Remembering Carl Woese.

My first memory of Carl Woese was as an undergraduate at the University of Illinois. I attended a seminar course in which Carl was an instructor along with Professors Dick Davenport, Dave Stocum, and Dave Nanney. The first class was on a cold winter morning and Carl arrived late, looking hurried and wearing an impossibly large, puffy winter coat. I turned to my classmate, Doug Melton, and said “The Iceman Cometh,” to which Doug chuckled and replied “That was a good one.” I later asked why Carl, a microbial evolutionary biologist, was tasked to teach a seminar in the area of developmental biology. He said that the other professors just liked to hear Carl’s take on things so they might incorporate some of his ideas into their own research. Undoubtedly, Carl found an opportunity to get a teaching credit where he didn’t have to do too much work. Carl famously didn’t like teaching; he felt it took him away from his real love, understanding the origin and evolution of life.

I started working in Carl’s lab as an undergraduate and continued on as a graduate student. When I was an undergraduate senior I wanted to pursue basic research and was looking for a lab to do an honors project. Some of my classmates were looking as well and we all did what typical students do, apply to the labs of the most popular professors that taught the most popular courses. Carl, having avoided most teaching duties, was not well known among the undergraduates. In addition to soliciting professors whose courses I had taken, I also approached Carl, probably the only undergraduate that year to do so. Carl accepted me into his lab while the others did not. With Carl, I likely received a much better learning experience than any of my peers; in this turn of events serendipity was my friend.

Scientists are typically trained in a specific biological system or technology and then seek to answer questions that are approachable with their training. Carl did what great scientists do, determine the fundamental questions to be answered and then develop the tools to answer them. Carl sought to know how microbes evolved from the primordial soup, and to explore that question he wanted to trace the ancestors of extant organisms as far back in time as possible. He sought out a macromolecule that would be a proxy for evolutionary change. Walter Fitch at the time was doing molecular phylogeny with cytochrome C, but that protein is not ubiquitous among all organisms and probably evolves too quickly to be of use in looking far back (billions of years) in time. rRNA, however, is ubiquitous and evolves slowly given its importance to the basic function of protein synthesis. Thus, it was ideally suited to look back billions of years, a sort of “ancient molecular text” as Carl would say. Carl spent years setting up the sequencing technology needed to create a 16S rRNA database of common bacteria, yeast and algae, all in order to look for organisms whose ancestors diverged early after life began. Blazing a trail in this way entails risk; if the approach was flawed he could have ended up with useful but mundane data on microbial relatedness. He made a giant leap of faith and was amply rewarded with some great discoveries.

The technology to completely sequence the long rRNAs was not available when Carl started his studies in the late 1960s. But a great deal of sequence information could be obtained by digesting the 16S rRNA with a ribonuclease, releasing oligonucleotides ranging from two to many nucleotides, separating the oligos by paper electrophoresis, and then further sequencing each short (≤ 25 nt) oligo. The resulting “catalog” of 16S rRNA oligonucleotide sequences from each microbe could then be compared in order to evaluate phylogenetic relatedness; the greater the number of common oligo sequences, the closer the evolutionary relatedness. With this information a phylogenetic tree could be constructed.

In order to create a 16S rRNA oligonucleotide catalog, bacteria were grown in the presence of 32P orthophosphate after which the radiolabeled total RNA was purified and its components separated by PAGE. The 16S rRNA was purified and digested with ribonuclease T1 (which cleaves after every G residue) and the resulting oligos were separated by 2-D electrophoresis. Electrophoresis in the first dimension was done on a cellulose acetate strip. The strip was then “pasted” to a large sheet of DEAE paper and the transferred oligos were separated further by electrophoresis in the second dimension. The DEAE paper was dried, exposed to X-ray film and the separated oligos were detected by autoradiography. The collection of “spots” on each film was referred to as a “fingerprint” (see Fig.  1).

Figure 1. Carl is working on a 16S rRNA “primary” fingerprint at his light board, 1976. To the left of Carl is a “secondary” sequence analysis of oligonucleotides. Photograph by the author.

The ribonuclease T1-digested 16S rRNA fingerprint was comprised of scores of spots, each representing one or more oligonucleotide sequences. In addition to knowing that every oligo terminated with a G residue, there were two additional pieces of information that could be deduced from the oligonucleotide mobility: the oligo length and the number of uracil residues it contained. The oligo length correlated with slower mobility in both electrophoresis directions and the U-content resulted in waves of oligos, termed “isopleths” (increasing U content correlates with decreasing mobility in the second dimension).

