For 40 years beginning in 1963 as a new college grad, laboratory technician and graduate student at NYU, and lasting until I closed my research laboratory in the early part of the 21st century, I was a biomedical research scientist. This is a brief summary of those years written primarily for a scientific audience. To tell this story, I have adopted the style used by The Society for Neuroscience (SFN) in its wonderful history of the field of Neuroscience through autobiography:
All of the work described below was done in collaboration with numerous colleagues, students and assistants. As I recall the significant scientific moments, I have mentioned those co-workers who come to mind, the ideas behind the problems we tackled, and the times in which the work was done. I know that I have left out many valued associates and students in this summary. I apologize to anyone who worked on any of these projects and feel slighted (I expect to hear from you if you are in this group). I have written this too, with, optimism; the hope that a young scientist will become interested in one of these ‘scraps’ of research findings, get as excited as we did back then when the ideas and experiments were new, pick up the gauntlet and become ‘obsessed’ (if you’re lucky enough to find an obsession) with finding the answers to any one of these still unanswered questions.
I have not given specific references to the individual findings. But most of the work has been published in peer-reviewed journals and can be accessed through Pubmed
If you have specific questions or suggestions about this work, please contact me. If you are a lay reader and would like to see this work presented in more easily understood terms, let me know and I’ll see what I can do. I am mostly away from science now, but I find that the neural circuits formed during the period when we attacked the problems described below are still firmly connected and would love to hear from anyone in whom this work elicits excitement and who says ‘Wait, I have an idea!’
Now, read on and welcome to this personal slice of scientific adventure.
Introduction to laboratory research
When I graduated from the University of Rhode Island in May of 1963, I knew nothing about scientific research. This is a remarkable statement since I spent four years as a biology major/math minor. I take most of the blame for this and only little fault do I place at the feet of my teachers or University. What I did learn in those years were social skills: how to drink beer from a fish bowl, how to make and wear a toga for Fraternity parties (think of the scene in Animal House), the excitement of watching Rhody basketball (the football team was terrible and RI winters were cold!), how to navigate the dark in night swims at the beach in Point Judith (where I lived with two classmates) well into the fall and, I learned a little (not much) about girls.
By the end of those four years I was a solid ‘B’ student with no career path in mind. My father was a family physician and at one point I thought this might be my calling. But I wasn’t serious about it, not working hard enough to get the requisite ‘As’ and frankly, I didn’t find the coursework that interesting, nor did the material come easily to me. Most of my science grades were ‘Cs’. Where I raised my GPA was in Math, English and History; ironically, I always felt more comfortable reading that literature than chemistry, physics or biology.
I learned about biomedical research in the summer of 1963 when, as a new college grad, I went looking for a job in New York City and discovered that I qualified for entry-level technician positions. I filled out applications at Sloan Kettering Cancer Research Center and Cornell Medical School. Someone in one of the HR office suggested that as long as I was in the area I should apply for a job at the Rockefeller Institute. (I am sure someone suggested it, rather than going there on my own, because I had no idea what the Rockefeller Institute (or Rockefeller University, as it was soon to become) was. In fact, I didn’t even know it existed!
When I passed through the gates on 68th and York, I thought I had entered the estate of a wealthy Eastside socialite family. The grass was neat and impeccably manicured and the gardens in front of the stone laboratory buildings were right out of the pages of Home and Garden. When the woman in personnel told me that one of their new scientists was looking for a technician and that he was available to interview me, I almost dropped. I had gone on one other interview in the late spring at Yale for a graduate program in mathematics. The head of the program described the coursework and the program at Yale and then said that the field of mathematics led to a rather sedentary life and I had to be sure that was for me. A what? I thought, nodding but not letting on that I had no idea what the word ‘sedentary’ meant. I remember that interview today, 50 years later and still feel embarrassed at my ignorance.
The interview at Rockefeller went somewhat better than the one at Yale, and I got the job. I was hired as a technician for Cecil Yip, a researcher in the actions of thyroid hormones. Since Dr. Yip was also in charge of radiation safety, I was given the job of handling the distribution of radiation detection badges to those students and faculty who wanted to use radioactive isotopes in their work.
Dr Yip’s lab was my first introduction to biomedical research, and how scientists lead their day-to-day lives. Most of my college friends who found their way to NYC had found jobs in banking, Wall Street or finance. The restrictions on their dress and their regimented working hours seemed so oppressive to me that I couldn’t imagine living that kind of life. In my new position, I wore khaki pants and a white pullover shirt (like those worn by Interns of that era). Neither required much laundering and were perfect for the lab work that could be dirty and sloppy at times. As for hours, Dr. Yip’s routine was my glimpse into what it was to be a scientist allowed free reign to tackle a problem and try to solve it without anyone looking over your shoulder.
Dr Yip came to the lab around 10 or 10:30, usually in tennis clothes and perspiring from a squash match. His opponents were other faculty or students who also liked to start the day off in a dynamic way. Dr Yip set up his experiments and then around 1:00, went to the cafeteria for lunch with the Rockefeller faculty, several of whom were Nobel laureates. The faculty included physicists, biochemists, mathematicians, and clinician-scientists – all working to provide answers to problems in biomedical areas. I didn’t know it then, but I had stumbled into one of the premier scientific research institutes in the world.
