Long‐Term Potentiation: The Accidental Discovery

ABSTRACT

Long‐term potentiation (LTP), is a type of synaptic plasticity now considered essential for learning and memory. Here I tell the story of how I accidentally discovered in 1966 in the laboratory of Per Andersen in Oslo, Norway, because I was not looking for it. It just emerged. I recount how I came to work with Per and why my result was not immediately followed up. Then, in 1968 Tim Bliss joined the lab and, on his urging, from 1968 to 1969 we did the experiments that resulted in Bliss and Lømo, 1973. I explain why I think the experiments later failed in Oslo, and for a few years also in Tim's lab in London, before it became a readily observable phenomenon. I also describe how Tony Gardner‐Medwin and I in 1971 failed to reproduce the results that Tim and I had obtained 2 years earlier in the same lab and the same type of anesthetized rabbit preparation. I tell how this failure caused me to leave the LTP field and, instead, continue exploring mechanisms of nerve–muscle interactions, which I had studied with much success during my postdoc period in London from 1969 to 1971. I reflect on Donald Hebb's influence on LTP studies and on my experience when after many years of neglect, I became interested in LTP and the hippocampus anew and started to write about it, though without doing lab experiments. Finally, I report briefly on the experiments I am doing now in retirement, studying how the nervous system regulates body temperature through varying amounts of muscle tone.

1. Introduction

The story of long‐term potentiation (LTP), in my view, started with Per Andersen in whose laboratory the phenomenon was first seen. As a medical student in Oslo in the early 1950s, Per was much impressed by his teachers in neuroanatomy, Professors Jan Jansen senior and Alf Brodal, and took time off from his studies to become a research assistant at the Institute of Anatomy. He was also strongly influenced by Theodor Blackstad and Birger Kaada, who were there at the time. Blackstad introduced Per to the hippocampus and to his detailed anatomical studies of its afferent inputs and Kaada, who had set up a section for Neurophysiology within the Institute of Anatomy, introduced Per to the electrophysiology of the brain. In 1961, Per published four single author papers in Acta Physiologica Scandinavica on interhippocampal connections based primarily on his pioneering field potential studies that wonderfully matched Blackstad's anatomically demonstrated layered inputs to the hippocampus. Per then spent 2 extraordinarily successful and productive years with John Eccles in Canberra, from where he returned to set up his own lab in Oslo in early 1964 in what then became the Institute of Neurophysiology with Kaada as its head in a building next to the Institute of Anatomy in downtown Oslo.

I joined Per in August 1964 after 8 years of medical studies, including 1.5 years of obligatory postgraduate clinical training, a year off (1958–1959) doing research on single unit activity in the visual cortex of awake rabbits at the Institute of Physiology in Pisa, Italy (Lomo and Mollica 1962), and, finally, 13 months as a doctor in the Norwegian Navy as part of my 1.5 years of compulsory military service. The first year in Per's lab, I assisted Per on projects related to synaptic excitation and inhibition in the hippocampus of anesthetized rabbits and cortical spindle activity in cats under barbiturate anesthesia. He showed me how to prepare the animals and position the stimulating and recording electrodes and, after that, would come into the lab when all was set for the actual experiments—he, a virtuoso in front of the oscilloscope, playing his instruments, the stimulators and Grass camera, I, sitting at the side of the animal moving the electrodes.

Per's lab was animated by his enthusiasm for science and his inspiring and encouraging attitude toward others in the lab. Our discussions I remember as lively and wide‐ranging. However organized discussions of results and their implications for further experiments were not part of the lab culture as I remember it. Per taught me the basic techniques and how to think about what we were doing, but once on my own for my PhD thesis, I worked independently from then on with little involvement from Per, from doing the experiments to writing the papers, except for his comments on drafts of the resulting papers, and I liked it that way.

I have been asked why I wanted to become a doctor, and later, a scientist. I did not have a special desire for either. It was a given that my two older brothers and I would go to the only University in Norway, then in Oslo, and it was somehow expected in my family that I would become a doctor. To do Science was not on my mind. It was Professor Alf Brodal, who introduced me to scientific research. During one of his lectures in anatomy, he invited a group of six students to come to his office once a week over a period of 2–3 months, not to do research, but to discuss Science and what he was working on at the time, the reticular formation of the brain stem, giving us access to his experimental material and relevant literature. I got interested, read some of the literature, and, at the end, was asked to give a talk to the class about our experience. Two years later, well into my clinical studies, I felt I needed a break and went to visit Brodal.

