Latitudinal differences in the circadian system of Nasonia vitripennis Floessner, Theresa
DOI:
10.33612/diss.102037680
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Floessner, T. (2019). Latitudinal differences in the circadian system of Nasonia vitripennis. University of Groningen. https://doi.org/10.33612/diss.102037680
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Latitudinal differences in the
circadian system
Latitudinal differences in the circadian system of Nasonia vitripennis. Flößner T.S.E. November 2019. PhD Thesis, University of Groningen, Groningen, The Netherlands.
This research has been carried out at the Department of Neurobiology, Groningen Institute of Evolutionary Life Science (GELIFES) of the University of Groningen (The Netherlands), according to the requirement of the Graduate School of Science (Faculty of Science and Engineering, University of Groningen, The Netherlands).
This research was founded by the EU Marie Curie Initial Training Network INsecTIME. Additional support was received from the research school of Behaviour and Cognitive Neuroscience (University of Groningen, The Netherlands).
Layout: Theresa Flößner Cover design: Nele Zickert
Printed by: Ridderprint BV, www.ridderprint.nl
ISBN: 978-94-034-2078-3
ISBN: 978-94-034-2077-6 (electronic version)
Copyright © by Theresa Flößner
All rights reserved. No part of this thesis may be reproduced or transmitted in any form or by any means without prior permission of the author.
Latitudinal diff erences in the
circadian system
of Nasonia vitripennis
PhD thesis
to obtain the degree of PhD at the University of Groningen
on the authority of the
Rector Magnificus Prof. C. Wijmenga and in accordance with
the decision by the College of Deans. This thesis will be defended in public on Friday 22 November 2019 at 11.00 hours
by
Theresa Sophie Elisabeth Flößner
born on 2 January 1986
Prof. R.A. Hut
Prof. D.G.M. Beersma
Assessment Committee
Prof. B. Wertheim Prof. J.A. Harvey Prof. C.P. Kyriacou
Chapter 1 General introduction 7
Chapter 2 Lifespan is unaffected by size and direction of daily phase shifts in Nasonia, a hymenopteran insect with strong circadian light resetting
23
Chapter 3 Increased circadian light sensitivity in a northern line of Nasonia vitripennis
39
Chapter 4 Photoperiod response corresponds to different circadian entrainment properties in northern and southern Nasonia vitripennis lines
57
Chapter 5 In search of the hymenopteran circadian light resetting mechanism
67
Chapter 6 Conclusion & Discussion 83
References 91
Summary 109
Samenvatting 113
Acknowledgments 117
Academic Curriculum Vitae 121
Chapter 1
General Introduction
Abstract
Throughout evolution life is maximising reproductive output by optimising behaviour to the environment. Rhythmic fluctuations in the environment have facilitated the evolution of timing mechanisms to generate an internal representation of these fluctuations for the organism (Hut and Beersma 2011). Conceptually, such an internal representation allows the organism to anticipate the reoccurring rhythmic changes in the environment. The most prominent timescales at which rhythmicity in the environment occurs are tidal (0.5173611 day), daily (1 day), lunar (29.530556 days) and annual (365.256363004 days) rhythms. Internal biological oscillators for tidal, daily, and annual rhythms have been described for various organisms, and recently understanding the molecular basis for tidal and annual rhythms has progressed considerably (Dardente et al. 2010; Zhang et al. 2013; O’Neill et al. 2015; Sbragaglia et al. 2015). Nonetheless, the molecular oscillatory mechanisms and their entrainment are best understood for daily (circadian) rhythms. Here we will focus on circadian rhythms in insects and mammals to illustrate concepts and definitions of daily rhythms, free-running rhythms, phase-response curves, principles of entrainment, and light resetting mechanisms.
*This chapter is modified after
Theresa Floessner & Roelof A. Hut (2017) Basic principles underlying biological oscillations and their entrainment. In: Biological Timekeeping: Clocks, Rhythms and Behaviour (Kumar V., ed.) Chapter 3, Springer, India. p.47-58.
Synchronisation and Entrainment
Circadian oscillators are internal biological mechanisms located in many tissues and cells throughout the body. They provide timing signals to the organism. Behaviour, physiological processes, and cellular processes are affected by such oscillating mechanisms. In animals, including vertebrates and arthropods, a central neuronal pacemaker optimally synchronises these various tissue and cell autonomous oscillators. In insects and mammals, the central pacemaker consists of clock neurons with strong molecular rhythmicity and which are closely communicating amongst each other to maintain coherence. In mammals, the neurons of the central circadian pacemaker are located in the suprachiasmatic nuclei (SCN), located dorsally on the optic chiasm in the ventral hypothalamus.