The dimer and trimer sequences were easy to identify based on mobility but yielded no useful comparative information. The longer oligos were rich in information but their sequence could not be determined based solely on mobility, and thus each had to be sequenced individually. To determine the oligo sequences, the autoradiography film was aligned to the electrophoresis paper and the radiolabeled oligo was excised with a razor blade, placed in buffer, and a second ribonuclease was added (such as RNase U2, which cleaves after A residues, or RNase A, which cleaves after pyrimidines). The DEAE paper containing the digested oligo was macerated and pressed onto a long sheet of DEAE paper and electrophoresed in one dimension. The digested oligos were separated again by length and sequence content. For longer oligos, the digestion sub-fragments could be excised from the secondary DEAE paper, digested with a third ribonuclease, and evaluated in a “tertiary” analysis. The entire catalog would take a few weeks to complete because material loss and 32P decay necessitated longer (up to 2 wk) film exposures for the secondary and tertiary analyses.

For oligos that were just a few or several bases, it was pretty easy to determine the sequence. For longer oligos (≥ 8 nt), deciphering the sequence could be quite challenging, especially if there were runs of the same base or if the primary spot contained more than one oligonucleotide sequence. Carl was a master at piecing together the puzzle of each oligo to determine its sequence. He could often be seen jiggling a film against the light board in order to detect faint spots from a poor exposure or incomplete ribonuclease digestion. The skill and determination of deducing a composite sequence for each of the scores of oligos in a catalog should not be underestimated; you really have to believe in what you’re doing to come in every day and do the same tedious thing over and over. And woe be to the researcher who underexposed a tertiary analysis or didn’t prepare the X-ray film processing reagents correctly. If you didn’t get all of the sequence information before the 32P fully decayed you pretty much had to start over. And anything that slowed down the march to data was a problem for Carl, and thus, a problem for you.

By the time I got to the lab, there were rows of film boxes each containing all the fingerprints from a microbial catalog. Remember, much of this was work was done before computer entry was easy or even available; bioinformatics was a term not yet in general use. It seemed sometimes that Carl kept most of the sequence information in his head. I was flabbergasted when he would determine the sequence a long oligo, remember in what bacteria he had seen that sequence before, and then pull out the film to prove it.

When I started in Carl’s lab in 1974 he gave me a desk in his (very small) office, so he “could keep an eye on me.” I thought my presence would be an imposition, but I shouldn’t have worried because he rarely spent much time in his office. The real work was done in the lab, reading 16S rRNA fingerprints at his light board. He was, however, very generous with his time speaking to me, trying to teach me what he was doing and why. At the time, I knew very little science and I’m sure Carl was cognizant of that, but he nevertheless made me feel that my learning was important.

For my first project he paired me with Larry Zablen, then a graduate student in Carl’s lab. My job was to prepare 32P-radiolabeled 16s rRNA for fingerprint analysis. We would dump ~10 mCi of 32P orthophosphate into a 50 ml bacterial culture and after overnight incubation, run the pellet through a French press, isolate the nucleic acids by phenol/chloroform extraction, and separate the rRNAs on an acrylamide gel. The rRNAs were leached out of the gel and then concentrated by ethanol precipitation. What a mess that often was! The Geiger counter was always screaming a path to wherever you had been.

After Larry received his PhD, I was tasked with fingerprinting the 5S rRNAs and preparing the 16S rRNAs for analysis by Carl, postdoctoral fellow George Fox, and research assistant Linda Magrum. The first bacteria I worked with extensively were the Clostridia, doing the culturing and radiolabeling. One of the Clostridia I cultured was Clostridium perfringens, the causative agent in gangrene. Carl never mentioned gangrene to me. When I found out what I was working with I was a bit peeved and when confronted he just chuckled and said not to worry, he had gone to medical school for “2 years and 2 days” and that the bacterium would be dangerous only if it infected an open wound. I took this as an important lesson: know what you’re working with and don’t count on others to always tell you what you need to know.