My life at Rockefeller was closer to 9-5 than was Dr. Yip’s. But still, I knew that I could take liberties. And, if I needed to be somewhere for a good reason (graduate school classes at NYU, for example), I could leave early or arrive late, as long as I got my work done. Cecil Yip usually got his own work done well after I had gone home. He liked to stay late into the night to read pertinent papers or work on a manuscript or sometimes finish an experiment I had begun. It was his routine that initially attracted me to science as a career and I thought that perhaps I too could lead an unregimented, free but responsible, life like his.
The research we were engaged in involved the iodination of thyroid hormone and its precursor. I remember little of the theory of the work but do remember having to prepare a series of chromatography columns and setting up more than a dozen of these small columns to run simultaneously. When I got the experiment going, carefully not allowing any of the columns to run dry, Dr Yip was so impressed that he brought several colleagues around to watch me run the columns. As the months passed, I began to feel competent at this job and proud of my abilities in the lab. I continued to think that a life in science, at a college or university was not a bad one; not having to ‘punch a clock’, not having to get dressed in a suit and tie, not being tied to a desk all day (I spent most of my day on my feet running experiments).
I was making enough money to get a small apartment on 66th and 1st with a roommate so I didn’t have to commute from my parent’s house in Queens. In those years I did not have a real appreciation of what it meant to be a scientist nor had I yet found a passion for scientific inquiry; that didn’t come fully until I had finished my PhD, and postdoctoral training and was running my own lab ten years later.
My parents had strongly urged me to go to graduate school so I could reapply to medical school the following year. I half-heartedly enrolled in the graduate program at NYU along with over a hundred other recent grads (NYU had an open door policy in 1963 – as long as you had the prerequisites, you got in. You then were required to maintain a ‘B’ average to stay in school). Of the hundred or so students I started with in 1963 I only became friends with a few, and none of them completed the PhD degree. There must have been others who did, but I didn’t know them.
Graduate study at NYU
In my first semester at NYU I took Biochemistry, Endocrinology and a seminar course, while working full-time at Rockefeller. My life was occupied with work and attending classes on weekdays and studying on weekday evenings and most weekends. Friends from the old neighborhood in Queens couldn’t understand it. “Weekends, too?” They asked. But, I had to put in long hours if I wanted to get those required ‘Bs’ and understanding biochemical reactions did not come easily to me. Only after seemingly endless repetitive sessions did I get it. The Krebs cycle haunted me for weeks; the glycolytic pathway, memorizing all of the amino acids, the enzymes and nucleic acids involved in protein synthesis were equally daunting tasks. Even then the information seemed endless (the amount of material poured onto today’s graduate students has probably increased 50 fold since the 1960s).
Endocrinology was not as difficult for me as was biochemistry – perhaps because I was spending my days working with many of the same compounds we studied in class, the material seemed to sink in with less effort. Another factor was that the teacher, Albert Gordon, was legendary for his teaching ability and for the care that he took in making sure that all his students learned the material. For the most part, his teaching of endocrinology was a labor of love. His class was at 5:00 on Friday afternoons (all our classes were late in the day or in the evening allowing those of us who worked to make it to class).
One Friday afternoon in late November, I was in Washington Square Park before class when a news-buzz rippled through the crowd. Girls broke into tears, collapsing into their boyfriend’s arms. Long-haired, bearded males and longer haired beaded girls were strewn all over the grass sobbing, many locked in each others arms. John Fitzgerald Kennedy, our dashing, vibrant leader, the hope of America’s youth, the promise of a new way of conducting international affairs, the ideals of our generation come to the highest office, the presidency, the man who many of us felt held the power to change the course of American politics, had been assassinated.
We got to Gordon’s class, but he was not there. When he did arrive, ten minutes late, he came to the podium fighting to control his own emotions. Then he announced what most of us already knew. There were gasps from those who had not yet heard. He said he felt horribly about our president’s death but he considered it his responsibility to go on, to present tonight’s lecture, He understood if anyone felt that they had to leave. Several of my classmates broke down in tears and many left the room. Albert Gordon began his lecture. Most of us were not listening. I think he sensed this and dismissed us after 30 minutes (it was normally a two hour class). We dispersed back into the park.
I remember that no one had taken the American flag down and it still flew over Washington Square. I thought to go out and lower it to ½ staff but didn’t. I have regretted not taking that action to this day; it would have been the right thing to do.
Working towards the PhD degree – the remarkable Vincent P. Dole, MD
I only stayed with Cecil Yip for a year before he left for a faculty position in Toronto and I transitioned to the lab of Dr Vincent Dole. I soon realized that I had gone from the lab of a brilliant young scientist to one of a middle-aged genius. Dole’s genius was proclaimed to me one afternoon by one of his mentors, Donald Van Slyke. Dole, who worked with Van Slyke in the 1940s, had invited him, now an aging gentleman, to visit his lab.
Dole left Van Slyke at my bench so I could tell him about my work. I began to describe my experiments when Van Slyke, who I realized was not really listening but was watching Dole as he went to answer his phone in the next room, interrupted.
‘He’s a genius, you know,’ Van Slyke said to me.
‘Yes, sir,’ I said. ‘I know. I have a lot of respect for him.’