The central oscillator needs to become synchronised to the 24-h rhythm of day and night in the environment in order to provide a reliable representation of external time to the internal body. The synchronisation process of the internal rhythm of the body to the external daily rhythm in the environment is called entrainment. Entrainment (in biology) or resonance (in mathematics) refers to the interaction between two oscillators that leads to mode or phase locking, a notion first observed by the Dutch astronomer, mathematician, early science fiction writer, and inventor of the pendulum clock Christiaan Huygens (1629-1695). In 1665 he wrote a letter to his father Constantijn Huygens, a close friend of Descartes and Galileo Galilei, in which he described his first observation of oscillator entrainment as odd sympathy:
While I was forced to stay in the bed for a few days and made my observations on my two clocks form the new workshop, I noticed a wonderful effect that nobody could have thought of before. The two clocks, while hanging side by side with a distance of one or two feet between, kept in pace relative to each other with a precision so high that the two pendulums always swung together and never varied. While I admired this for some time, I finally found that this happened due to an odd sympathy: when I made the pendulums swing at different paces, I found that half an hour later, they always returned to synchronism and kept it constantly afterwards, as I let them go.
He published this observation in `Horologium oscillatorium sive de motu pendularium´ (1673): It is quite worth noting that when we suspended two clocks so constructed from two hooks imbedded in the same wooden beam, the motions of each pendulum in opposite swings were so much in agreement that they never receded the least bit from each other and the sound of each was always heard simultaneously. Further, if this agreement was disturbed by some interference, it re-established itself in a short time. For a long time I was amazed at this unexpected result, but after a careful examination finally found that the cause of this is due to the motion of the beam, even though this is hardly perceptible.
Indeed, the beam on which Huygens` clocks were suspended transferred a small force between both pendula, leading to synchronisation of the two oscillators. In analogy, there also must be some kind of biological force, or better signalling, from the daily environment to the internal circadian pacemaker causing synchronisation between the two oscillating processes. Such entraining stimuli are called Zeitgebers (time givers), and they will effectively generate a precise 24-h rhythm in the circadian oscillator, even if the intrinsic period of the oscillator
deviates from 24 h under constant conditions (free-running conditions). In general, the daily light-dark cycle can be considered as the most prominent external stimulus (Zeitgeber) that entrains circadian systems in various organisms.
Daily rhythms versus free-running rhythms
In nature, rhythms in activity patterns are most obvious (Fig. 1A). Activity rhythms of animals can be either diurnal (e.g. ground squirrels, butterflies, jewel wasp, most birds), nocturnal (e.g. mice, rats, moths, owls), crepuscular (e.g. Drosophila, most ungulates), or arrhythmic (e.g. arctic reindeer, mole rats). Although such activity rhythms may change depending on environmental conditions (Hut et al. 2012), they can be helpful to define markers and timescales for circadian research (Fig. 1A). Under entrained conditions the period of the activity rhythm equals the period of the Zeitgeber cycle (T). The time axis of one Zeitgeber cycle can be divided into 24 Zeitgeber hours (ZT0-ZT24), and the light onset is defined as Zeitgeber time 0 (ZT0) or light offset is defined as ZT12 (Fig. 1A).
Figure 1 Daily and circadian activity patterns and time scale definitions in chronobiology in entrained (A) and free-running continuous (B,C) conditions.
