Fifty years my mentor: Harold Atwood

Abstract

While readers of Journal of Neurogenetics may be familiar with Harold Atwood’s work with Drosophila, most may know little of his previous work on crustacean neuromuscular systems that prepared him to utilise Drosophila neuromuscular junctions. Here, I will give brief overviews of his academic career, one line of his research that persisted throughout his career and his entry to the Drosophila field. This is not a review paper. Finally, I will relate my experiences with Atwood since 1967 as an undergraduate, Postdoctoral Fellow, and Faculty member and finish with some personal anecdotal observations.

Keywords:

• Synapse
• transmission
• presynaptic differentiation
• contraction
• crustacea

Academic history

Harold Atwood graduated with a B.A. in Biology from University of Toronto in 1959. He proceeded to University of California, Berkeley, where he graduated with the M.A. in 1960 having completed a thesis titled ‘Effects of Temperature and Thermal Acclimation on Neuromuscular Performance in Crabs’ with Professor R. I. Smith. Next he completed a Ph.D. in 1963 working with Professor C. M. Yonge, F.R.S., and Dr. G. Hoyle at University of Glasgow on ‘Investigations on Excitation and Contraction in Crustacean Muscles’. Thus, Atwood’s formal education was completed in three very different educational systems. He went on to do postdoctoral work at University of Oregon with G. Hoyle and at Caltech with C. A. G. Wiersma. In 1979, he was awarded the D.Sc. at University of Glasgow with the thesis title: ‘Crustacean Neuromuscular Systems’. In 2010, he was awarded an honorary D.Sc. from University of Waterloo.

In 1965, he joined the Zoology Department at University of Toronto and remained until 1981 when he became Chair of the Physiology Department in the Faculty of Medicine at University of Toronto. He served in that capacity for two terms of five years. He retired at the mandatory age of 65 years in 2002 but kept his laboratory running for several more years.

Awards

Atwood is a Fellow of the AAAS and a Fellow of the Royal Society of Canada. He was president of the Canadian Association for Neuroscience. He was awarded the Fry Medal of the Canadian Society of Zoologists for outstanding contributions to zoological science in Canada and the Medical Research Council of Canada Distinguished Scientist Award. He was Canadian member of the Council Scientists of the Human Frontiers Science Program and was on the Council of the International Union of Physiological Scientists (IUPS).

Professional activities

Atwood has served an Editor role on nine journals and is a member of many professional societies. He has served on grant review committees of the National Research Council of Canada, Canadian Heart Association, Muscular dystrophy Association of Canada, Medical research Council, Human Frontier Science Programme Organization and others.

Research

Crustacean neuromuscular systems

In the early 1960s, the nervous system was even more of a mysterious black box than it is now. There was great utility in the pursuit of fundamental questions by means of experiments conducted on simple systems. This notion is just as relevant today as it was then: ‘It is far better to achieve a rigorous solution to a simplified problem than to remain frustrated by confusion. – A first approximation may permit a sound and rigorous attack on the more salient features of the problem.’ (Davidson, Lindsey, & Davis, 1987; and see Conn, 2017). In this era, Comparative Physiology was in its heyday and there were several approaches used. Harold Atwood focussed on the discovery of general principles that resulted in unifying concepts. He chose to use crustacean neuromuscular systems where, unlike vertebrate skeletal neuromuscular systems, there is complex distributed innervation sometimes with multiple excitatory and inhibitory inputs on single muscle fibers. Once one appreciates a muscle cell as a general postsynaptic cell, this arrangement appears to mimic to some extent the complexity of innervation in the mammalian central nervous system (CNS) but is much easier to study. For instance, the muscle cells are large and permit easy penetration by sharp microelectrodes and long experiments can be performed at room temperature without oxygen supplementation. Moreover, entire muscles have few motor neurons and they can be individually identified. For example, the claw opener muscle of crayfish has just one excitor and one inhibitor motor neuron. The same identified neuron can be studied in successive animals and they are large enough to permit penetration by sharp intracellular microelectrodes to allow measurement of membrane potential, stimulation and intracellular injection. Atwood became the major contributor in this field. Indeed, he showed that crayfish and their classmates were good for more than bisque or bait.