‘No, I mean, he’s a real genius.’ A look of profound deference covered his face.
Vincent Dole had worked on lipid metabolism in the late 1940s and early ‘50s. This research later led to the development of some of the first weight control medications and was part of his accomplishments that resulted in him being awarded the Lasker prize years later. While I was in the lab from 1964 to 1969, his work combined addressing a social issue, the plight of heroin addicts on the streets of New York, with a medical issue, the treatment of those heroin addicts as patients with medical needs, and the scientific issue of where in the brain heroin and similar drugs were acting. My job was to help him try to answer the scientific question.
Over the following months, I began to appreciate what Van Slyke had said. Vincent Dole, like my previous boss at RU, was athletic (he skied, climbed mountains and every day walked the four flights up to our lab in Founders Hall), was urbane, being a main part of the life of the city, working with the NY Health Commissioner, the mayor’s office and other leaders in city government. Dole, along with Marie Nyswander, a psychiatrist (they married shortly after I arrived in the lab), had developed the methadone maintenance program to treat heroin addicts. This treatment program was conceived by Dole who years earlier had reasoned that methadone, a synthetic narcotic – developed in Germany during WWII – that caused analgesia but was much less addictive than heroin might share the same receptor sites in the brain as morphine and thus might compete with it. This meant that methadone might be able to be used in addicts to block the action of heroin without producing some of the unwelcomed side effects (the emotional highs and lows, and the craving for increased doses to attain the high). Since methadone could be given orally, it had the promise to be a treatment to remove the craving in heroin addicts, stabilize their moods and potentially bring addicts back into society as useful, productive citizens. This concept had huge medical and societal implications and was implemented in NYC despite strong political and medical opposition.
While Dole was fighting to introduce the treatment program in a largely unreceptive and skeptical environment, I had the task of determining how methadone worked and where in the brain it exerted its effect. Dole designed the experiments I was to perform and only after I was well into the work did I realize the brilliance of the idea.
Methadone contains an asymmetric carbon atom and thus exists in both levo (l) and dextro (d) forms. Only the l-form is biologically active. Dole reasoned that perhaps we could find the site of action for methadone (and a clue to localizing the site of action for other opioids with similar pharmacological effects) by first chemically synthesizing l and d forms of methadone and then radioactively labeling each with a different isotope, i.e., 3H l-methadone and 14C d-methadone. The idea was to inject a mixture of the two isotopically labeled forms of methadone into the lateral ventricles of the brains of mice. Then at various times after injection, sacrifice the mice and make slices of their brains. We then analyzed the brain slices for the presence of the 3H and 14C labels, expressing the data as a ratio of the isotopes for given regions of the brain. Shifts in isotope labeling favoring the active ‘l’ form would suggest stereospecific binding of the active over the inactive form of the drug.
After working out the details of the experiments and solving most of the major technical problems, I began a series of experiments that would last more than five years.
The basic design was to lightly anaesthetize a mouse, puncture the skull with a small needle, and lower a cannula into the brain according to stereotaxically determined coordinates. The contents of the cannula, the isotopically labeled, d and l forms of methadone, were then released into the ventricle to be distributed throughout the brain. At various times after injection, I removed the brains and submerged them in liquid nitrogen to quick-freeze them and prevent postmortem diffusion. Then I went into the cold room and cut the brain on a microtome. (I think one of the reasons I remember this detail so well is that at the time I was seeing a woman who was an aspiring model. Surprisingly to me, she was intrigued by the work I was doing. When she found out that frequently I went into the lab on Saturdays to cut the brains she begged to come along. One Saturday, she came to the lab and listened intently as I described what I had done previously and that now, I’d be going into the cold room to section the brain. As I put on my coat, she pleaded to come along. I was impressed, thinking that she wanted to learn some brain anatomy. But that wasn’t exactly the case. She put on her fur coat and sat next to me as I dissected the mouse brain, ‘Oh,’ she said. ‘This will be so good for my fur,’ fluffing her coat in the 40 temperature.)
Elaborations of those basic experiments formed the foundation for both my master’s and later my PhD theses. The publication (only one short paper in 1970, 6 years after I had begun) reported a few places in hypothalamic regions where there was a small but statistically significant shift in the isotope ratio we were following. These data showed that a specific part of the brain was showing stereo-specific binding of the active l-methadone in preference to the inactive ‘d’ form, suggesting the presence of stereospecifc opiate receptors that might be mediating the clinical symptoms of analgesia and dependence. But the differences were so small, we concluded that the stereo-specific binding we were searching for might not exist (there could be other reasons why one form of methadone was active and the other was not) or the receptors that bound the compounds existed at levels that were lower than we were capable of detecting. The latter turned out to be the case.
When several different groups discovered the opioid receptors in the late 1970s, they were shown to be present at several orders of magnitude lower concentrations than we were able to detect with our methods. We did get credit for our work, however. In a review of the history of the discovery of the opiate receptors, our paper was described as ‘opening a crack in the wall of ignorance…’ When Dr. Dole pointed this out to me, complementing me on my work, I told him that opening a crack in a wall requires someone who knows where to strike the wall and someone to wield the sledgehammer. I was clearly the one with the sledgehammer.