This, however, poses a problem in studies where different day lengths (duration of the light phase) are used. Daan et al. (2002) therefore defined another environmental time axis, external time, which, similar to our local timescale, uses midnight as the phasereference point
(ExT0, Fig. 1A). Interestingly, data so far indicate that the light-entrainable pacemaker in the SCN of nocturnal and diurnal mammals roughly maintains the same phase angle relative to the external light-dark cycle (Fig. 1A, lower panel), which leads to the conclusion that activity patterns are driven by a slave oscillator or a relay downstream of the SCN (Hut et al., 2012) and that the SCN under entrained conditions forms a reliable internal representation of the external light-dark cycle independent of the overt activity rhythm. To establish the endogenous origin of a rhythm, it is essential to show that the rhythm persists under continuous conditions, albeit with a different period (Fig. 1B, C). To define phase markers under continuous conditions, we need to know the activity rhythm under entrained conditions. Under continuous conditions, where the activity rhythm of the animal free-runs, we define the subjective day and night as the active and rest phase in a diurnal animal and as the rest and active phase in a nocturnal animal (Fig. 1B, C). The circadian timescale (CT) is defined by phase markers drawn from the Zeitgeber timescale, where CT0 is defined as activity onset in diurnal animals and CT12 is activity onset in nocturnal animals. Again, this may pose a problem when different durations of activity (α) or rest (ρ) are compared. As a solution, Daan et al. (2002) also defined an internal time axis (InT) where activity midpoint of a diurnal animal is defined as InT12, while activity midpoint of a nocturnal animal is defined as InT24 (Fig. 1B, C) (Daan et al., 2002). Both the circadian timescale (CT) and the internal timescale (InT) cover one circadian cycle (with duration τ) and are divided in 24 equal time steps (circadian hours, ch). When τ>24 h, a circadian hour lasts longer than one clock hour (local timescale), while τ<24 h results in a circadian hour shorter than one clock hour (thus, ch=τ/24).
Phase-response curve and phase transition curve
To understand circadian entrainment, it is useful to understand the perturbing effect of a single stimulus (e.g. light pulse) on a certain circadian rhythm. The effect of such single stimulus depends on the phase (φ) at which the stimulus is applied. The perturbation (advance or delay) of the observed rhythm can be described in terms of a change in phase or a phase shift (Δφ). The relationship between Δφ and φ is called the phase-response curve (PRC, Fig. 2). Numerous experiments have described PRCs to a variety of stimuli in a multitude of organisms (Johnson 1999). One rule of thumb can be learned from this collection of PRCs: the phase-shifting response to a stimulus is largest at those time points when the organism is usually not exposed to such a stimulus under entrained conditions. Indeed, in many organisms the phase-shifting response to light is maximal during the subjective night (i.e. the dark phase under entrained conditions, Fig. 2; Aschoff and Pohl 1978).
Figure 2 The circadian Phase Response Curve represents the phase shifting response to a stimulus plotted
against the phase of the rhythms at which the stimulus was given. Advances of the rhythm are defined as positive, delays as negative. Under constant conditions, the rhythm of an organism will free-run with a period that can deviate from 24-h. The phase at which the pulse was applied thus has to be expressed on a circadian time scale (ct), where 1 circadian hour (ch) = (period of the free-running rhythm)/24. Subjective day is defined between ct0-ct12, subjective night between ct12-ct24. The deadzone (no response) falls in the middle of the subjective day where normally light would lead to stable entrainment of the rhythm. The largest responses are found in the subjective night where it would be normally dark when the rhythm is entrained by a light-dark cycle. When activity is used as a phase marker, the subjective day is defined as the active phase in diurnal animals, while the subjective night is defined as the active phase in nocturnal animals.
Two types of PRCs can be distinguished: weak resetting PRCs (type 1; Fig. 3A, C green) and strong resetting PRCs (type 0; Fig. 3 red). The terms type 0 and type 1 PRC originate from an alternative way to present the phase-shifting effect of a stimulus. Here, the phase of the applied stimulus is expressed relative to the rhythm before the perturbation (‘old phase’) and relative to the rhythm after the perturbation (‘new phase’). This phase transition can be described in a phase transition curve (PTC; Fig. 3B, D) when new phase is plotted against old phase. The average PTC slope of a strong resetting stimulus is 0, while the average slope of a weak resetting stimulus is 1, respectively, type 0 and type 1 resetting.
The PRC and PTC can be considered as a property of the pacemaker and its shape will depend on the response and sensitivity to the stimulus. With increasing stimulus strength (or increasing sensitivity of the pacemaker to the stimulus), a weak resetting PRC (or type 1 PTC) can change into a strong resetting PRC (or type 0 PTC). Theoretical arguments predict a discontinuity during this transition, which becomes clear from the 0 transition during the subjective night (between CT12 and CT24) in a weak resetting PRC, which should jump to a 12-h phase advance (or -12-h phase delay) in a strong resetting PRC when stimulus strength increases.
Entrainment explained from the PRC: Parametric or phasic entrainment
PRCs are useful tools to understand entrainment and, more importantly, phase angle of entrainment. Because free-running rhythms usually do not occur in nature, natural selection could not directly select for a certain intrinsic period of the circadian system.