The visit resulted in Brodal writing to Professor Giuseppe Moruzzi at the Institute of Physiology in Pisa, Italy, and an invitation to spend a year there (1958–1959) and live rent‐free in a small room in that institute. Back in Oslo, I gave a talk on my work in Pisa to among others Per, Kaada, and Brodal. Five years later, looking for a job in a hospital in Oslo, preferably one combined with research, I met Per accidentally in a street near the University. We started talking, I joined Per, stopped clinical work, which I had come to like better than expected, and embarked on what turned out to be a long career in basic research.

For many years I found academic research stressful. I lacked self‐confidence, was prone to depressive thoughts, and felt no strong ambitions. But I liked being in the lab on my own, noticing when something unexpected happened and following up hunches. I was most pleased when I managed to set up my experiments in such a way that the results became clear without the need for statistics. Or to quote a saying attributed to the physicist Ernest Rutherford: “If your experiment needs statistics, you ought to have done a better experiment.” By seeking what clearly stood out from the noise, aiming at critical physiological processes rather than modulatory ones, being in a “discovery mode” rather than a “follow up” mode, I have been able to largely dispense with statistical analysis. If something is important, it should be there for all to see, like night and day. Looking back, I see certain independence in my approach to Science. Of course, I admire and become inspired by good work and good people, but I don't see any real role models in my life. Perhaps I became too aware of the flaws of human nature in myself and others, which sometimes could lead to strained relationships.

2. The Discovery

At the end of 1965, I started my independent work for a PhD, which we agreed should deal with the perforant path input to the dentate area and the phenomenon of frequency potentiation, which Per had just described (Andersen, Holmqvist, and Voorhoeve 1966), based on experiments he did before leaving Canberra. In the dentate, frequency potentiation occurred as a striking potentiation of granule cell discharges appearing as population spikes during stimulation of the perforant path usually at 10–20 Hz, called high frequency then. In those days, PhDs were commonly published as single‐author papers, as were both Per's and mine. PhD work was also meant to be largely independent. Consequently, I first used Per's only setup on days when he was otherwise busy, and then moved to another setup across the corridor as Per's lab expanded.

Early on in my experiments on rabbits anesthetized with urethane and chloralose, I noticed what is now called LTP. I saw the frequency potentiation described by Per, but also that potentiation could last for hours when tested by brief trains of stimuli at increasingly longer intervals after the previous train. The response to the first stimulus in the later trains was strongly potentiated with respect to amplitude and shorter latency of the population spike and with respect to an increased rate of rise of the monosynaptic population EPSP. In addition, potentiation appeared first as one, then two, and later three population spikes at short intervals in response to each stimulus during the train. From that time, I still have the abstract I wrote for a meeting in August 1966, of the Scandinavian Physiological Society in Turku, Finland (Lømo 1966), some slides I presented at that meeting, and some handwritten notes (see Lømo 2018). My presentation (10 min) was about more than the long‐lasting after‐potentiation, mentioned only at the end of the abstract (Figure 1). Further, according to my handwritten notes: “The effect is long lasting and the induced increased efficiency shows no tendency to fade after periods of rest up to 22 min in this experiment. In other experiments increased efficiency is seen for hours.” And “If it is correct that the hippocampus is involved in memory function, this is a region where one should expect long lasting changes in synaptic efficiency to occur. The phenomenon may represent a kind of bahnung [facilitation or opening] of individual synapses and may have relevance to theories of learning.”

FIGURE 1.

This abstract shows the first mention of LTP, as it is now called, published in Acta Physiol. Scand., as indicated.

The LTP described above was unexpected. In Per's earlier study (Andersen, Holmqvist, and Voorhoeve 1966), the potentiation following tetanus lasted “for several seconds, sometimes for as long as half a minute.” Synaptic plasticity was not mentioned and, evidently, not looked for. That paper was submitted in November 1965. By then, I had already been in Per's lab for more than a year and synaptic plasticity was not a matter of interest. Accordingly, my own follow up study set out to study the processes engaged during high‐frequency stimulation, not after, which included making numerous recordings of the large changes in extracellular DC potentials (large shifts in field potential between evoked responses) that occurred within and beyond the tetanized pathway during and after the high‐frequency stimulation (see Lømo 2009). But apart from the 1966 abstract I never published those results. I switched to simpler approaches based on paired‐pulse activation of granule cells by orthodromic perforant path and antidromic mossy fiber stimulation to explore the organization of excitatory and inhibitory processes within and on either side of a narrow band of activated granule cells.