Throughout my time in the Dole lab, I was influenced profoundly by the intelligence, wit and cultural interests of Enoch Gordis. Enoch was a medically trained scientist in Dole’s lab working in the field of alcohol abuse. He, like Dole, was interested in the clinical and social aspects of alcohol addiction as well as the scientific basis for alcohol dependence. Enoch was a classically trained pianist and opera lover as well as a scientist. He was also one of the funniest, cleverest people I had met and showed me the cultural side to New York that I was just beginning to appreciate.
In 1969 I completed my degree requirements at NYU (my thesis research was done under Dole’s guidance at Rockefeller, but with a co-mentor, Milan Kopac, chair of the biology department at NYU) and went to a post-doctoral position one block uptown, at Cornell Medical School and the laboratory of Bernice Grafstein. I had met Grafstein during her years at Rockefeller University (remember, I took care of the radiation badges). When she heard I was graduating and looking for a post-doc position she invited me to join her as she was leaving Rockefeller to join the physiology department at Cornell. This suited me since I wanted to remain in the New York area and while I wouldn’t be pursuing research in addiction (as I had hoped) I would be learning lots more about the brain and the field that was soon to become known as ‘Neuroscience’. Before 1971, researchers who studied the nervous system were classified as neurophysiologists, neurochemists, neuroanatomists, etc., and were not yet thought of as neuroscientists as they would be in 1971, when the Society for Neuroscience was formed. From then on, if you were studying anything to do with the nervous system from experimental psychology to molecular neurochemistry, you were a neuroscientist.
In the Grafstein lab I was given the task to determine if RNA could be synthesized in a neuronal cell body and then exported into its axon. Grafstein had shown that this was true for proteins and published her findings in landmark papers in the late 1960s. Little did I realize then that this basic question would consume the rest (approximately35 years) of my scientific life. Grafstein’s original work had been conceptually simple but led to important results. The experimental model she had chosen was the goldfish visual system. For a variety of reasons this was an ideal system. For brevity’s sake, I’ll mention two of those reasons. First, the system was easy to use; fish could be anesthetized with ice water and injected with radioactively labeled precursors (amino acids in Grafstein’s experiments) in one eye with little trauma or disturbance to the fish. The labeled proteins could then be detected in the terminals of the retinal ganglion cell axons in the fish brain (the optic tectum) days, weeks or months later. The system allowed the experimenter to control for diffusion and blood borne distribution, distinguishing those mechanisms from the transport of protein molecules along optic nerve axons. Using this system, Grafstein showed that proteins could be synthesized in neuronal cell bodies in the eye and then move along axons to their terminals. ( see:http://www.sfn.org/index.aspx?pagename=HistoryofNeuroscience_autobiographies for a more complete description of Dr Grafstein’s work)
The second reason this is an ideal model system is because these axons – unlike optic nerve axons in higher vertebrates including mammals – are capable of complete regeneration and restoration of function following transection. Thus, the goldfish visual system is an excellent model-system to study axonal transport of molecules and the transport of those molecules during the growth of the axon. In later papers, Grafstein (along with other labs) would show large increases in the axonal transport of total and specific proteins during axonal growth.
I left the Grafstein lab in 1971, just as her work was being praised in the neuroscience community.
My own lab – New Jersey Medical School –
At my new lab at the New Jersey Medical School of the University of Medicine and Dentistry of NJ at Newark, (I was hired as an Assistant Professor in the department of Physiology), I continued to study the axonal transport of RNA in neurons of the goldfish visual system. This work, which I consider the most important of my scientific career, had a complexity that we only began to recognize well after the work began. The main problem laid in the differences in the way the retinal ganglion cells (and all other neurons as far as we know) handle small molecules (amino acids and nucleotides). Following injections of labeled amino acids (aa) into the eye, the retinal ganglion cells take up the radioactive aa from the vitreous where they join the pool of endogenous intracellular aa. These radioactive aa are then incorporated into proteins and a portion are exported into the axon as axonally transported proteins. In none of these experiments was there any evidence for the export of free amino acids into the axons. To my knowledge there have been no experiments reporting the axonal transport of free amino acids into either intact or regenerating (growing) axons in any species.
Surprisingly, this was not the case with nucleotides. Like aa and proteins, radioactively labeled injected nucleotides are taken up by cells of the retina, incorporated into a macromolecule (RNA in this case). But, they are also transported as small molecules into the axons of retinal ganglion cells and then transferred to surrounding cells where they are incorporated into RNA. [The observation that axonally transported nucleotides could be delivered from the axon to cells surrounding it and then be incorporated into RNA led to a series of experiments reported in 1983 in The Journal of Neuroscience. These experiments showed the differential axonal transport and subsequent transfer of small molecules out of the axon into cells surrounding the axon. In this paper we speculated that part of the mechanism used by axons to communicate with neighboring cells is by selectively ‘feeding’ them some small molecules (nucleotides, polyamines, lipid precursors) while amino acids could not be derived from the axon but must come from the supporting cells themselves. To my knowledge, this observation (that I still find intriguing), and speculation remains unchanged; not challenged, confirmed nor advanced in the subsequent decades!]