Figure 3 Circadian phase response curves (PRC) and phase transition curves (PTC) for strong (type 0) and weak (type 1) resetting. The phase shifting effect of a stimulus can be plotted against the phase of the
circadian rhythms when the stimulus was presented to generate a PRC (A). The phase resetting response can either be strong (A, red curve) or a weak (A, green curve). When plotted as a PTC (new phase against old phase, B), the average slope of the weak resetting stimulus becomes 1 (type 1 PTC; B, green), while the average slope of the strong resetting stimulus becomes 0 (type 0 PTC; B, red). The PRC of a Jewel wasp (Nasonia vitripennis) iso-female line was measured with two different stimulus strengths to observe the transition from weak to strong resetting. Comparing a weak 1-h light stimulus (C, green) with a strong stimulus 8-h light stimulus (C, red)) indeed reveals larger phase shifts for the strong stimulus. When plotted as a PTC, the type 0 (D, red) and type 1 (D, green) can be distinguished by their average slopes.
Natural selection could possibly only select for a certain intrinsic circadian period because of its influence on the phase angle of entrainment. Oscillator theory predicts that the ratio between intrinsic period and the period of the Zeitgeber (τ/T) determines the phase angle of entrainment (ψ). To see this, we can take the weak resetting PRC (Fig. 2) and let it entrain to a 24-h light-dark cycle (T=24 h). Under entrainment, the period of the rhythm = T ≠ τ. This means that daily phase shifts are needed to obtain stable entrainment according to the following relationship: T=τ-Δφ. Thus, if τ<24 h the circadian rhythm needs daily delays to entrain to T=24 h, whereas if τ>24 h daily advances are needed (Fig. 4, arrows).
Figure 4 The PRC as a tool to understand phase angle of entrainment. If τ>24 daily advances will lead to
entrainment, if τ<24 daily delays will cause entrainment. From this it follows that the phase angle of entrainment is determined by the shape of the PRC and the τ/Τ ratio.
From this it follows that τ/T, together with the shape of the PRC, determines (1) the phase angle (ψ) of entrainment and (2) the limits of the range of entrainment to different T cycles (Fig. 3, Fig. 5A, (Aschoff and Pohl 1978)). This can be simply understood in common language: With a relatively fast clock (τ/T<1), one tends to be too early, and with a slow clock (τ/T>1), one tends to be too late (see Fig. 4). Because the amplitude of the PRC increases with increasing Zeitgeber strength, the range of entrainment will also widen with increasing Zeitgeber strength (Fig. 5B). This widening of the range of entrainment is visualised in Fig. 5B, where Zeitgeber strength is plotted against the τ/T ratio, thereby creating an area where entrainment will occur: the so-called Arnold tongue, after the Russian mathematician Vladimir Arnold (1937–2010), who also happened to be a great admirer of Christiaan Huygens.
Figure 5 Phase period rule, range of entrainment and Arnold tongue. (A) Given a certain shape of the PRC,
the ratio between pacemaker and Zeitgeber period (τ/Τ) will affect phase angle of entrainment (see Fig. 3). If
τ/Τ<1, the rhythm will lead the Zeitgeber (positive ψ), if τ/Τ>1, the rhythm will lag the Zeitgeber (negative ψ). The exact shape of this relationship depends on the shape of the PRC. Because PRC amplitude will increase with increasing stimulus strength, the range of entrainment will increase with increasing Zeitgeber strength (A, compare red and blue bar). For a range of Zeitgeber strengths, the ranges of entrainment are collectively called an Arnold tongue (B).
Parametric versus non-parametric entrainment
The theory of entrainment explained by phase shifts (as explained above) is called non parametric or phasic entrainment and relies on timing of the light-dark transitions (Pavlidis 1967, Pittendrigh 1981). Despite its exceptional explanatory power, non-parametric entrainment theory was challenged by the observation that exclusively diurnal ground squirrels never see dawn or dusk under natural entrainment (Hut et al. 1999). Non-parametric entrainment can be contrasted with parametric or tonic entrainment, where the intrinsic period of the pacemaker is affected by light in a phase dependent manner (Pittendrigh and Daan 1976, Aschoff 1979, Daan 2000, Wever 1966). This phase dependency of light affecting the speed of the circadian cycle could work in synergy with PRC-based entrainment in such a way that light in the early subjective night would slow the pacemaker down, whereas light in the late subjective night would speed the pacemaker up. This could improve entrainment stability especially in diurnal burrowing animals (like the ground squirrels), which receive large fluctuations in their light environment during the day. In an attempt to combine both models of entrainment, it could be shown that entrainment stability indeed increases in diurnal animals when non-parametric entrainment works in synergy with parametric entrainment (Beersma et al. 1999; Daan 2000).