Some readers may be interested in the technology used in Atwood’s early years. For instance, muscle resting potential and synaptic responses were measured with intracellular sharp microelectrodes filled with 3 M KCl. Often experimenters built their own amplifiers using vacuum tubes. Atwood, Hoyle, and Smyth (1965) show a diagram of a setup using a vacuum tube circuit called a cathode follower connected to a microelectrode. This circuit (Figure 1) provided the high input impedance required to match that of the microelectrode and was built in the lab. An RCA 5734 moveable anode vacuum tube was used to detect muscle contractions. Applied force caused the anode inside the vacuum tube to move relative to the cathode thus changing the transconductance. Electrophysiological signals were displayed on a cathode-ray tube oscilloscope and recorded with a Polaroid camera or Grass 35 mm motorized camera on paper or film. When funds were low or there was little choice of commercial scientific instruments, it was a great advantage to understand electronics enough to design and build instruments in-house (Atwood & Parnas, 1966; Atwood & Walcott, 1965).

Figure 1. Preparation of single muscle fibers for recording tension. (a) The force transducer is an RCA 5734 vacuum tube to which fine forceps tips (b) have been attached to grip a muscle fiber after (c) one end had been cut from the shell. Microelectrodes inserted in the fiber (a, d) passed stimulating current and recorded changes in membrane potential via a cathode follower vacuum tube circuit. Motor axons were also stimulated (d) (from: Atwood et al., 1965). Inset: The RCA 5734 vacuum tube mechanical transducer was 8 mm in diameter and 24 mm long (picture from https://www.radiomuseum.org/tubes/tube_5734.html).

Contraction speed, innervation and synaptic differentiation

Some crustacean muscles are innervated by two axons but stimulation of one causes rapid contractions while stimulation of the other causes slower contractions. Early theories to explain these observations suggested that each axon released a different transmitter. However, Atwood and collaborators showed that not all the muscle fibers were the same; those that produced rapid contractions had short sarcomeres (2–3 µm) while the slower contracting tonic muscle fibers had much longer sarcomeres (11 µm). As a generality, the phasic short sarcomere, fast contracting muscle fibers are innervated by synapses that release many quanta and produce large excitatory postsynaptic potentials (EPSPs) that may depress with repetitive stimulation. The longer sarcomere, slow contracting muscle fibers are innervated by synapses that release fewer quanta producing EPSPs that facilitate (increase) with repetitive stimulation. The next time you tuck in to a dinner of Homarus americanus (found on the north Atlantic coast of America) examine the two claws (chelipeds). One, called the cutter, is slim, has sharp protrusions, fast contracting muscle and is specialized for catching prey. The other claw is fatter and knobby, contains many tonic muscle fibers specialized for prolonged slow contraction; this is called the crusher and functions to break up prey.

Atwood and colleagues produced many studies on the basis of neuromuscular differentiation using many techniques. The results are summarized in reviews (Atwood, 1967, 1976). An outstanding feature of Atwood’s research is the combination of electrophysiology and ultrastructural studies. This provided physical correlations with electrical measurements. This effort was facilitated by the beautiful ultrastructural work of Ms. Irene Kwan, Dr. S. S. Jahromi and Dr. Leo Marin.