Returning to the question of the possibility that like proteins, RNA could also be transported along axons, we injected 3H-Uridine (as a precursor to RNA) into the eye of fish, and found labeled RNA in cells of the retina as well as along the length of the axon and at the axon terminals in the goldfish brain. However, we also found labeled precursors of RNA along the length of the axon and at its terminal. Remember, if one does this experiment using radioactive amino acids, radioactive proteins are found only in the axon, but not in cells surrounding the axon; hence, a major difference in the ways a neuron handles protein and RNA precursors.
That nucleotides (as indicated by the radioactive tracer) could be transported along the axon, transferred to surrounding cells and there be incorporated into nucleic acids, created a dilemma not faced by the protein transport people. Since the wave of precursor always preceded the appearance of labeled RNA, it was likely that the 3H-labeled RNA along the axons and at their terminal was the result of the axonal transport of the precursors followed by their transfer to surrounding cells where they were utilized for RNA synthesis. This view was confirmed by autoradiography at both the light and EM levels. Thus, results demonstrated labeled RNA in cells surrounding the axon, but none (or very little) within it. [The inability to demonstrate a significant label associated with RNA in axons fit well with traditional views of axon biochemistry of the 1970s that axons contained no protein synthetic machinery – ribosomes had not been seen using electron microscopy in the axon – and thus could not synthesize their own proteins.] Most of the other biochemical evidence supported the morphological data (with a few notable exceptions) so that while the axon delivery of nucleotide precursors to neighboring cells to be used in RNA synthesis was an interesting finding, the far more intriguing hypothesis that neurons could transport RNA into their axons, was either wrong or occurred at very low levels.
However, dramatic findings were on the horizon; these results catapulted the rest of my scientific inquiry.
The beginning of a passion for science- a low molecular weight RNA is present in regenerating axons
We next posed the question; what happens if nerves are damaged and allowed to begin to regenerate prior to the injection of labeled uridine into the eye? The results were astounding, the kind of data that occur only a few times in a scientist’s life (if we’re lucky).
Results of the first series of experiments showed that the amount of labeled RNA in the tectum was more than ten fold greater in fish with regenerating axons vs controls. Therefore, the fact that the axons were growing led to a huge increase in the export of label from the retina. Electron microscopic autoradiography (performed in the lab of Pierluigi Gambetti) using stringent quantitative data analysis, showed that at least ½ of the labeled RNA was in the growing axons, with a significant amount localized to the axon growth cone.
[The story behind the collaboration between my lab and that of Pierluigi Gambetti began in 1973 and is worth telling. I was new to the field, having just opened my own lab and, no longer associated with Grafstein. But armed with what I considered strong data generated over the past two years from biochemistry and light autoradiography, I submitted an abstract to the meeting of the Society for Neuroscience in San Diego. I was assigned to give a 10 minute talk followed by 5 minutes of questioning (most of the presentations followed this format in what was only the third annual meeting of the Society).
Nervous but confident in the data, I gave my first presentation as an independent investigator. I reported that using the goldfish paradigm – injecting labeled uridine into the eye of fish with regenerating nerves – resulted in a large increase of labeled RNA in the brain. The light autoradiographic evidence showed linear arrangements of labeled RNA in the brain, “suggesting,” I emphasized, that the RNA was in axons and that it had been transported from the eye along the growing axons.
Unfortunately, my talk was preceded by Pierluigi who showed elegant EM autoradiograms, using the same experimental paradigm in the rabbit visual system that revealed no RNA in axons. Pierluigi concluded that the labeled RNA seen along the length of the axon must be in surrounding glia and not in the axon. When I finished my presentation I was peppered with questions, like, ‘Dr Gambetti has just shown in brilliant experiments using sophisticated techniques that the RNA is not in rabbit optic axons, but in the surrounding cells. Now you stand up here, showing inconclusive light autoradiograms and have the audacity to propose that the RNA is in axons.’ Although I pointed out two things, first that we saw the same thing as Pierluigi in intact optic nerves of fish and that these are regenerating axons, and second that our data was ‘suggestive’, clearly not definitive, my 5 minutes of audience questioning was up and I was left standing in front of a packed room, like the rookie who just missed two foul shots losing the game for the home team. Pierluigi was lauded. I was vilified. [It was around this time that Grafstein said to me, Nick, ‘you don’t have to put my name on any of your RNA papers, nor do you have to thank me on them. I prefer that you not associate my name with any of that work.’]
Ironically, it was Pierluigi who came to my rescue. He cornered me outside the presentation room at the break and asked if we could talk. I expected to be ridiculed a little more. Instead he said, in his charming Italian accent, ‘Nick, I have to tell you, we have seen the same as you in baby rabbits during optic nerve development. There appears to be RNA in those growing axons too.’
‘What? Why didn’t you say something in the room?’ I said. ‘Why did you let them attack me like that?’
‘I’m not sure yet,’ said Pierluigi. ‘The data are too preliminary and it’s too difficult working in baby rabbits. How would you like to collaborate? You supply the specimens and we do the EMAR on the regenerating optic axons of goldfish.’