Light input to the transcriptional-translational negative feedback loop: two alternatives Type 0 phase response curves for circadian light resetting may theoretically be present in all animals when the light stimulus is strong enough, but so far they were described mostly in
insects (i.e. Drosophila). Even with very long bright light pulses, mice do not respond with strong resetting to light (Comas et al. 2006). The explanation may lie in the neurobiological molecular mechanism of light entrainment in different vertebrate and invertebrate species groups.
In mammals, it has been well established that melanopsin plays an important role as circadian photoreceptor in the ganglion cells in the retina, while classical photoreceptors like rods and cones also provide light input (Provencio et al. 1998, 2000, 2002; Berson et al. 2002; Hattar et al. 2002, 2006). The pacemaker cells in the mammalian SCN themselves are not intrinsically light sensitive. In Drosophila, however, circadian pacemaker cells in the brain are intrinsically light sensitive. Drosophila Cryptochrome (d-CRY) is a photosensitive clock molecule and plays an important, albeit not exclusive, role as circadian photoreceptor (Adewoye et al. 2015, Saint-Charles et al. 2016, Yoshii et al 2015). Mammalian Cryptochrome (m-CRY) has lost its light sensitivity over the course of evolution. Also in fish and crustacea, Cryptochromes seem to have lost their role as a light receptor and, so far, a role for Cryptochrome as a circadian light receptor has only been discovered in Diptera (flies, Drosophila melanogaster) and in Lepidoptera (butterflies, (Merlin et al. 2013). Next to the photoreceptive d-CRY, Lepidoptera also express the non-photoreceptive m-CRY (Table 1; Haug et al. 2015) for a complete phylogenetic Cryptochrome review).
Figure 6 CRY dependent and CRY independent circadian light input to the circadian molecular clock. The
transcriptional-translational feedback loop consists of central transcription factors (CLOCK and BMAL in mammals (A); CLOCK and CYCLE in Drosophila (B)), which drive transcription of period (per) and timeless (tim) genes. Light induces per expression in mammals (A), whereas light activated CRY causes degradation of
TIM, leaving PER susceptible to CK1e phosphorylation and degradation in Drosophila.
From Table 1 it can be concluded that there are at least two basic circadian light input mechanisms: CRY-dependent and CRY-independent (Fig. 6) light input. The striking difference between these mechanisms is that in CRY-independent resetting, the per gene is induced, leading to an increase of the PER/m-CRY repressor complex over the first 1–2 h of a light pulse (Fig. 6A). In CRY-dependent circadian light resetting, however, light-activated CRY causes degradation of the PER/TIM protein complex, resulting in a reduction of this repressor complex (Stanewsky et al. 1998; Emery et al. 2000b, 2000a). As a result, PER protein expression under normal light entrainment should have an opposite phase in CRY-dependent and CRY-independent circadian light entrainment systems. This seems to be the case indeed. In mammals, per mRNA in the SCN is peaking around the second half of the light phase (Sosniyenko et al. 2009), whereas per mRNA levels in Diptera only start to rise after lights off (Meireles-Filho and Kyriacou 2013).
Another important difference between type 0 and type 1 resetting mechanisms is that the direct light sensitivity of clock neurons in Diptera may lead to strong resetting because of its direct interactions with clock proteins. In mammals, all circadian photoreception is at least one synapse away from the pacemaker neurons. Second messenger pathways within the circadian pacemaker neurons offer more possibilities for the circadian entrainment system to attenuate and regulate circadian phase-shifting properties in response to a light pulse. Such processes may have helped to regulate and stabilise entrainment to light in mammals, which had to rely on a nocturnal and burrow-dwelling lifestyle in their early evolution (Gerkema et al. 2013).