Further studies showed that although the transmitter release probability is 100–1000-fold higher in phasic synapses than in tonic synapses, they have smaller pools of readily releasable vesicles than tonic synapses (Millar, Bradacs, Charlton, & Atwood, 2002). Phasic synapses also have fewer mitochondria. Imaging experiments showed that differences in Ca²⁺ entry per active zone cannot explain the difference in transmitter release between phasic and tonic synapses (Msghina, Millar, Charlton, Govind, & Atwood, 1999). The relative properties of phasic and tonic synapses are reviewed in Millar and Atwood (2004). The major hypotheses for presynaptic differentiation were set out in detail in a review (Atwood & Karunanithi, 2002). A hypothesis emerged that phasic synapses are more sensitive to Ca²⁺ than tonic synapses. Andrew Millar, a graduate student in Atwood’s lab, performed a series of experiments in which spatially uniform steps of intracellular [Ca²⁺] were imposed by photolysis of a Ca²⁺ chelator injected into phasic or tonic axons. When intracellular [Ca²⁺] and transmitter release were measured, it was clear that for similar Ca²⁺ loads, phasic synapses released the readily releasable pool of vesicles at a higher rate and shorter delay than tonic synapses. Robert S. Zucker (University of California, Berkeley) provided essential expertise in the Ca²⁺ release experiments and extensively modelled possible explanations of the results. The two types of synapse may differ in Ca²⁺-dependent priming of vesicles (Millar, Zucker, Ellis, Charlton, & Atwood, 2005). This leaves us with the question of which molecular interactions in the transmitter release process differ between phasic and tonic synapses. Atwood’s work inspired us to examine the possibility that SNARE zippering differed in the two types of synapses. We approached this by presynaptic injection of catalytic light chains of tetanus toxin and botulinum-B and -D which cleave VAMP but not in the fully zippered configuration of the SNARE complex. Only botulinum-B blocked transmitter release under low stimulation conditions. The resistance to cleavage by tetanus toxin and botulinum-D indicates that the SNARE complex is partially zippered in both synapses. The physiological effects of the proteases were similar in both types of synapse so we concluded that SNARE zippering was not much different and therefore the huge difference in release probability does not arise from differences in the degree of SNARE zippering (Hua & Charlton, 1999; Prashad & Charlton, 2014). The molecular basis of phasic–tonic differentiation remains obscure.

This research summary has focussed on a narrow range of Atwood’s work and is not intended to be complete. Other aspects are discussed in other perspectives in this issue. A broader summary of Atwood’s work is provided by Satterlie and Cooper (2004).

Drosophila era

The many experimental advantages of Crustacea for the study of neurobiology have been exploited to the full by Atwood. However, some aspects of crustacean biology such as their long generation time, high chromosome number and lack of genome data have limited their usefulness in studies that require molecular genetics analysis. For instance, crayfish reach sexual maturity in 3–7 months while the time to maturity of marine Crustacea such as the lobster, Homarus americanus, is even longer. The time to sexual maturity is longer than the usual models such as Caenorhabditis elegans, Drosophila, mouse, rat but similar to zebrafish (Vogt, 2018). Decapod Crustacea have a large number of chromosomes; the crayfish Procambarus clarkii has 2n = 188 while the lobster Homarus americanus has 2n = 136 (see Table 1 in Mlinarec, Porupski, Maguire, & Klobučar, 2016). The large number of chromosomes may make reliable karyotyping and gene locus determination difficult although other crustaceans have fewer chromosomes. As of 2018, only five crustacean full genomes have been sequenced including that of the parthenogenic marbled crayfish, Procambarus virginalis (Baldwin-Brown, Weeks, & Long, 2018; Gutekunst et al., 2018). This species is likely to be an important research model because it produces offspring (>50) that are clones (Vogt, 2018), can be easily cultured and its genome is sequenced. The genome of P. virginalis is about the same size as humans. Low coverage genomes for several other crustaceans are available (see review in Vogt, 2018) and more complete genomes are appearing (Baldwin-Brown et al., 2018). Despite the lack of genomic data, the sequences of several proteins in the crustacean nervous system have been deduced from cDNA (e.g., see Prashad & Charlton, 2014). Gene editing by CRISPR/Cas (KKumagai, Nakanishi, Matsuura, Kato, & Watanabe, 2017; Nakanishi, Kato, Matsuura, & Watanabe, 2014) and other techniques (Hiruta et al., 2014) have been applied in Daphnia whose genome has been sequenced. As other crustacean genomes are sequenced and thoroughly annotated, powerful techniques will be applied and crustacean biology will take a huge leap forward. I believe we can look forward to a new era of crustacean mutants and altered genes that will allow testing of profound molecular hypotheses. Atwood’s work provides the framework with which to exploit these new developments.

These advances in crustacean biology were not available in 1991. Therefore, it was a wise decision to expand research to Drosophila, a model which provided many mutants of interest to synaptologists. His excellence in experimental design and technique and vast experience with Arthropod neuromuscular systems put Harold Atwood in a perfect position to exploit the relative experimental advantages of Drosophila. In 1991, he went to the lab of C.-F. Wu to learn how to do experiments on Drosophila larval neuromuscular junctions.

