Climbing the ladder from neuron to brain in Harold Atwood’s laboratory

Abstract

Twenty years spent in one laboratory is sufficient to build a legacy of publications and a body of work to make an impact. However, the impact of our work was highest at the personal level, and time spent in Harold Atwood’s laboratory was not a culmination of my career but rather a crucial path toward learning and maturing as a researcher. During that time, I experienced discoveries and lessons that shaped the next steps of my career. This article is written in gratitude for wonderful experiences and describes a few highlights that were especially memorable and influential.

Keywords:

• Crayfish
• neuromuscular synapses
• electrophysiology
• synaptic transmission

The book ‘The Organization of Behavior’ published by Donald Hebb (Hebb, 1949) had a strong influence on my interest and career in Neuroscience. Even though the book stressed neuropsychological theories of brain function, a considerable attention was devoted towards synapses. Hebb’s theories, emphasizing the brain’s cellular mechanisms, were surprisingly forward looking. Even though the term silent synapse was never used in the book, the author seemingly proposed their role and described them as synapses that are ‘near enough to excite…’. He further proposed a possible role of structural changes at synapses and growth processes within neurons as a cellular basis of learning (Hebb, 1949). This work was futuristic, since it was published at a time when mainstream psychology treated the brain and its components as black boxes and the connection between synaptic mechanisms and neuronal networks was not generally acknowledged or established.

Using Hebb’s ideas for guidance, neurophysiologists were making step-by-step progress in asserting the role of synaptic transmission in the balance of excitation and inhibition and the modifiability of neural networks in brain processes. The work of P.D. Wall served as a more recent example of modern thinking about neural networks and their plasticity. Central to this thinking was the proposed presence of ‘ineffective synapses’ and their unmasking in the context of plastic changes in the spinal cord after experimental lesions of afferent inputs (Wall, 1977). The collaboration of P. Wall and R. Melzack led to the influential gate control theory of pain, also based on early evidence of how neural circuits could change as a result of past activity (Melzack, 1973).

My knowledge about the brain’s functions was building against the background of such work but under strong direct influence of Barnard Katz’s book ‘Nerve, Muscle and Synapses’ (Katz, 1966) and Kuffler and Nicholls’s 1976 book entitled ‘From neuron to brain’ (Kuffler & Nicholls, 1976). At that time modern Neuroscience was in its infancy and there was a pervasive and wide gap between the biophysical approach toward studies of single synapses and psychological approaches toward the functionality of the brain. During my years of working with Harold Atwood I reached several ‘eureka moments’ which had strong influence on my subsequent efforts and provided guidance in my studies of mammalian hippocampal plasticity.

Eureka! A reliable preparation is the key

I arrived in Harold Atwood’s laboratory at a time when the crayfish claw opener preparation had already been well established by my predecessors S. Jahromi (Jahromi & Atwood, 1974) and L. Swenarchuk (Swenarchuk & Atwood, 1975). During my first interview/meeting with Harold we discussed some of the possible projects to be carried out, taking my electrophysiological experience into account. Voltage clamping was high on the agenda, the reason being that recording synaptic currents was a more direct way to measure synaptic transmission in comparison to traditional intracellular voltage recordings. In order to facilitate this approach, we chose to dissect very small walking legs from young crayfish and use intracellular recordings from electrically short muscle fibers. The rationale was to use these relatively small muscle cells to obtain a uniform voltage clamp along their whole length. Although this approach seemed to work, the signal-to-noise ratio obtained with such a voltage-clamp recording was not much better than with the current clamp. One possible reason was revealed several years later by the discovery of the subsynaptic reticulum and its role in filtering of the ‘missing quanta’. The subsynaptic folding of the muscular membrane formed a complex network of de facto ‘spines’ that apparently filtered the postsynaptic currents, making them inaccessible to postsynaptic voltage clamp (Figure 1). Thus, the idea of voltage clamping was abandoned and the expensive voltage-clamp apparatus sat deep in the cupboard for the remainder of my crayfish recording career. Still, the idea of missing quanta was not inconsequential and it was followed up by others using the Drosophila preparation (Nguyen & Stewart, 2016).

Figure 1. Complex membrane structure of the crayfish muscle fibers. The spine-like structures complicated voltage clamping efforts by the author. However, the observed structures may have functional significance in synaptic transmission analogous to dendritic spines in neurons. The phenomenon of ‘missing quanta’ may be attributed to the structures.

At some point, the idea of presynaptic axonal recording came about and this was a winner. Perhaps by the fortuitous choice of using preparations from young crayfish we found that axons were surrounded by a relatively tender connective tissue that was easy to penetrate. This was helped by my past experience and familiarity with an electronically controlled micromanipulator which propelled sharp micropipettes into the axon. Once inside the axon the pipette was very stable and the tight seal between the glass electrode and the axonal connective tissue ensured that the movement of the preparation during muscle contraction did not break the bond. Needless to say, this technique was later taken to a higher level by younger students in the laboratory who, like Jodi Dickstein, Don Dixon and others, managed to inject various compounds, such as calcium buffers and second messenger inhibitors, into the axon using presynaptic recording electrodes.