Per and I were aware of the potential relevance of my findings to learning and memory and I remember well how excited we both were when I first showed him the results. But we were not interested in learning and memory and possible underlying synaptic plasticity, which may appear strange now, given Per's close relationship with Eccles and Eccles' long‐standing interest in synaptic plasticity as necessary for learning and memory. Seeing those early results, Per could have turned his lab into an LTP lab, as others did later when their importance began to sink in (Bliss and Lømo 1973). But he did not. Nor did I follow them up. We were interested in a multitude of physiological problems relating to our experimental preparation and, surely, did not appreciate then how the observed plasticity stood out from all the other results. Per had other ongoing projects and interests. After Canberra, he was invited to many meetings, wrote many papers, attracted an increasing number of researchers from inside and outside Norway to his lab, and was also interested in much outside science (see Bliss and Lømo 2024). Nor did I follow up on those early results but switched to lesser and simpler approaches to better understand the function of the underlying basic circuitry before embarking on more complex processes, such as those occurring during and after tetanization. I did not see myself as a potential great scientist searching for answers to the big questions. I was primarily interested in making my experiments work and conduct them in such a way that I could understand what was going on.

3. Lessons From the Literature

The literature that meant the most to me at the time was Per's papers on the hippocampus, Eccles' The Physiology of Synapses (1964), and his many papers on the cerebellum, which relied strongly on paired‐pulse stimulation. In Eccles' book I read about his interest in synaptic plasticity and his disappointment in failing to demonstrate it in the spinal cord beyond post tetanic potentiation and some processes lasting an hour or more after extremely prolonged unphysiological conditioning. And yet also Eccles spent several years (1964–1967) working out the circuitry and short‐time interactions between projection neurons and interneurons in the cerebellum, resulting in the book “The Cerebellum as a Neuronal Machine (1967)” with Masao Ito and Janos Szentagothai as co‐authors. None of this work addressed the issue of long‐term synaptic changes after activation. That was left to Masao Ito, who first demonstrated LTD many years later (1982) at synapses between parallel fibers and Purkinje cell dendrites. Thus, I feel my decision to focus on basic features of the system before moving to more complex phenomena, such as synaptic plasticity, to be partially vindicated.

The editors of this issue in Hippocampus have asked for clarification on some issues relating to the time around the discovery of LTP. I am quite certain that I knew little, if anything, about Donald Hebb and Hebbian synaptic plasticity in 1966 and certain that it did not influence what I was doing. Tim Bliss and I hardly discussed it, if at all, during 1968–1969 when we did the experiments that resulted in Bliss and Lømo (1973). Nor did I read or was interested in any matrix memory models that I now understand existed at the time. I may have read about patient HM but, again I am certain, it had no influence on the way I did or thought about those experiments in 1966 and I think the same can be said about the experiments Tim and I did in 1968–1969. In an email to me in 2016 Tim writes: “I am coming to the conclusion that I will never have a clear idea of the influence that Hebb had on us—and, to be specific, on me—in the early days. At McGill, I never took a course from Hebb, I heard him lecture only once, but nevertheless was aware of his book, and knew he was a very significant figure in psychology at that time. But he certainly did not have a day to day influence on the experiments I did as a PhD student with Ben Burns. It was Burns who was the influence, and I'm pretty sure I never had a discussion with him about Hebb. To Burns (as to Hebb), it was obvious that activity regulated synaptic function, either homosynaptically or heterosynaptically, and our experiments examined both potential forms of plasticity. In my thesis, and in the paper that it gave rise to (Bliss, Burns and Uttley, J Physiol, 1968), Hebb is quoted in support of the general idea of activity‐dependent plasticity, along with Pavlov, Burns and Eccles. Hebb's status I'm sure derives from his brilliant ploy of formulating the zeitgeist as an hypothesis involving two cells, A and B, and putting the resultant statement in italics (I jest, almost). I don't remember us talking about Hebb, and the fact that we don't mention him in the 1973 paper must surely reflect a lack of any particular influence on our thinking. We did wonder whether frequency potentiation of the population spike was a requirement for long‐lasting potentiation, but the expts at 100 Hz when postsynaptic firing was largely suppressed, suggested otherwise. I think it's fair to say that we were thinking broadly in what would now be called a Hebbian framework, without thinking at all about Hebb or the neurophysiological postulate. Hebb and Hebb's heirs have hijacked an idea that was widely, in fact almost universally, accepted.”