Improved saline

Back in Toronto, Atwood remarked that Drosophila larval neuromuscular preparations developed vacuoles and deteriorated rapidly, and that the corresponding brevity of experiments severely limited the amount of data that could be collected and the kind of questions that could be investigated. As in general society, an immigrant in science can make major contributions when unencumbered by prevailing cultural assumptions. He saw no obvious reason why Drosophila larval preparations should be so different from those of crayfish neuromuscular preparations and surmised that the standard physiological saline was to blame. A fine graduate student, Bryan Stewart (Professor, VP Research, University of Toronto Mississauga), evaluated the literature about Drosophila hemolymph and analyzed hemolymph with ion-selective electrodes. This technique was difficult and required a great deal of perseverance. I encouraged its use because my previous experience with intracellular K⁺-selective electrodes (Charlton, Silverman, & Atwood, 1981) convinced me that this was an excellent way to determine ion activity (free ion concentration) as distinct from total ion concentration, in very small volumes. Ion activity is more relevant in the design of physiological salines than total concentration. The results showed that existing saline recipes were not very similar to hemolymph and suggested that a new formulation might work better. The new formulations called hemolymph-like salines (HL) preserved physiological parameters and worked much better in electrophysiological experiments (Stewart, Atwood, Renger, Wang, & Wu, 1994) than previous salines. This development has had a major influence on Drosophila neurophysiology and the paper is Atwood’s most cited publication with about 500 citations (Web of Science). Further tweaking of HL3 saline has made it more applicable in some situations (e.g. Feng, Ueda, & Wu, 2004).

Presynaptic differentiation

As with the crustacean projects, the approaches were electrophysiological and correlated structural studies. The study of presynaptic differentiation was extended to Drosophila. Atwood, Govind, and Wu (1993) described the comparative ultrastructure of type 1a and 1b synapses which have different transmitter release probabilities. Kurdyak, Atwood, Stewart, and Wu (1994) investigated differential anatomy and physiology of type 1a and 1b synapses.

Calcium imaging

Calcium-dependent modulation of transmitter release is a central theme in many studies of synaptic transmission. We wondered whether Ca²⁺ dynamics could be measured in nerve terminals of Drosophila neuromuscular junctions. Previously my lab had measured Ca²⁺ signals in presynaptic terminals of the squid giant fiber synapse, frog neuromuscular junction (Robitaille & Charlton, 1992), and crayfish neuromuscular junction (Elrick & Charlton, 1999). We decided to apply some of these techniques to Drosophila neuromuscular junctions. First, with an Atwood postdoc, Shanker Karunanithi (now at Queensland Brain Institute) and John Georgiou (Charlton lab, now at Lunenfeld-Tanenbaum Research Institute, Toronto) we used permeant Ca²⁺ indicators to load presynaptic boutons (Karunanithi, Georgiou, Charlton, & Atwood, 1997). Later in experiments conducted by Greg Macleod (joint postdoc, now Associate Professor, Biological Sciences, Florida Atlantic University) we forward-filled cut motor axons with indicator conjugated to dextran (Macleod, Hegström-Wojtowicz, Charlton, & Atwood, 2002). The work described fast Ca²⁺ clearance and resting intracellular [Ca²⁺]. Then, we showed that Ca²⁺ signals could be detected in single CNS neurons with backfilled indicators (Macleod, Suster, Charlton, & Atwood, 2003). These were very early shots in the study of Drosophila Ca²⁺ dynamics. The techniques were also used to ask whether synaptic vesicles buffer intracellular Ca²⁺. To do this we compared Ca²⁺ signals before and after depleting vesicles in the dynamin mutant shibire (Macleod, Marin, Charlton, & Atwood, 2004). A former Atwood Postdoctoral Fellow, Greg Lnenicka (Professor, Biological Science, SUNY, Albany), took the study of Ca²⁺ dynamics much further in his own lab. Later studies used genetically encoded Ca²⁺ indicators (Chouhan, Zhang, Zinsmaier, & Macleod, 2010; Reiff et al., 2005), and expanded the investigation to octopaminergic type II synapses (Martin et al., 2016; Xing & Wu, 2018).