The success of this approach was the result of combining new techniques and the lucky choice of the crayfish preparation. The reliability of this preparation was a welcome change from my past experience working with finicky mammalian preparations. In contrast to many such in vitro preparations, it was reassuring to know that the crayfish opener muscle would respond in the same way from one day to next, enabling experiments based on past results and new ideas and skills without being frustrated by the unknown state of the cells.

Eureka! Functional and structural changes are inseparable

There were plenty of ideas to work on. Sherman, Swenarchuk and Atwood had established the existence of Long Term Facilitation (LTF) in their seminal papers (Sherman & Atwood, 1971; Swenarchuk & Atwood, 1975), but there were many questions as to the mechanism of induction and expression of the long term synaptic changes produced during facilitation. Jahromi and Atwood (1974) proposed the existence of silent, or inactive, synapses at the crustacean terminals and the obvious next step was to determine whether some of these inactive synapses become recruited to the active state in the process of LTF. Sherman, Swenarchuk and Atwood (Sherman & Atwood, 1971; Swenarchuk & Atwood, 1975) discovered that LTF was induced in conditions favoring influx of sodium into the axon, with sodium being primarily responsible for the mechanism of LTF induction. This idea was pursued and expanded in several subsequent papers (Wojtowicz & Atwood, 1985).

The mechanism of LTF was relevant not only within the field of crustacean physiology but also in the context of the mammalian nervous system plasticity. Bliss and Lømo’s discovery of the Long Term Potentiation (LTP) in 1973 (Bliss & Lømo, 1973) in the mammalian dentate gyrus, closely resembling the long term facilitation in the crayfish, presented an interesting parallel. The similarities and differences between LTF and LTP led me towards better understanding of the cellular processes underlying learning and memory and helped to open up future areas for my research.

A big advantage of crustacean preparations was the availability and accessibility of the identifiable neurons. The opener muscle in the crayfish leg, for example, is innervated by one excitatory and one inhibitory axon. Thus, there was no confusion as to the identity of the cells and an experimental idea could be pursued in successive experiments until resolution and without uncertainty due to random or not so random sampling. There were many ideas to work on. Many experimental suggestions were offered by the unique application of electron microscopy (EM). With the identifiable structures of the inhibitory and excitatory axons and the underlying muscle we got very familiar with the synaptic junctions and their unique features visible under EM. Harold’s foresight was to hire a full-time EM microscopist of considerable talent just about the time when I joined the lab. Dr. Leo Marin, who was a botanist by training, was an inspired choice since he was meticulous, very skilled and approachable. I eagerly participated in collaborative projects with Leo, since I always had a particular fascination with EM as a tool to study the nervous system. Harold and Leo took this technique to the next level by making it 3 D and quantitative. It was a great opportunity for me to join their efforts by contributing my electrophysiological expertise.

Some of the morphological features of synapses, including synaptic perforations and dense bodies, were previously unknown to me, so it was a wonderful opportunity and a steep learning curve to study them in correlation with the electrophysiological parameters yielded by our experiments. The main guiding idea was that the crayfish neuromuscular terminals possessed many silent synapses which could be recruited by an intense period of repetitive stimulation. Although I was somewhat familiar with this idea, based on my previous reading of the literature, I had never had an opportunity to work on it directly until I joined the Atwood lab. Curiously, the concept of silent synapse was not widely known and certainly not attributed to Atwood’s early work, even though Jahromi and Atwood (Jahromi & Atwood, 1974) were the first to suggest the idea. Subsequent reviews also missed this contribution of Atwood’s early work to this field. A review by Crawford and Mennerick (Crawford & Mennerick, 2012) for example, referred to presynaptically ‘dormant’ synapses giving reference to (Wojtowicz, Smith, & Atwood, 1991). Although not insignificant, this later paper was just an elaboration of the original contributions by Jahromi and Atwood (Jahromi & Atwood, 1974), that presented a detailed account of ultrastructural features. Among other things, they proposed that the synapses with prominent presynaptic dense bodies are likely to be more active, since they were often a focus for aggregation of presynaptic vesicles. These ideas proved to be seminal for our further work showing synaptic restructuring during LTF (Wojtowicz, Marin, & Atwood, 1989). I am often reminded of this work when I read in current papers that synapses in the brain are never static. They constantly change in their strength and function to accommodate both past experiences and future needs. Our crayfish studies were crucial for the reinforcement and consolidation of my early ideas along the lines of Hebb’s hypotheses. It was indeed an eureka moment proving true that one could understand the function of the nervous system on the basis of functional and structural synaptic plasticity.