Recently, Hebb was declared a hero of neuroscience along with six others and the Hebbian synapse has become a catch‐all phrase for synaptic plasticity as expressed by LTP. Hebb's book (Hebb 1949) and Darwin's “On the Origin of the Species” (1859) have been considered “two of the most influential books in the history of biology” (Adams 1998). However, in The Physiology of Synapses (1964), Hebb's book is only one of nine papers cited by Eccles for being behind the idea that “activation of synapses increases their efficacy by some enduring change in their fine structure,” and he adds two other references for a similar suggestion. Furthermore, according to Hebb (1949), Pavlov “had already formulated a simple rule for the occurrence of learning. Any stimulus that acts repeatedly at the same time as a response will form a connection between the cortical cells formed. Subsequently, the stimulus will be sufficient to arouse the response,” a formulation that seems not that different from Hebb's. True, Pavlov, like Cajal (see Jones 1994) invoked the formation of new synapses, whereas Hebb proposed the strengthening of existing synapses. But LTP may involve both silent and existing synapses, as well as newly formed ones, as axonal sprouts quickly contact nearby spines or induce their formation.

In his brief history of “Long‐Term Potentiation,” Roger Nicoll cites 29 publications for major LTP discoveries between 1966 and 1995, but only four of them cite Hebb with the first of these citing Hebb together with eight others for proposals of models related to associative memory (Nicoll 2017). We also now know that LTP is a much more complex phenomenon than simply firing together leading to being wired together. Because Hebb's postulate is now generally applied to so many results on synaptic plasticity, it appears unfalsifiable. And yet, on specific points, it has been falsified. As we know now, postsynaptic firing is not required nor does the tagging‐capture hypothesis require coincident pre‐ and postsynaptic firing. Furthermore, Hebb's postulate of cell A repeatedly activating cell B and causing some growth process or metabolic change implies that A must occur before B, not simultaneously as the later demonstration of simultaneous activation of NMDA and AMPA receptors showed. To me, it appears that Hebb's speculation has been picked out for celebration on the back of experiments over which he had no influence, like those by Tim and me, and that the term Hebb synapse then spread like a meme to make Hebb a cult figure. Referring to an important phenomenon by a person's name is jargon in my view. It perpetuates the habit of referring to a medical diagnosis by the name of the person who first described it, rather than using a descriptive term that also informs people outside the immediate field. One might say that this use of Hebb's name reflects a general human desire to see one's name connected to something positive and important, thereby making the use of names to represent a discovery desirable. I am now, of course, aware of Hebb's towering influence for bringing the emerging field of neurophysiology and that of general psychology together to better understand human behavior in more mechanistic terms, but I do not quite see the justification for letting a particular statement of his, expressing thoughts common at the time, have such a revered status in the field. Is it really on par with Darwin's work? Hebb himself would probably agree. According to Bruce McNaughton (McNaughton 2003) Hebb expressed surprise that his hypothesis caused such excitement. The idea was an old one, he said, dating back at least to Lorente de No, and the principle was obvious to anyone who had considered the principles of associate learning. Something like cooperativity had to be present in the nervous system.

4. Bliss and Lømo, a Wonderful Collaboration

After mid‐1966, I worked on what became my thesis, which I defended in October 1969. But in August 1968 Tim Bliss came to Oslo to spend a postdoc year in Per's lab. As he describes it in this issue (Bliss 2025), upon arrival, he quickly persuaded me to set off 1 day (and night) per week to obtain the results published in Bliss and Lømo (1973). At that time, I was doing my experiments in “lab B” across the corridor from Per's “lab C.” I prepared the rabbits the same way I did for my experiments. By then I had established the lamellar organization of the perforant path input to the dentate gyrus (Lømo 1971a) and shown that the perforant path fibers pass in the angular bundle as if through a bottleneck before they fan out to innervate a medial to lateral series of lamellae in an orderly fashion. This allowed us to place a test stimulation electrode in the angular bundle to activate perforant fibers targeting granule cells in both a medial lamella (the experimental pathway) and a more lateral lamella (the control pathway). Placing the tetanizing electrode “up‐front” where the perforant path fibers enter exclusively into the medial lamella, we could then tetanize the experimental pathway without affecting the perforant path fibers in the control pathway. We worried at the time that delivering tetanizing and test stimuli through the same electrode, as we did in our initial experiments, might cause changes at the site of the tetanizing electrode, which could lead to increases in the number of nearby axons activated by the post‐tetanic test stimuli and thus be responsible for the observed potentiation at least in the shorter term (see Bliss, this issue). By placing the electrodes the way we did, we avoided any complications of this nature.