The similarities between crustacean and Drosophila neuromuscular systems have been reviewed by Atwood and Cooper (1995). There is considerable sequence homology between crustacean and Drosophila proteins. For instance, crayfish (GenBank accession number KF773142, GI: 575488803) and Drosophila VAMP are 86% identical and share the critical SNARE motif region including binding and cleavage sites for the catalytic light chains of tetanus toxin, botulinum-B and botulinum-D (Prashad & Charlton, 2014). Several antibodies made for Drosophila proteins react in crayfish. Thus, results from studies in Drosophila in addition to being important contributions, will fertilize crustacean studies.

Not just Arthropods!

Although his major research was on Arthropod synapses, Atwood published 22 papers using mice or rats. These included studies on basic muscle physiology, nerve terminal sprouting, membrane properties, denervation effects, ion regulation and synaptic plasticity in the hippocampus. He also published some papers on the giant fiber synapse of squid with me. Clearly, Atwood is no chauvinist in his choice of research preparations.

The big picture

Highly successful scientists are usually pretty good at putting their work in context with the big concepts in their field. Even in early papers, Harold Atwood pointed out how crustacean neuromuscular junctions were useful models for events in the vertebrate CNS. This was a frequent theme in the introductions and discussions of his papers. He often draws parallels between his data and conclusions and those from mammalian neuromuscular junctions and CNS. The review published in 1967 (Atwood, 1967) was highly cited and was designated as a Citation Classic in the publication Current Contents (Atwood, 1981). Atwood noted in this brief missal that the review paper was highly cited because it demonstrated ‘features shared by neuromuscular systems of vertebrates and arthropods’ and it ‘established some general features of crustacean neuromuscular systems, emphasized the utility of these systems for research on synaptic mechanisms akin to those in central nervous systems, and provided a timely summary of a rapidly developing field of research’. Current Contents was a weekly paper publication that reproduced the table of contents of many journals with authors’ addresses so we could send a reprint request card to them by snail mail. This little publication allowed us to avoid extensive trawling in a library to find out what was happening and had the additional benefit that journals in disparate areas with important papers became easily visible. The paper form of Current Contents lasted until the age of the Internet and electronic publishing.

If you only read one of Atwood’s papers, be sure to digest the review ‘Diversification of synaptic strength: presynaptic elements’ by Atwood and Karunanithi (2002). This review is well cited with 182 citations and 40 since 2013. Who is citing this paper? The records in Web of Science show that at least 27 of the 40 citations since 2013 were in papers about some aspect of mammalian or other vertebrate synapses. Besides the obvious importance of the information in this paper I think there are additional aspects that make this paper special. For instance, the articulation of Case 1 differentiation in which a single presynaptic neuron makes diverse synapses on multiple postsynaptic cells and Case II differentiation in which different presynaptic neurons make diverse synapses on the same postsynaptic cell give an important framework for theories in this area. Moreover, the examples from crustacean neuromuscular systems are paired with examples in the mammalian CNS. That triggered a short research program on cerebellar synapses in my lab and triggered our investigation of SNARE zippering in phasic and tonic synapses. Finally, the methodical examination and falsification of several hypotheses for presynaptic differentiation was critical to any researcher in this area who wished to avoid too much time in dead ends.

Trainees

In addition to his prodigious scientific output, Atwood has had many trainees including 29 M.Sc. students, 16 Ph.D. students and over 30 postdoctoral fellows and visiting scientists. Atwood gave his trainees lots of attention with frequent conferences about all aspects of their work. Atwood was a patient and kind supervisor of trainees encouraging sharp thinking and good writing. Sometimes one would see a trainee slouching down the hall carrying a manuscript or thesis draft that was almost covered in comments made with Atwood’s red pen. As drafts progressed, the overall colour of the manuscript would gradually turn from red to pinkish and the trainee would brighten up. Atwood was good at gently explaining why certain kinds of writing were better than others. He is a master of concise prose and writes beautifully.

The result of all that care with trainees is that many of them became very successful independent scientists although a few did go to the dark side and became Deans. A partial Atwood academic tree may be seen here: https://neurotree.org/beta/tree.php?pid=9062. Trainees can be found in Canada, the USA, France, Sweden, Austria, Australia and England.