Involvement of the circadian system in photoperiodic time measurement Latitudinal variation of the circadian system
Circadian systems can be involved in seasonal responses to the change in duration of the light phase (photoperiod) or dark phase (scotoperiod) over the year. Due to the tilt of the Earth’s rotational axis, photoperiod changes predictably over the course of the year. Besides time of year also latitude plays an important role in annual changes in photoperiod. Close to the equator the fluctuation in photoperiod is very small over the year which becomes greater further to the poles, up to constant light or constant darkness in the (ant-)arctic summer or winter, respectively. Annual increase in photoperiod causes a delayed increase in temperature, leading to an annual hysteresis pattern that expands with latitude; the latitudinal expansion of the photoperiod-temperature ellipsoid (Fig. 7, Hut et al. 2013). Due to the different temperature - photoperiod relationship at different latitudes, the circadian system is challenged in different ways and needs adaptation to the geographical location to enable circadian synchronisation to the environment and appropriate seasonal responses to changes in photoperiod.
Figure 7 Annual patterns of (A) photoperiod, (B) temperature, for different latitudinal ranges and (C) Ellipse-like annual PPT curves predict changing photoperiodic responses with changing latitude and altitude; (A) Photoperiod calculation based on civil twilight times at dawn and dusk. Civil twilight (solar altitude
68 below the horizon) is the moment when log light intensity changes most rapidly and is often considered as the time of ‘lights on’ and ‘lights off’ for biological systems. (B) Monthly mean dry bulb air temperatures (mostly between 1960 and 1990) from 873 weather stations around the world obtained from the World Meteorological Organization (http://www.wmo.int), and globally averaged over 108 latitudinal bands. Colours indicate mid-latitude of each band (hemispheres plotted separately). (C) The annual temperature hysteresis leads to an ellipse-like relationship between temperature and photoperiod, with higher temperatures in autumn than in spring (dots indicate the mid-point of each month). The dotted lines indicate a hypothetical threshold temperature at 10 °C at which a certain species starts winter dormancy (e.g. diapause), resulting in a shift towards longer critical photoperiod (CPP) earlier in the year when this species moves north. This fundamental process forms the basis for the expectation that latitudinal clines in photoperiodic response mechanisms may exist in nature (figure and figure caption taken from Hut et al. 2013).
One aspect of latitudinal adaptation in the circadian system can be visualized by mesuring its intrinsic period. For example, field lines of linden bug Pyrrhocoris apterus show latitudinal variation in free-running periods from ~21 h up to 28 h with short periods in lines from southern Europe mainly and long periods in lines from northern Europe (Pivarciova et al. 2016). Opposite latitudinal clines were reported in Drosophilid flies. D. subobsura (Lankinen 1993), D. littoralis (Lankinen 1986a) as well as D. auraria (Pittendrigh and Takamura 1989) all showing shorter free-running periods in northern regions. It is not clear why there are opposite circadian latitudinal trends in free-running period, but it might relate to different light resetting mechanism in various insects (Fig. 6). Different circadian light reception properties may perhaps also lead to different selection pressures on circadian capacities. For instance, temperate zone Drosophila species can show arrhythmicity in long photoperiods or constant light. Natural selection in Drosophila may thus have favoured mechanisms to avoid circadian arrythmicity to occur at higher latitudes. Pittendrigh therefore hypothesized that reduction in light sensitivity with increasing latitude may prevent weakening of the circadian pacemaker amplitude (Pittendrigh et al. 1991).
