Gerald Lincoln: A man for all seasons

Gerald Anthony Lincoln died after a short illness on 15 July 2020 at the age of 75 years. Gerald was Emeritus Professor of Biological Timing at Edinburgh University and a Fellow of the Royal Society of Edinburgh. He was an outstanding scientist and naturalist who was a seminal figure in developing our understanding of the neuroendocrine mechanisms underlying seasonal rhythmicity. This review considers his life and some of his major scientific contributions to our understanding of seasonality, photoperiodism and circannual rhythmicity. It is based on a presentation at the online 2nd annual seasonality symposium (2 October 2020) that was supported financially by the Journal of Neuroendocrinology.


| E ARLY LIFE AND C AREER
Gerald grew up on farms in Norfolk, perhaps explaining his lifelong fascination with the natural world and with conservation.
As a teenager, he bought a moth trap and wrote up a project investigating how weather conditions affected moth numbers on the farm. This was awarded the Prince Philip Award for Zoology and, as an undergraduate, Gerald studied zoology at Imperial College, University of London. Gerald's work on moths came to the attention of Roger Short who was working at the University of Cambridge Veterinary School, so he was offered a PhD project studying the seasonality of reproduction in red deer on the Isle of Rum, Scotland. Rum has been a world-leading centre for research subsequent to being acquired by Nature Conservancy Council in 1958. Although much of the focus on Rum has been on behavioural ecology and population dynamics and genetics of red deer, Gerald's specific interests were to investigate how behaviour and physiology were regulated by endocrine systems. 1 He showed how the stags cast their antlers in the spring when testosterone levels are at their nadir and grow new antlers when levels remain low in the summer. As the daylength decreases in autumn, testicular synthesis of androgens increases. This facilitates rutting behaviour, and the antlers stop growing and become mineralised, resulting in hard bony weapons crucial in competitive encounters with other males. 2 When he was still a PhD student on Rum, Gerald had the rare distinction of publishing a letter anonymously in Nature. 3 He had been weighing his shavings daily and noted that his beard growth increased in anticipation of returning to the main land and resuming sexual activity. This was widely reported in the newspapers at the time, although it had the important message that that testosterone production could be influenced by the higher centres. When Roger Short was appointed director of the new MRC Unit of Reproductive Biology in Edinburgh, he recruited Gerald in 1974 to study mechanisms underlying male fertility.
Arguably a major factor in Gerald's success was his decision to work with the Soay sheep as an animal model for this work.
Although the laboratory rat and primates were the established animal models at the time, Gerald's work on seasonality in deer underscored the value of studying a species in which fertility changed dramatically over the course of the year, such that the underlying neuroendocrine mechanisms could be appreciated. The Soay is a small semi-domesticated breed of sheep retaining a marked seasonal cycle, which is also convenient in practical terms for col-

| THE GONADOTROPHIN -RELE A S ING HORMONE PUL S E G ENER ATOR AND S E A SONAL CONTROL OF FERTILIT Y
Gerald developed a protocol where male Soay sheep (rams) were maintained indoors on long days comprsing 16 hours of light and 8 hours of dark for 16 weeks, then on short days comprising 8 hours of light and 16 hours dark for 16 weeks. Many overt aspects of annual seasonality were apparent in the sheep but compressed into a 32-week cycle, including testicular diameter, inguinal skin colouration, moulting of the wool and the rate of horn growth. Collection of serial blood samples from indwelling jugular cannulae at different points in this photoperiodically driven cycle showed clearly that the cycle in testis size and function reflected changes in the pulsatile pattern of luteinising hormone (LH) secretion. 6 Specifically, the sexually inactive phase in long days reflected a low frequency of LH pulses, although those that occurred were of a relatively high amplitude. By contrast, exposure to short days induced an increase in LH pulse frequency and a consequential reduction in pulse amplitude; this was associated with an increase in plasma testosterone concentrations. Although pulsatile LH secretion had been described in a number of other mammalian species by the mid-1970s, this was clear evidence that, in the sheep at least, an increase in LH pulse frequency was the primary signal to drive steroidogenesis and, alongside increased folliclestimulating hormone, gametogenesis. 7 The implication of this is that a change in hypothalamic function, culminating in a change in frequency of GnRH secretion, must be causing the change in the frequency of release of LH from gonadotrophs in the anterior pituitary. Gerald exploited a number of experimental strategies to confirm the relationship between gonadotrophin-releasing hormone (GnRH) and LH, as well as their role in control of the testis.
For example, using small portable infusion pumps, he showed that intermittent stimulation of sexually inactive rams with synthetic GnRH at 2-hour intervals would drive testicular function 8 ; thus, GnRH alone was sufficient to induce seasonal cycles in reproductive function. Conversely, he used an immunoneutralisation strategy to demonstrate that loss of GnRH led to reproductive failure. 9 The development of surgical procedures to collect blood from the portal capillaries in the median eminence of sheep confirmed the precise relationship between GnRH and LH secretion experimentally in the following decade. 10,11 Gerald interpreted this GnRH pulse frequency-modulated seasonal control of fertility in sheep as a clear indication that the key mechanistic questions about seasonality were really about central nervous system function. An example of his creative and lateral thinking was that he had found that use of an opiate-based anaesthetic widely used in veterinary practice (Immobilon) markedly suppressed LH secretion in sheep. 12 Because the Immobilon did not impair the response to exogenous GnRH and treatment with the opiate antagonist diprenorphine rapidly restored LH secretion, he inferred that these were central effects. 13 Gerald was aware of the work of Hughes and Kosterlitz with respect to identifying enkephalin peptides as endogenous opioids, 14 and so investigated in depth whether the seasonal suppression of GnRH secretion in sheep might reflect enhanced endogenous opioid activity in the hypothalamus. This was an eminently plausible hypothesis, and was exhaustively tested using the opiate antagonist naloxone as a powerful pharmacological tool. Contrary to expectations, blockade of endogenous opioids with naloxone failed to restore suppressed gonadotrophin secretion in reproductively quiescent rams, yet it enhanced pulsatile LH secretion during the breeding season. 15 Subsequent studies confirmed a key role for opioidergic systems in mediating gonadal steroid negative feedback to the GnRH secretory system rather than a role in mediating seasonal quiescence, 16 although it is worth noting that these studies on endogenous opioids preceded the more specific identification of the kisspeptinneurokinin B-dynorphin (KNDy) system by almost two decades.