Apart from the systematic and sometimes tedious work on such big ideas there were daily thrills and enjoyment of seeing phenomena not ever seen before. Our novel application of presynaptic axonal recordings with the use of small postsynaptic muscle fibers or, with macropatch recordings from axonal terminals, presented unique opportunities to observe quantal transmitter release with unprecedented resolution. It was possible, for example, to ‘tickle’ the presynaptic terminals with a depolarizing current to induce barrages of miniature synaptic potentials onto the muscle fibers (Figure 2). Quantification of this effect and the interaction of the presynaptic depolarization with the evoked action potentials were the subject of my first paper with Harold which put us at the forefront of the crustacean field (Wojtowicz & Atwood, 1984). More importantly, the paper was the basis of our progressively more ambitious experiments on quantal analysis. This theoretical approach, treating synaptic responses as randomly produced events according to the binomial distribution, was a perfect complement to the physiological and morphological approaches we were already using. It was irresistible to equate the binomial parameter ‘n’ with the number of synapses observed under EM, and ‘p’ with the probability of vesicular transmitter release. However, the binomial theory proved to be more complicated than we initially thought and we could never overcome the uncertainty of underlying assumptions. Nevertheless, the theoretical approach was intellectually stimulating and provided a basis for several publications. I believe we were also influential in promoting this approach in studies of synaptic plasticity in mammalian neurons. Several research groups around the world were using quantal analysis in studies of LTP in the hippocampus, and it was very rewarding to interact with these researchers and to have full understanding of their research and, by extension, to follow the advances in the mammalian LTP. This knowledge served me well in my transition towards research on the mammalian brain using hippocampal slices.

Figure 2. Depolarizing pulses applied through presynaptic recording electrode release barrages of quantal currents recoded by macropatch electrode (arrowheads). The technique was developed and put into use in several publications (adapted from Atwood et al., 1987).

Eureka! Collaboration is the best way to go

Harold’s research was recognized and valued by scientists around the world. Some of these researchers visited the laboratory and I was privileged to meet them in Toronto and at scientific meetings. Especially memorable were our collaborations with a dynamic duo of the Israeli scientists Hanna and Itzchak Parnas from Hebrew University in Jerusalem. Itzchak Parnas was a clever experimentalist who combined his zeal for novel ideas with the theoretical approach developed by Hanna Parnas. The Parnases visited our lab a number of times and on occasion participated in electrophysiological recordings. Together we produced three papers combining our complementary expertise (Atwood, Parnas, Parnas, & Wojtowicz, 1987; Wojtowicz, Parnas, Parnas, & Atwood, 1987, 1988). An important albeit indirect, player in this work was Professor J. Dudel (Physiological Institute in Munich, Germany) who supplied us with an ingenious ‘Dudel Amplifier’, specifically designed to stimulate and record quantal currents at the crayfish terminals (Figure 3). It had a built-in feature to compensate for the stimulus artefact and to unmask the quantal currents that were otherwise difficult to observe. The contentious idea under investigation was a ‘voltage theory’ of transmitter release promoted by the Parnases (Parnas, Dudel, & Parnas, 1986) and vigorously debated in papers and in scientific meetings.

Figure 3. ‘Dudel amplifier’ is shown in the author’s rig. The amplifier was introduced to us by Itzchak Parnas and used in collaborative experiments addressing mechanisms of quantal transmitter release.

An important scientific adversary in this field was Professor Bob Zucker (Univ of Berkley, California), who scrutinized our work and on occasion provided necessary critique of the theories developed by I. Parnas. The controversy and antagonism among these two researchers provided welcome entertainment and a stimulant for our research, but most impressive was Harold’s ability to maneuver among these people with very strong personalities and views, and to maintain good personal and scientific relations. For me, the lesson was to establish good collaborations with researchers with similar interests but different approaches and expertise. This served me well in my scientific future.

In summary, my work with Harold Atwood contributed some valuable experimental and theoretical evidence for the existence and importance of silent synapses in the crayfish neuromuscular junction. In doing so we also added substance towards building a case for considerable physiological parallels between the crustacean junctions and the mammalian central nervous system. Our work culminated in a paper (Wojtowicz, Marin, & Atwood, 1994), which made a solid case for some synapses being nearly silent while others being always active. The concept is clearly applicable to various forms of synaptic plasticity such as LTF and LTP in the mammalian brain (Bliss & Lomo, 1973). The idea of silent synapses being activated during LTP has been adopted by researchers studying the mammalian brain and expanded to include other multiple types of synapses utilizing a variety of glutamate receptors, such as AMPA and NMDA receptors (Hanse, Seth, & Riebe, 2013).

Acknowledgements

I thank Marianne Hegström-Wojtowicz for reading and commenting on the manuscript and Yao Fang Tan for help with figures and references.

Disclosure statement

No potential conflict of interest was reported by the authors.