Five years later, after ‘disaster followed by ‘disaster’ [I would call Pierluigi periodically to ask how things were going with his part of the experiment and he would greet me with that single word], we published unequivocal data based on extensive and well-controlled experiments. We were confident in reporting that during regeneration of the goldfish optic nerve (but not in intact nerves), large mounts of RNA could be synthesized in retinal ganglion cell bodies in the eye and transported axonally in growing optic axons. We also reported, that much of this RNA was in the furthest extent of the growing axon, the axonal growth come. Although both our labs were elated with the outcome, the scientific community barely noticed. This was my introduction to what is axiomatic in science; that a new, radical, anti-dogmatic hypothesis is met with scorn and disbelief at first. Then, once it is proven, the response is ‘Oh, that’s nice – so what’s it doing there?’
Of course, that is the question we too wanted to answer. Why was RNA in the most advanced tips of growing axons? This question, still unanswered, occupied much of the rest of my scientific life. I remember thinking then that I was sure of this RNA data and was convinced of its importance. And so, I decided that this would be the focus of all of my future experiments.
Trying to ascribe function to observation: What role does RNA play in the regeneration of goldfish optic nerve axons?
What follows is a description of the approaches we took to try to answer this question. But, I’ll state at the outset, more than 40 years later, there is still no answer to this question.
The next step was to determine the species of RNA present in these growing axons. We tackled this problem using SDS PAGE. This was not so simple (nothing was in these experiments). We had to find a way to distinguish the radioactive RNA inside the axon from that in the surrounding cells. For this part of the story I refer the reader to a series of papers published in the late ‘70s and early eighties. (see especially Science, 1979). The unavoidable conclusion (the culmination of about 5 years of work) was that only a low molecular weight RNA of approx. 76nt, was being transported from the eye into growing axons. This conclusion was consistent with results published by Ray Lasek in 1973 using the squid giant axon, where, there too, the preponderant species of RNA isolated from extruded axoplasm, was localized to a single band in the molecular range of a 4S, tRNA, marker (approx. 76 nt).
In subsequent years, several graduate students extended these findings in goldfish to peripheral nerves of mammals (Tom Lindquist, rat sciatic nerves), and a central mammalian nerve, optic nerves of rats during development (Mike Politis), with very much the same results. Intact vertebrate nerves contained little (if any) axonally transported 76nt RNA. But when the nerve was growing, either during development or regeneration, relatively high amounts of a ~ 76nt RNA were present in the axons.
Research at the Marine Biology labs at Woods Hole
Some of the research done in these years (late 1970s early 1980s) was performed on the giant axon of the squid obtained from the waters off the coast of Martha’s Vineyard and brought to the Marine Biology Labs (MBL) in Woods Hole, Ma.
[The MBL is an idyllic research facility that simultaneously was: a) a scientific inspiration (constant seminars, top quality scientists available for discussion, many Nobel Prize winners, a dedicated support staff that could get you almost anything you needed within 24hrs), and b) a huge distraction from being in the lab (Woods Hole is on a peninsula surrounded by Vineyard Sound leading out to the Atlantic ocean to the east and south and Buzzards Bay to the west and north. The coastline is rocky with sandy beaches scattered in spaces between the rocks and the woods (of Woods Hole) which reach down almost to the waters edge (fresh water springs are abundant in the area supporting the rich vegetation).
Nature’s beauty tended to lure me outside whenever possible. The sunsets over Buzzards Bay viewed from a small beach we called ‘Sunset Beach,’ were some of the most spectacular I’ve seen anywhere in the world. I frequently missed the evening lectures (Fridays and Mondays) in favor of these sunsets – perhaps you can see why, all of the photographs on this website come from the area around the MBL at Woods Hole.
In the first summer at the MBL I worked alone, collecting axoplasm from the giant axon of squid. In subsequent summers, I took my colleague, Goutam Chakraborty, (a post-doc and then Research Associate who played a major role in the lab in the late 1970s and through the 1980s and is responsible for much of the data described in those years) and medical students from NJMS with me to help with the experiments.
The first medical student to come to Woods Hole with me was Armen Babigian. In the second summer, Laurie Slotnick, also an NJMS student, replaced Armen as my lab assistant. In the beginning of the second summer, Armen spent some time with us up in Woods Hole. He also showed Laurie Sunset Beach and the rock ¼ mile offshore that we called Paradise rock. Several years later, these two graduated med school, got married and went into separate practices (neurology and dermatology) in western Connecticut. What should we call this ‘Love and Loligo’ Passion and the Giant Axon? I still grin at the memory of the two of them getting to know each other over squid giant axons.
In our initial effort to ascribe function to the ~76 nt axonal RNA we posited that the axonal RNA was transfer RNA. This was logical since the size was the same and no other RNA of that molecular weight had yet been described in the literature (but see developments later in this essay). Since little if any protein synthesis was thought at the time to go on in axons (we now know that view is too limited; the current view holds that some proteins are made within axons) we hypothesized that tRNA in axons, specifically argtRNA acted posttranslationally to add Arg to the N-terminal amino acid of endogenous axonal proteins. Posttranslational arginylation of proteins had been described previously, was ubiquitous in eukaryotic cells, and it seemed possible that this was its function in axons.