Experiencing Harold Atwood

I have had the pleasure of interacting with Harold Atwood as an undergraduate, a postdoctoral fellow and as a professor in the Physiology Department while he was Chair.

Undergraduate

My experience began in 1967 early in his career as an Assistant Professor in the Zoology Department at University of Toronto. I took his neurobiology course with a few other students. As the first class met, one of the students asked ‘who is the kid in the lab coat at the back of the room?’ This turned out to be Harold Atwood. His lectures were extremely coherent, and there was a point of view. This was my first exposure to primary scientific literature. I do not remember his course being dependent on a textbook and perhaps that was part of its appeal. He did however encourage us to obtain a great book written by Bernard Katz entitled ‘Nerve, Muscle and Synapse’. I got the hardcover edition in November 1967 and still have it; it is heavily marked up and underlined. Atwood presented arguments for work on simple systems and this appealed to me. I subsequently took his laboratory course and measured resting potential in crayfish muscle, contractions of frog muscle and sensory neuron activity in cockroaches for example. There was also a project component in which I recorded electrical activity in fish muscle. “I was so much older then” (Dylan, 1964) and had taken up smoking cigarettes. Atwood caught me taking a smoke break from a frog muscle experiment and asked ‘Aren’t you afraid the NICotine will harm your experiment?’ The first syllable was emphasized by a little upward nod of the head. These two courses convinced me that I could think about neurophysiology and do something in this field. For that I will always be grateful. Atwood’s influence was evident in the work I did for my M.Sc. in the lab of Dr. Valerie Pasztor in the Biology Department of McGill University. There I used electromyography, nerve stimulation, histology and electron microscopy to study the movements and muscles of the maxilliped exopodites of the Blue Crab.

Postdoctoral experience

After a Ph.D. at the University of Texas I returned to Toronto in 1975 as Post-Doctoral Fellow in Atwood’s lab in the Zoology Department. I had a fellowship from the Muscular Dystrophy Association of Canada and we began to work on some aspects of muscular dystrophy in mouse models. We decided to study ion regulation and with another postdoc, Harold Silverman (SUNY Senior Vice Provost), I embarked on a lengthy development of double-barreled ion selective microelectrodes and a physiological preparation system that allowed penetration of muscle fibers in anesthetized mice. The system allowed measurement of membrane potential and intracellular [K⁺] with minimal standard deviations in a population of muscle fibers. Moreover, muscle contractions remained robust (Charlton et al., 1981). The techniques were used in three further papers. This episode exemplifies one of the important aspects of Atwood’s supervisory style: He allowed trainees to follow their dreams, exercise their creativity and develop their potential. A little later, as an Assistant Professor at Ohio University, I obtained a research grant from the Muscular Dystrophy Association of America that helped to establish my lab; I am sure that my experience in Atwood’s lab helped me get that grant. This supervisory style was further exemplified when he collaborated with me to restart my study of transmission at the giant fiber synapse of squid. In his lab, I designed and built a voltage clamp that allowed measurement of presynaptic Ca²⁺ currents. Again this was time consuming and costly. The development of the voltage clamp was important to my future career because it was used in projects in my own lab and formed the basis of a later design which could clamp both presynaptic and postsynaptic cells simultaneously. That instrument was used in several studies. Atwood had previously linked the squid giant fiber synapse to a quotation from Sherlock Holmes: ‘It is crushing the nut with a triphammer – an absurd extravagance of energy – but the nut is very effectually crushed all the same’ (in Atwood, 1976). We went to the Marine Biological Laboratory at Woods Hole, MA where squid are available. Harold was a good sport and dissected many squid to get the best synapses and even carted squid in a huge tub of seawater from the trawler to our lab. We published three papers on the squid giant synapse, one of which was a voltage clamp study of temperature effects on presynaptic Ca²⁺currents (Charlton & Atwood, 1979). Another paper showed that, with no presynaptic action potential, presynaptic depolarizations of only a few millivolts could evoke significant transmitter release and thus there was no threshold for release related to presynaptic action potentials (Charlton & Atwood, 1977). The ability to restart my giant synapse project was critical to getting an independent career started. My experience was not unique. Atwood helped trainees to start their careers.