In this issue of the Hippocampus, Tim has more to say about our experiments in Oslo. Here, I add some notes on the lab's connection with Eccles, responding to questions by the editors. After I started working in Per's lab in 1964, Eccles visited many times. Once in 1965 or 1966, Per took us in his car, a rather small Peugeot for the occasion, on a 2 days tour to the Norwegian west coast, me sitting in the back seat with Sir John in the middle and his wife from Australia on his other side, feeling the close presence of such a great, and people say, dominating man, somewhat intimidating, Per driving and Victor Wilson (Rockefeller Institute, New York) in the passenger front seat (Figure 2).

FIGURE 2.

Per Andersen, John Eccles, and Victor Wilson on a ferry on Sognefjorden, the longest fjord in Norway, 1965 or 1966.

Eccles was encouraging and helpful (see below). His book, The Physiology of Synapses (1964), was essential reading for me and the main source for what I knew about synaptic plasticity. His later papers on the cerebellum were important for how I did my experiments. He was very supportive of Per and his lab. and they appeared as very good friends despite their age difference. Tim is telling the story of how excited Eccles became when in late 1968, he saw one of our early results showing LTP and then went on to show one of our figures in his books Facing Reality (1970) and in The Self and its Brain (1977) co‐authored with Karl Popper. But I cannot remember that he was told by me or Per about my earlier results in 1966 or that we ever discussed them, which may seem strange now.

Doing those experiments with Tim in 1968/1969 was a wonderful experience, full of excitement, fun, and laughter. We became, and remain, close friends and, despite some complications discussed below, we still are. Last year we came together for 3 weeks to write a biographical memoir on Per Andersen for the Royal Society and this year we have just been together for a week to write our contributions to this issue of Hippocampus, enjoying our company with as much fun and laughter as in the good old days.

5. Failing to Reproduce the Early Findings

In October 1969 Tim returned to his lab at Mill Hill outside London and I went to the Department of Biophysics at University College London (UCL) in downtown London as a postdoc to do work on nerve–muscle interactions, as suggested by Ricardo Miledi, who, at Per's request, had invited me. About once a week during my first months in London, I went up to Tim's lab at Mill Hill to continue experiments on LTP in anesthetized rabbits, where, to our consternation, we failed to induce it (Bliss 2025). I returned to Oslo in April 1971 and, in August of that year, Tony Gardner‐Medwin joined me for a year for further studies of LTP. Tony's lab was in the Physiology Department at UCL in the same building where I had just been, and where Tim and Tony had demonstrated LTP in awake rabbits with chronically implanted electrodes (Bliss and Gardner‐Medwin 1973). But in Oslo, in the very same lab where Tim and I had demonstrated LTP 2 years earlier using the same anesthetized rabbit preparation, Tony and I failed to induce LTP and, after much effort, gave up.

In lab C, across the corridor, Knut Skrede and Rolf Westgaard, who first published that transverse hippocampal slices contain a functional trisynaptic circuit from the perforant path onto dentate granule cells and then to CA3 and CA1 (Skrede and Westgaard 1971), were still working. Some comments on their paper may be of interest. They showed that perforant path stimulation evokes synaptic responses in dentate granule cells, and they write, though without showing it, that “it was possible to excite CA1 cells trisynaptically by perforant path stimulation,” and further that it was necessary to increase the stimulation frequency from 1/s to 10/s to see it, and that “trisynaptic activation was obtained in only one experiment.” They refer to C. Yamamoto's paper in Proceedings of the Japan Academy, 46 (1970) 1041–1045, entitled: “Synaptic transmission between mossy fiber and hippocampal neurons studied in vitro in thin slices” and write that Yamamoto's work was done in parallel with their own in the “same type of preparation,” namely, transverse slices rather than longitudinal slices along the length of the hippocampus, as used by Bliss and Richards (Bliss 2025). Skrede was a medical student who had taken time off from his studies to work in Per's lab. After a visit to Mill Hill in the summer of 1970, he developed the transverse hippocampal slice preparation in Per's lab C (Bliss 2025) that he and Westgaard were using when Tony and I struggled in lab B.