We tried to prove this hypothesis in axoplasm of the giant axon and exhaustively looked in vertebrate nerves for answers to these questions (Mike Zanakis, a PhD student was responsible for much of the work and planning of these experiments). We did demonstrate the reactions in axoplasm of squid and Mike was able to show, using peripheral nerves of rats, that in vivo injury to a nerve stimulated the arginylation of proteins (measured in nerve extracts). Later he showed that growth of the regenerating nerve also elicited a spike in the arginylation of proteins. Around the same time, arginylation was being proposed as a mechanism used by cells to target damaged proteins for degradation by the ubiquitin proteolytic pathway. This set up an intriguing possibility that arginylation in axons was a prelude to ubiquitination of that protein which made it a target for degradation. However, we were never able to demonstrate the sequence of damaged proteins arginylation ubiquitination degradation, in any of the systems we used nor, were we able to determine the protein targets for arginylation. A final frustration was that we were not able to show in vertebrate nerves that the reactions were occurring within the axons and not in surrounding cells.
During this frustrating period (mid 1980s) our thinking shifted to the clinical problem of spinal cord injury in humans. Could any of the work we had done so far provide clues as to why some axons – central axons of fish and frogs, for example and peripheral nerves of higher vertebrates, including humans – can regrow severed axons while axons of higher vertebrates like those in the spinal cord and optic nerve of humans could not? Susan Shyne-Athwal joined the lab as a PhD student and set about testing the hypothesis that the posttranslational arginylation of proteins (whatever its function) was a critical early event in successful regeneration. The hypothesis stated that peripheral nerves of mammals (tested in experimental rats) were capable of initiating these reactions while central nerves were not.
Sue set out to test this hypothesis by crushing optic and sciatic nerves in rats, then assaying the section of nerve just proximal to the crush site for posttranslational arginylation. Initial results were thrilling. Sue was progressing along a time course and reporting the results as she completed the experiments: sciatic nerve level of arginylation in controls = 1; two hours following nerve crush = 1.4; 4 hrs = 2.0; 6 hrs = 5.0; 12 hrs = 10.0; 24 hrs = 10.0. Optic nerve, controls, arginylation is 0.5 (1/2 control values of sciatic nerves); at all other time points after optic nerve injury segments proximal to the site of injury showed values that were = or less than controls. Eureka!! The sciatic nerve, a nerve capable of regeneration is able to activate arginylation reactions while an optic nerve is not. Perhaps this means that if we could stimulate arginylation in optic nerves we might be able to stimulate those axons to grow.? Certainly a testable hypothesis. But Sue did some more experiments that made us think again (remember, nothing is simple in this research area).
When Sue looked at time points, over several days, sciatic nerve arginylation showed a second wave of increases, this time the increases occurred in the distal nerve stump where newly formed axons were growing. That was interesting, but the optic nerve was the real surprise. We expected the reactions to continue to be low and show little change (remember these axons were dying, not regrowing). But that’s not what Sue found. Beginning a few days after nerve injury the reactions in optic nerve proximal stumps (still connected to their cell bodies), began to show increases in arginylation. They were being activated by some physiological changes occurring in that injured proximal nerve stump but not until days later, whereas the same reactions were enhanced in sciatic nerves within hours of injury. We knew this was important but the possible significance of this didn’t occur to me till years later.
Axon resealing following transection
In the late 1990s I stumbled on a literature that I had not been aware of (too much time enjoying sunsets, perhaps). The experiments dealt with the resealing of cut axons following axotomy. The literature was small but elegantly simple and provocative. The experiments showed that severed axons of neurons incubated in medium reseal according to a specific time course and require a number of components without which they will not reseal. We began to reason; what if there was a resealing problem in central vs peripheral nerve axons, that peripheral nerves readily resealed their axons but central axons lacked some component that would delay resealing.
Thus, in our arginylation experiments, we reasoned, peripheral axons reseal rapidly creating an internal axonal milieu suitable for arginylation and the activation of this reaction works, in some way, to trigger axon growth leading to successful regeneration. Conversely, central axons lack a resealing component and so the closing of the severed axon is delayed, leading to the lack of activation of arginylation until a much later time, when conditions for successful regeneration have passed.
This again was a testable hypothesis and after lengthy conversations with my longtime colleague and friend, Sansar Sharma, he agreed to put one of his post-docs on the project. The idea was to cut optic and sciatic nerves in anaestherized rats, then apply dye to the cut ends of the axons at various time after axotomy and record the uptake of the dye into the cut axons; the logic being that resealing time could be determined by assaying for the exclusion of dye from the axon (the failure to take up the dye would indicate that the cut end of the axon had sealed). These experiments (results published in 2001) showed that central axons had a significantly longer resealing time than did axons in a peripheral nerve environment. This tantalizing data remains just that – we were not able to pursue this line of research for a variety of reasons (mainly rejection of several grant proposals) and unwillingness on my part to champion another far-out venture and confront skeptics for whom I had developed dwindling respect as scientists. In my view, this remains another area of research worth pursuing, especially in light of the many failures of other approaches that have tried to solve the clinical questions of central nerve regeneration in humans.
Returning to the question of the role of a 76 nt RNA in axons.
By the early 1990s we had completed several series of experiments with two other PhD students (Mujun Yu, a remarkable young graduate student from China, who is now a pathologist living in Seattle, Wa., and later Doug Steinberg, an equally remarkable guy (who already had a law degree, and degrees in journalism and neuroscience when he joined my lab) which (through no fault of theirs) made little headway in our quest to discover a function for the 76 nt RNA in regenerating axons.
