaryotic genomes.

Prior to my time in Fred's lab, Jeroo Kotval, a postdoc, had isolated revertants of cyc1-512 (23). While a few of the revertants were translocations that brought in new termination sites (24), I was more interested in the sequences of “strictly local” reversion mutations and deciphering signals in the DNA. To rapidly clone multiple alleles in the pre-PCR era, I would digest the genomic yeast DNAs, run them on a gel, purify the DNAs of the appropriate general size, use the DNAs to make “enriched” phage libraries, and screen for the correct CYC1 clones by plaque hybridization. Furthermore, I found it possible to get sufficient phage DNA to perform dideoxy sequencing directly on the yeast fragment in the recombinant bacteriophage, without having to subclone the fragment into a plasmid. Sequencing off the recombinant phage was not feasible with the Maxam and Gilbert method. These methodological adaptations illustrate the wide-open nature of technology development when working in the “pre-kit” era of molecular biology.

Upon sequencing various cyc1-512 revertants, I found that one of them retained the mutant sequence and yet terminated strongly at several of the aberrant sites seen in the cyc1-512 mutant. My subsequent crosses of the revertant strain revealed it to have an extragenic suppressor of termination (sut2) (24). A suppressor at last! Later studies in Fred's lab showed that SUT2 encodes UPF1, a major component of the mRNA surveillance complex (25). Wild-type SUT2 was apparently degrading the long cyc1-512 mRNAs that harbored unusual, extended 3′ noncoding regions, and the sut2 mutation stabilized the transcripts.

Back then, Fred Sherman was constantly mapping genes, such as new suppressors, by genetic means. He tried creative new approaches, such as those developed by Susan Dutcher (Washington University, St. Louis), to allow chromosome transfer between yeast nuclei (26), and by Sue Klapholz and Shelley Esposito (University of Chicago), employing recombination-deficient yeast (27). Yet I remember when Fred told us about a phone call he just had with Charles Cantor, then at Columbia, who with David Schwartz had invented pulse-field electrophoresis and could use it, with Southern blotting, to map genes to chromosomes (28). Even though the initial technology was rudimentary, Fred immediately realized that the days of mapping genes by genetics were over and he eagerly sought the new technology in his lab. It was impressive to see how readily Fred could leave the past behind and embrace a new world.

Another phone call came in, this time from John Carbon. He and Louise Clarke had cloned yeast centromere III into a circular plasmid and showed that the centromeric sequence conferred mitotic stability in yeast (29). At lunch, which Fred had with his trainees almost every day, we discussed how this violated precepts we considered axiomatic: first, that the cloned centromere III wouldn't cause the normal chromosome III to be lost frequently during mitosis, and second, that the chromosome could be so stable as a circle, without telomeres. A few years later, Andrew Murray and Jack Szostak at Harvard invented a clever way to insert telomeres into yeast artificial chromosomes (30); these further stabilized centromere-bearing plasmids. But at the time, we were perplexed that someone would even try to clone centromeres directly. This starkly illustrated another important lesson: you can know too much.

Fred knew this well enough to embrace the recombinant DNA revolution, with everything new to learn, and derive new insights into biology.

These basic lessons have stayed with me in the time since I left Fred's lab, in 1982, and have provided a measure of confidence when moving into new fields. Fred loved science at its highest levels. He was one of the most critical scientists I have known and yet could patiently explain the complex workings in his mind to someone new to the field. This facet of his personality explains how he and Gerry Fink, and later with Jim Hicks and others, could so successfully initiate and run the Yeast Genetics Course at Cold Spring Harbor. Fred spawned generations of yeast geneticists, many of whom were attracted to yeast from other model organisms. And for those of us who moved from yeast, his wisdom has proven to be universal.