Failing to induce LTP in live rabbits, Tony and I decided to make our own hippocampal slice preparation. I wanted to use slices kept between two nets submerged in the physiological solution for two reasons. First, I wanted to have the slices submerged in a small perfusion chamber on the stage of a compound microscope with transmitted light from underneath. With that set up it was wonderful to be able to see individual neurons and other structures in much greater detail than with Skrede's setup, which involved a dissection microscope, illumination from above, and slices on the interface between the artificial cerebrospinal fluid underneath and the humidified and oxygenated air above. Second, I wanted to avoid what I saw as an artifact in Skrede's setup. Watching his experiments, I noticed that droplets of water accumulated on the glass recording electrode, occasionally running down the shaft and onto the slice, resulting in a long barrage of spike discharges that could be heard in my lab across the corridor.

To avoid this problem, Tony and I continued with our submerged slices but without success. The slices were in poor condition. Axonal impulses were easy to evoke, but not postsynaptic potentials. We then started to dilute the perfusion fluid with distilled water, taking our cue from the increased excitability observed in lab C when, presumably, the extracellular fluid at the site of recording became diluted by the drops of water flowing down the recording electrode. The excitability of the submerged slices did improve, but when we reached half of normal perfusion fluid tonicity, we agreed that pathology was not what we wanted to study and gave up our slice venture. Skrede and Westgaard (1971) refer to the problem this way: “Preliminary experiments indicate that the excitability of the preparation in vitro is highly dependent on the ionic composition of the bathing medium (Gardner‐Medwin and Lømo, personal communication).” We cut slices by hand then. Probably, we failed mainly because we did not master the technique of getting the slices reasonably intact and quickly into the chamber. Some years later, others obtained beautiful results in submerged slices (Alger and Nicoll 1979).

So, I see this as another missed opportunity along with not pursuing LTP when I first saw it in 1966. We could have moved to rats, but rats (or mice) did not exist as experimental animals in the institute then and we had no reason to think that rats might be more resistant to stress than rabbits and, therefore, perhaps more favorable for studies of LTP (see below). Also, we could have moved to chronic experiments with which Tony had a good experience, as I did, having set up rats for chronic stimulation of muscle during my time in London. But we did not.

I can only explain the success in demonstrating LTP in 1966 and 1968–1969 and the subsequent failures in London and Oslo by thinking that the rabbits we used in Oslo, to begin with, were less responsive to stress than those we used later. The first ones were brought to the lab by a local farmer for about £5 each, whereas the later ones were bought from professional dealers. Their backgrounds were therefore very different, and this might have affected their ability to express LTP. We know now that stress or exposure to corticosteroids can severely suppress LTP expression in the dorsal hippocampus whereas it enhances LTP in the ventral hippocampus (Segal, Richter‐Levin, and Maggio 2010). LTP expression is a subtle phenomenon sometimes difficult to elicit. Before an acute experiment, we brought the rabbit to the lab for injection of the anesthetics into an ear vein. Though the rabbits gave little obvious indication of being stressed, they must have been, as also rats and mice must be in a similar situation. But rabbits are different. In our lab setting, they tended to go into a state of torpor when injected with the anesthetic, whereas rats or mice would tend to fight, making a similar procedure for injection of the anesthetic impossible. Presumably, the degree of stress would also vary with how the animals were treated before the injection, contributing to the large variation in LTP expression that we observed among individuals. So, given the similar backgrounds for the rabbits used in London, why did LTP occur in the chronic animals but not in the anesthetized animals? Presumably, because in the chronic situation, the rabbits habituated to it over many days and were awake and not stressed in the same way when receiving the tetanus. Yet, also in the chronic animals, the variability in LTP expression was very large among individuals (Bliss and Gardner‐Medwin 1973). It appears from all this that rats and mice are more robust than rabbits and more likely to express LTP under stressful conditions. Interestingly, no one, as far as I know, has used rabbits for studies of LTP after 1971.