Out of ideas, bereft of thoughts for significant experiments, unwilling to ‘sacrifice’ any more animals for this cause, and being hinted at strongly (very strongly) by funding agencies that perhaps it was time to throw in this towel, I threw in the towel.
That same year (1999) I assumed a leadership role at the graduate school at UMDNJ on the Newark campus and found a second passion and renewed inspiration (see the Education and Social Issues links.)
Ironically, less that two years later I stumbled on another possibility for what a 76nt RNA could be doing in a growing axon.
RNA interference as a possible function for axonal RNAs
In 2002, my father-in-law was hospitalized for chest pain and subsequently underwent several surgical procedures in Albany, NY. My wife and I visited him but I had to return to New Jersey early and so booked a seat on Amtrak back to Newark. On the taxi ride to the train station I started talking with the cab driver, a ZZ-Top kind of guy in his early fifties, I guessed, who told me that he didn’t always drive a cab.
“What did you do before this?” I asked.
“I used to be a cell biologist,” he answered.
I told him about my research in neuroscience and we both commented on the remarkable coincidence.
When he told me about his PhD work and his mentor, I realized that he had been with one of the country’s premier Drosophila genetics labs; his mentor had won the Nobel Prize.
Then he asked me what was new in cell biology because, he said, while lots was published every year, there wasn’t all that much really new. I told him I agreed. [When I later told this story to a bunch of colleagues, they too all nodded in agreement].
I told him about a field I had just heard and read about, the field of RNA interference. In these groundbreaking experiments (three of the original researchers were awarded the Nobel Prize in 2006), experiments had shown that a 22nt RNA could silence mRNAs and thus control translation well away from the nucleus. I told him about our experiments and that when I started reading about these miRNAs (one form of inhibition comes from a class of RNAs termed micro RNAs), I began to wonder if this is what we had been looking at all along – an RNA in axons whose job it was too regulate translation by suppressing the expression of mRNA (this fit well with the emerging evidence that some protein synthesis did occur in axons and with our ideas about the mechanisms associated with axonal growth). But, I said. I don’t think I could confuse a 76nt band with a 22 nt band. I didn’t consider my lab a great biochemistry lab (or myself a great biochemist), but I couldn’t imagine we could have been that far off.
“What are the precursors of these 22nt miRNAs?” asked my cabdriver. I had no idea and really hadn’t thought about it. But he had planted a seed. When I got back to my office I began looking for papers about the origins of the miRNAs and found none. Then later in the summer of 2002 came a report that all of the 22nt miRNAs were derived from a higher molecular weight precursor and in all cases studied, that RNA had a length of 76nt! What? Sure enough, when I looked at the gels in those publications, the researchers had used E. coli tRNA as a marker and the putative precursor of the miRNA migrated at exactly the same spot on the gel. The pre-miRNA (as it was called) turned out to have a hairpin structure, contained the unique miRNA sequence in one of its stems and, in all cases was 76nt in length. This was it. I felt sure about it.
We had not been looking at tRNA at all; it had been 76nt pre-miRNA and functioned by releasing specific miRNAs which regulated local translation in the axon. This was another tantalizing hypothesis, but like the ones before it, remained to be proven.
One evening during this period my wife noted that I looked particularly happy. “Had something happened at work?” She asked.
“I think I’ve been testing the wrong hypothesis for the last15 years,” I said.
“And, you’re happy about that?” she said.
But I was. Much of science opens doors and raises questions that never get answered in our lifetime. Now there was a possibility that we could identify the purpose of a 76nt RNA in growing axons.
Around 2003 I decided to get back into the battle but only in a reduced, ancillary way. The experiments needed to test the hypothesis that the 76 nt RNA in axons was the pre-miRNA could best be done, I reasoned, in the squid giant axon and the analysis of that RNA could best be done, not by me, but by an RNA molecular biologist.
The best person I knew to isolate axoplasm from the squid was Harish Pant (of NIH). As an RNAi molecular biologist I approached Tom Tuschl of Rockefeller University to see if he was interested in assaying the RNA from axoplasm for the presence of pre-miRNA . Both agreed and I became the facilitator, intermediary of the group. These experiments went on for several summers: Harish collected the axoplasm and the following falls the Tuschl lab analyzed the data). The Tuschl lab was able to show that axoplasm contained miRNAs and several distinct species of miRNA were identified. However, for a number of reasons, his lab was not able to answer the original question of whether the 76nt RNA that we and Pierluigi Gambetti had found in the growing tips of regenerating axons was the precursor of miRNA, pre-miRNA.
Thus, the question: is the 76nt RNA in axons of squid and regenerating optic axons of goldfish, pre-miRNA? remains unanswered.
This is disturbing. But, in a way (perhaps I have rationalized), is not surprising. In my scientific life, several different exciting experimental pathways have not led to anything even remotely approaching a complete answer. We don’t know why 76 nt RNA is in axons and what it has to do with regrowth of axons (although I still suspect it is pre-miRNA). We don’t know why, neurons move select small molecules out of their cell bodies and transfer them to surrounding cells. We don’t know what the significance is of the failure of central axons to reseal following injury and what, if any, answers to these questions might lead to clinical application. It is axiomatic in science that every successful experiment that answers a scientific question raises a myriad of new questions. These are the questions that I hope will be taken up by some of the scientists of the next decades.