Faculty experience

My next experience began when I became (1982) a faculty member in the Physiology Department at University of Toronto when Atwood was Chairman. His appointment to Physiology in 1981 was quite an innovation because he had no formal connection with the Department and he worked with invertebrates. However, he published internationally lauded research and, as I mentioned, he had successfully communicated the wider implications of his work. He was a very effective Chair and managed to strengthen the department in a time of retrenchment. He did that by increasing the emphasis on cell and molecular biology and by cross appointing faculty in other departments and institutes. His success was due to deep thinking about the Department and its future and hard work to integrate it with the medical establishment at the University of Toronto which includes hospital-based research institutes. A longer account of his chairmanship may be found in the history of the University of Toronto Physiology Department at https://indd.adobe.com/view/96d03f23-972d-451c-881d-e2a1384c3888.

Students, trainees and Faculty members appreciated that Atwood cared about them and their progress. Atwood and Charles Tator instigated our participation in the Network of Centers of Excellence on Neural Regeneration and Functional Recovery and founded and directed the Medical Research Council of Canada (now CIHR) group in Nerve Cells and Synapses. Both of these initiatives were very important to our labs and enabled us to enjoy larger grants and a couple of confocal microscopes.

Anecdotal evidence

Here, I just give a tiny bit of Atwoodiana. Atwood is quite a woodsman and naturalist. Being outdoors with him is a great pleasure and education; he knows the plants, trees and animals. Occasionally, we share observations of birds and other animals. As far as I know, he has suffered no major injuries from his chain saws. Once I asked him why he was carrying two chain saws. The answer was ‘if the big one gets stuck, I use the small one to cut it out’.

Harold has a wonderful memory and can reel off entire poems such as ‘The Cremation of Sam McGee’ (Robert Service) with just a little provocation. Sometimes in a conversation he will work in a few lines from a Gilbert and Sullivan song with a twinkle in his eye and mischievous grin. I have not found the correct command fiber to trigger this behaviour but have not given up. He can be quite funny and gave a little performance at the annual Physiology Department Christmas parties that had everyone laughing.

The Atwood lab was a friendly, welcoming place. Part of Atwood’s brilliance was to attract and most importantly, retain highly competent and pleasant lab technicians and collaborators for long periods. Ms. Irene Kwan performed electron microscopy from 1973 to 1981. Dr. Leo Marin then took over with superb results from 1981 until the lab was closed. Serial sectioning of nerve terminals was one of the strengths of the ultrastructure experts. Ms. Marianne Hegstrom-Wojtowicz was the lab manager and head technician from about 1981 until the lab closed. She ran a tight ship. My lab persuaded Leo Marin and Marianne Hegstrom-Wojtowicz to come out of retirement long enough to provide critical expertise for a project on cholesterol in Drosophila synapses (Dason, Smith, Marin, & Charlton, 2010).

A very funny lab event occurred in Atwood’s home with several students and postdocs. Participants were summoned to the celebration by Harold, blowing a hunter's horn! The event was to confer ‘The Order of the Crab’ on those who had worked on a crab species. Some got two awards because they had worked on two species.

There are reports that Atwood is a wily and fierce opponent in squash.

Finally, while imitation might be a form of flattery, in this case it represented love for Harold. I and a Postdoctoral Fellow in Atwood’s lab, Joffre Mercier (Professor, Vice President, Research, Brock University), would sometimes imitate Harold’s pleasing vocal mannerisms. We could carry on short imitation conversations and sometimes included imitations of Harold talking with his first graduate student Sabet Shokrullah Jahromi who was a loud and excitable Iranian. All this occurred rather surreptitiously. However, I once attended a meeting in Harold’s lab with several others. The lab telephone rang but Harold was too far away to answer it so I picked it up and said ‘HELL-oh’ in my imitation of Harold’s enunciation. The person on the line was fooled and Harold joined in with the general mirth.

The summing up

There is no adequate way to sum up my long experience with Harold Atwood or the importance of his work. His wisdom, advice and collaboration have enriched my career and life. He is an excellent man and scientist who I feel privileged to have known for so long.

Acknowledgements

Thanks to C. Charlton, Dr. J. M. Wojtowicz and Dr. J. Mercier and Dr. J. Dason for commenting on the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author.