6. Why Did It Take 4 Years to Publish Our Results?

There are several reasons. First, we did not see our results in Oslo as complete and planned further experiments in London. Second, moving to London in October 1969, we left behind records in Oslo that were still to be analyzed when I returned in April 1971. Third, failure to reproduce the earlier results introduced hesitations and delays. Fourth, I don't remember worrying about being scooped. We published abstracts of our results in Oslo and of Tim's and Tony's results in London in the Journal of Physiology in 1970 and 1971, respectively, and nobody seemed to pick it up. We had the field much to ourselves at that stage. Fifth, after 1971, Tim and I exchanged many versions of the manuscript by snail mail, trying to make the best of the results we had and arguing over how best to present them. And, as Tim writes in this issue, it was only when Tony insisted on publishing the findings from the chronic rabbits in London that Tim and I, realizing that our paper had to come first, we got our act together and wrote up a final manuscript for submission. Importantly, we never doubted our earlier findings. There had to be another explanation for our failures.

7. Leaving LTP

Unable to proceed with our projects, Tony, while also interested in mechanisms of learning and memory from a theoretical and modeling perspective (Gardner‐Medwin 1969) then used the rest of his time in Oslo on an entirely different, self‐initiated project and left the LTP field for good. Nor did I ever again do any experiment on LTP, largely, I think, because the experiments I had done in London (Lomo and Rosenthal 1971) had been so successful and were easy to follow up. The results in London came from stimulating rat denervated muscles in vivo over many days and then showing that the sensitivity to the transmitter acetylcholine outside the neuromuscular junction returned to the normally very low or zero values in the stimulated fibers, whereas the denervated unstimulated fibers remained strongly sensitive along their entire length. At one stroke, this result falsified the generally accepted concept at the time, that some nerve‐derived trophic factor(s) acting independently of nerve/muscle impulse activity controlled the sensitivity to the transmitter outside the neuromuscular junction. After denervation, no such trophic factor could be present; instead, the controlling factor had to be evoked by muscle impulse activity.

An initiative by Jan KS Jansen also contributed to my switch of field. One day during our struggles with LTP, he came down from his lab on the floor above mine to suggest a collaboration; to establish a rat preparation for studying ectopic synapse formation by transplanting a foreign nerve onto the soleus muscle, which later would be denervated or paralyzed by chronically blocking impulse activity in the soleus nerve to make the muscle receptive for such innervation (Jansen et al. 1973). Jan then went on using this and related preparations for his projects, while I focused on other issues; the role of impulse activity, particularly patterned impulse activity, in synapse formation, in regulating contractile properties, membrane properties, transmitter release properties, and the organization of cytoskeletal proteins. Some of this work was devoted to establishing whether the many lines of indirect evidence for nerve‐derived trophic factors had other and better explanations, which slowly became more and more obviously the case. Consequently, the entire issue has now almost entirely disappeared from the literature. In an invited review in Annual Review Physiology (Lømo 2015), I reflect on why the discovery of LTP now attracts so much attention, whereas the discovery of the “non‐existence” of the postulated trophic factor eventually lost all traction. And further the significance of LTP took many years to be generally recognized but then stayed at the center of interest, whereas the work on the trophic hypothesis gained immediate attention, but then disappeared after many years of continued controversies. All of this to say that I had good reasons to switch to study nerve–muscle interactions when my attempts at following up LTP failed.

After 1971 and the failures to induce LTP in anesthetized rabbits both in London and Oslo, Tim also stopped experiments on LTP but returned to the field about 10 years later with a paper in Nature in 1982 (Dolphin, Errington, and Bliss 1982). Like me, he also missed some opportunities, as I see it. He could have started chronic experiments in his lab after his experience with Tony in his lab at UCL or begun studying the transverse hippocampal slice, which contained the CA3–CA1 synapses rather than the longitudinal slice, which was restricted to the synapses between perforant path axons onto dentate granule cells and which, as was later shown, needed blockage of spontaneous inhibition within the slice to express LTP, or he could have moved from anesthetized rabbits to anesthetized rats, where LTP has turned out to be a more robust phenomenon (Bliss 2025).

8. My Return to LTP

For about 30 years after Bliss and Lømo 1973, I took little or no interest in LTP and did not follow the literature. I had more than enough with my experiments aimed at understanding the many ways motor neurons affect muscle properties. Having left the field entirely, it was natural that Tim “took over,” also because that was his particular interest. In 1994 I was invited to be the examiner at Fredrik Asztély's defense for the PhD, one of Bengt Gustafsson's PhD students in Gothenburg. In preparing for that I began to renew my interest in the hippocampus and LTP. In 2003 I was invited to give a historical review of the discovery of LTP at the Royal Society meeting for the 30th anniversary of our 1973 paper, which further raised my interest in LTP. Around that time Tim invited me to spend 3 months with him at Mill Hill. We did not do any experiments together, but I saw what went on in his lab, there was much discussion, and I began trying to bring myself up to date on the literature.