| MEL ATONIN
The realisation that seasonal changes in testicular function and just four ganglionectomised rams and four sham controls revealed that pineal function was essential for synchronisation of not just reproductive function, but also the prolactin axis to changes in artificial photoperiod. [17][18][19] Radioimmunoassay confirmed that ganglionectomy ablated circadian rhythmicity of melatonin secretion, although the assays still detected above baseline circulating concentrations of melatonin. 20 Later studies using subcutaneous implants that continually released melatonin ( Figure 1) confirmed that melatonin was a key neurochemical signal in that this treatment disrupted the timing of reproductive and prolactin-regulated cycles in Soay rams exposed to artificial lighting regimens, 21 and also advanced the onset of reproductive activity in red deer stags maintained outside. 22 In retrospect, we can appreciate that the constant release of supraphysiological concentrations of melatonin served to 'blindfold' the animals to the ambient photoperiod, although these dramatic actions of melatonin in ruminants underpinned later elegant studies using timed infusions to identify the precise characteristics of the nocturnal melatonin signal that is communicated to the neuroendocrine system. 23 Gerald and other research groups exploited a miniaturised version of this continuous-release melatonin implant strategy to identify potential target sites of melatonin. 24,25 Microimplants placed in the mediobasal hypothalamus but not in the preoptic area significantly accelerated the onset of testicular regrowth 24 and initiated a rapid decline in prolactin secretion. 26 Control studies using radiolabelled melatonin revealed that the hormone diffused quite a distance away from the implant site, so precise location of melatonin target cells was not feasible.
Autoradiographical analysis of radiolabelled melatonin binding identified widespread melatonin receptor distribution in the sheep brain, 27,28 and so these microimplant studies certainly helped to focus interest on melatonin regulating classical neuroendocrine regions in sheep, whereas, in many other mammalian species, the distribution of melatonin receptors was far more limited and, in some mustelids, restricted to the pituitary gland rather than the brain itself, 29 raising the issue of how the mediobasal implant experiments should be interpreted.

| THE PAR S TUB ER ALIS
Even in sheep, the pituitary stalk (pars tuberalis) was found to be the most dense region of melatonin binding, 30 and Gerald followed up other neuroendocrine axes, although there is now compelling evidence that this is the case. Studies by many research groups including Yoshimura in quail 35 and Hazlerigg 36 in sheep identified the beta subunit of the thyroid-stimulating hormone (βTSH) as a major paracrine factor from the pars tuberalis that acts on TSH receptors in hypothalamic tanycyte cells to regulate thyroid hormone processing and, ultimately, seasonal reproductive function. 37  showed clock gene phasing in the pars tuberalis that corresponded to the ambient melatonin phasing, yet downstream seasonal physiology reversed from the initial photoperiodic state. 45 Gerald considered this as important evidence for the existence of intrinsic circannual mechanisms that were independent of melatonin-regulated timers. 45

| CIRC ANN UAL RHY THMICIT Y
Throughout his career, Gerald had appreciated the contribution of endogenous circannual rhythmicity to the generation of seasonal cycles in behaviour and physiology, recognising that seasonality not only resulted from responses to changing ambient photoperiod, but also reflected innate long-term timing processes with a periodicity of approximately 1 year. This appreciation may have arisen from his fieldwork in tropical habitats, where annual rhythmicities in reproductive cycles exist in many species despite relatively small changes in daylength and, in some cases, these are out of synchrony in a population, although individuals retain quite precise annual timing. 46 In a very long-term study with his PhD student Osborne Almeida  50 Hazlerigg and Lincoln wrote that they "conceptualize circannual rhythm generation as a phenomenon involving cyclic tissue growth and remodelling", and noted that this may take place in many structures in the adult but it is in the pituitary and hypothalamic regions that also input photoperiodic information where this process is co-ordinated and can be re-synchronised. 50 The demonstration that thyrotrophs in the ovine pars tuberalis exist in one of two states, either βTSH positive in the long day state under the control of Eya 3, or βTSH negative in the short day state characterised by high chromogranin A expression, strongly supports the cyclical histogenesis theory. 51 46

PE E R R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1111/jne.12968.

DATA AVA I L A B I L I T Y
Data sharing is not applicable to this article because no new data were created or analysed in this study.