After I had been in the lab for a couple of years, Carl and Ralph Wolfe’s lab began a collaboration to fingerprint the 16S and 5S rRNAs of a very diverse and unusual group of bacteria, the methanogens. The methanogens were already recognized to be a distinct group of bacteria, unique because they lacked peptidoglycan in their cell walls in addition to containing metabolic pathways not seen in typical bacteria. A graduate student in the Wolfe lab, Bill Balch, had mastered the art of culturing these extreme anaerobes. We prepared radiolabeled rRNAs from several methanogens and Carl began to decode their 16S rRNA fingerprints. It was immediately apparent that the methanogen 16S rRNA sequences were unlike any of the common bacteria he had previously analyzed. The 16S rRNA sequences from other bacteria occupying extreme niches such as the Halobacteria and Sulfolobus acidocalderius were similarly distinct. The story of how the Archaeabacteria were discovered has been the subject of other publications,- so I won’t go into that here. But it was an exciting time in the lab when NASA announced in a press release that Carl had discovered a “third form of life.” Of course, that wording piqued the interest of the mainstream press and for about a week the lab was filled with TV cameras and reporters trying to get the “story.” From the New York Times to the CBS News, Carl was everywhere for a few days. I think Carl was a bit overwhelmed by the attention, but he also appreciated that the accomplishment was being recognized. And when press took photos, Carl was generous in including me, Linda Magrum, and others students involved in the work.

Carl was my first real mentor in science and much of what I accomplished could be traced back to his teachings, even if I didn’t realize that until much later in life. For example, I typically did several projects simultaneously and worked very fast; this often resulted in sloppy work. Instead of admonishing me to work more slowly, Carl told me a story. DNA polymerases, he said, have an optimal error rate, just enough to create variation for natural selection but not enough to kill the cell. Scientists need to find their own optimal error rate as well: those who work too carefully (slowly) will be overtaken by competitors while those that make too many errors will generate nothing but gibberish. The unexpected result from a mistake can be informative and sometimes lead to a greater discovery than the experiment was originally intended to produce … just like a mutation that confers a selective advantage to an organism. This explanation made perfect sense to me and was a far better learning experience than a simple rebuke or a directive to eliminate mistakes. And, of course, it was fitting that Carl used evolution as the backdrop for his story.

I also learned much by the way Carl approached his research. Carl exemplified maybe the best quality that a scientist can have—objectivity. Over the years I’ve witnessed many very smart people who failed because they believe what they want to believe, that their hypothesis must be correct because it originated in them. Carl was always data driven. He was certainly no different than most scientists—he had an ego and always wanted his hypotheses to be correct. But if the experimental data differed from his initial bias, the data always trumped ego. He only wanted the truth, no matter the source or who did the work. I have always tried to emulate that aspect of Carl’s temperament and, hopefully, I am a better scientist and a better person for it.

For me, Carl will always be the archetypal scientist, the person who most shaped the way I think about science and do research. He was a mentor and a friend but sometimes, in all honesty, we had a rocky relationship. When I was a graduate student, we often clashed about work habits and a few years into my thesis work he asked me to leave the lab for at least a year and, if I wanted to, come back and complete my degree. I think we both knew that if I left I would never return, and there were some sore feelings upon my exit. After running out of money, I took a job as a research assistant in George Fox’s lab at the University of Houston and so kept in contact with Carl. I would return to the Illinois campus each year at Christmas to visit friends and I always stopped by Carl’s lab to say “Hi” and to catch up on what he was doing. He would always greet me warmly and share his current thoughts on evolution and I would apprise him on how my own research was progressing. I think we were both relieved that no hard feelings persisted. He was also generous in including my name on publications both before and after I left the lab; to have a populated resume helped me immeasurably in furthering my career.

Over the years I would email him a congratulations whenever he won an award and he would always reply that the best thing about winning awards is that he would hear back from all his old students and colleagues. I recently had cause to update my resume and I thought about sending it to him, thinking that he would be pleased that I had found some measure of success over time. I never got around to doing that and, unfortunately, he passed before I could. But I think he would have felt some satisfaction knowing that I turned out OK.

I think Carl’s temperament mellowed much over the years. When I was his student in the 1970s Carl was in the midst of creating a phylogeny of life and the uncertainty of success, as well as the criticism and skepticism of his colleagues, probably weighed heavily on him. The acceptance of the Archaeabacteria (now the Archaea) by mainstream scientists and the recognition accorded to him (MacArthur Foundation “genius” award and the Craaford prize among them) gave Carl, I think, some peace of mind. In most of the photos from the 1970s he typically has a serious, almost stern look on his face. But in the photos from the 1980s and beyond, he’s generally smiling and looks much more at ease. Carl is rare among scientists who have big ideas and big goals in that he largely accomplished what he set out to do. His was a scientific pursuit well done.