I also began preparing for the publication of some of my unpublished results from the 1960s. After receiving my PhD in 1969, I submitted four papers to Experimental Brain Research. The editors asked for some changes, which I completed for the first two (Lømo 1971a, 1971b), but not for the last two. In March 1970 I received a letter from Eccles, the editor, saying: “I have been wondering for some time about your hippocampal papers. I am sorry that you have not made some of the suggested alterations and sent them into me because they would certainly have been accepted for publication. Perhaps though it is better that you think some more about them. Please do appreciate that the referee and I both think very highly of the papers so you should not be depressed about them.” Eccles was also one of the reviewers because I recognized his handwriting in the comments made in the submitted manuscript, which was returned to the author in those days. Anyway, the two unpublished manuscripts remained in my drawers until nearly 40 years later when I merged them into one and had it published in Hippocampus (2009). The data and the figures were the same, except for small changes in the layout of the figures, but the introduction and discussion were very different because so much more was known. It was interesting to note how much more interesting and understandable the new paper became.

Finally, it is instructive to look at the historical record and recognize how, through neglect rather than malice, it can come distorted. Writing up my paper for the Royal Society meeting in 2003, I became aware of the following statement in Science (1990): “It was in 1973 that the strengthening of synapses known as LTP was first described by Tim Bliss and his co‐workers at the National Institute for Medical Research in London” (Marcia Barinaga, Research News article). This was corrected soon afterward by Tim in a response to Science. In the textbook Neuroscience (2001, 2nd ed.) one reads: “Work on LTP began in the early 1970's when Timothy Bliss and his colleagues at Mill Hill in England discovered that a few seconds of high‐frequency stimulation can enhance synaptic transmission in the rabbit hippocampus for days or even weeks.” And similar statements were made elsewhere (Matthies 1989). Preparing for my paper on LTP published in Acta Physiologica in 2017, I began reading Tim's papers on LTP after 1982 in some detail and got the impression that my early work on LTP in 1966 was not properly referred to and that the introduction to many of his papers left the impression that the discovery emerged from the experiments we did together in 1968–1969.

I am confident this was not Tim's intention. I know he has always acknowledged my contribution to the field at meetings or elsewhere. It was also natural that he became the advocate of our work together when I left the field entirely for many years. But I also think that had I not renewed my interest in LTP and started to write about it and shown a figure from those early results as documentation, my contribution could easily have disappeared, as the above quotations seem to show. It is also important to know that had not Tim come to Oslo in 1968, it is very unlikely that I would have gone on to do similar experiments on my own, and a short abstract without any documentation would have counted for little. Tim knows all this, of course. We can strongly disagree on some issues, but they are few. We can discuss them and remain the good friends we have always been.

9. End Game

I retired from my job at the Institute of Basic Medical Sciences nearly 20 years ago but have been able to continue active research since then thanks to much goodwill from the Institute. Over the last 10 years or so, Arild Njå, another retiree at the Institute, and I have been studying mechanisms of body temperature regulation in rats aimed at determining the role of tonic motor unit activity, or muscle tone, in maintaining normal body temperature during rest or sleep. We find that the amount of muscle tone increases progressively with falling ambient temperature, representing an important heat source overlooked in the literature (Lømo et al. 2020). We are now trying to establish how much muscle tone and different types of motor units participate in heat generation as it gets colder and their importance relative to brown adipose tissue, which is attracting enormous interest today because it is thought of as a possible target for combating obesity. We are in new terrains with no competitors because, as far as I know, nobody is looking where we are looking. We do everything ourselves, occupy little space, a bench ~1.5 m long with what equipment we can stack above and below, equipment that was already in our labs when we retired. We operate on a low budget, approximately $5000 per year, mainly for animals, partly because we do not have any funds or an account at the Institute and partly because we do not need more. Expanding our project is unrealistic at our age and in our situation. As it is, it takes up most of my time and I find it engaging, exciting and meaningful.

Data Availability Statement

Research data are not shared.