Fruit flies are model organisms to study the genetic and behavioural background of the insect circadian system. Several tools were developed to manipulate Drosophila circadian genetics and to explore its neuronal circuitry and its specific clock gene profile that expresses latitudinal variation. By imitating the expression profile of d-CRY and neurotransmitter PDF (pigment dispersion factor) of the northern Drosophila species D. virilis, D. melanogaster also adapts to their locomotor activity profile. That indicates the important role of d-CRY and PDF in the process of circadian adaptation to different photoperiodic conditions and environments (Menegazzi et al. 2017). Furthermore, Drosophila population genetics also provided evidence for circadian adaptation to latitude. The clock gene per shows length polymorphisms in a threonine-glycine (Thr-Gly) repeat encoding region following a latitudinal cline (Costa et al. 1992) at which the longer type is more abundant at higher latitudes. This leads to the suggestion that its function in thermostability contributed to a circadian phenotypic selection (Sawyer et al. 1997) by better adaptation to colder temperatures in northern flies after colonisation from the warmer afrotropical climate (Stephan and Li 2007). Additionally, a study by Tauber et al. 2007 in D. melanogaster showed polymorphisms in the clock gene tim, encoding a light-sensitive circadian regulator (Suri et al. 1998, Yang et al. 1998). Here a latitudinal cline was identified along a European north-south axis in allelic frequency of an alternative start initiation variant (Rosato et al. 1997). Additionally, diapause induction followed a similar latitudinal cline, with induction later in the year in the northern population than in the southern (Sandrelli et al. 2007; Tauber et al. 2007). These mutual correlations between circadian modifications and latitude on the one hand and changes in photoperiodic response with latitude on the other hand form a piece of evidence for a modulating role of clock components in photoperiodic processing and response (Tauber et al. 2007; Zonato et al. 2017, 2018). In N. vitripennis, populations originating from different European latitudes, QTL analyses revealed polymorphisms in per showing a clinal distribution of two main haplotypes of per correlating with the latitudinal cline in diapause induction (Paolucci et al. 2013, 2016). Additionally, per RNAi in N.vitripennis confirmed the important role of the clock in photoperiodism as knock-downs of per causes a reduction of diapause incidence under short photoperiod (Mukai and Goto 2016). Together
these studies show that in insects, circadian organisation, clock genes, circadian gene expression patterns, activity profiles and photoperiodism show adaptation to latitude.
Photoperiodism and variation in photoperiodic response as a function of latitude
Anticipation to annual environmental changes is necessary for optimal timing of mating, growth and rest (diapause). Many organisms use photoperiod to drive seasonal responses, as it is a reliable, stable and recurring measurement (Hut et al. 2013). The transition from the active phase (mating and growth) to the inactive phase (diapause) is well studied in insects. When for a number of successive days photoperiod falls below a critical duration a switch occurs and a pathway is activated by which physiological processes induce diapause, a form of dormancy. A so called “photoperiodic timer” measures photoperiod and a “photoperiodic counter” determines the number of days of that photoperiod (Saunders 2002). The photoperiod, at which 50% of the population is in diapause, is defined as critical photoperiod (CPP). Diapause is characterized by developmental and metabolic arrest as well as stress resistance (Tauber et al. 1986). Diapause responses are genetically determined and may occur in species specific life stages (Tauber et al. 1986). Geographical variation in CPP was shown in many species (Danilevskii 1965, Hut et al. 2013; Fig. 8). The data express a clear north-south gradient with longer CPP at longer latitudes.
Figure 8 Latitudinal clines in photoperiodic and circadian timing mechanisms in insects. Latitude (in ° N)
correlates with critical photoperiod for diapause induction (populations above 800 m altitude were excluded); Brown, Sericinus montelus (pupae); black, Wyeomyia smithii (larvae); purple, Bruchidius dorsalis (larvae); pink,
Chrysopa carnea (adult); turquoise, Homoeosoma electellum (larvae); khaki, Tetranychus pueraricola (adult);
cyan, Orius sauteri (adult); dark blue, Acronicta rumicis (larvae); red, Nasonia vitripennis (larvae; maternally induced); green, Drosophila montana (adult); grey, D. phalerata (adult); blue, D. transversa (adult). (figure caption taken from Hut et al. 2013).
The involvement of the circadian system in photoperiodic time measurement is a controversial topic, well discussed over decades with different results for different species. Due to the fact that light has a key role in daily entrainment as well as in seasonal timing it is tempting that one could influence the other. Protocols were developed (Nanda and Hamner 1958; Bünsow 1963) to support that photoperiodic time measurement changes in resonance with the circadian period length (~24 h or a multiples of it). When such resonance patterns are observed, the species is called a positive responder, e.g. in N.vitripennis (Saunders 1968, 1974). A negative responder refers to a stable photoperiodic response independent from the external period length, as observed in, for instance the European red mite Metatetranychus ulmi (Lees 1953).
Erwin Bünning proposed in 1936 a model to explain the involvement of the circadian system by a dual-phasic oscillator with a frequency of approximately 24 hours. The oscillator consisting of a photophil phase of 12 hours that required light and a scotophil phase of also 12 hours requiring darkness. A short photoperiodic response, diapause for instance, is expected when light only occurs during photophil whereas a long photoperiodic response when the light phase is stretched into scotophil. This model was modified by Pittendrigh & Minis in 1964 and Pittendrigh 1966, into the ‘external coincidence model’ which states the occurrence of short photoperiodic induction (e.g. diapause) when a photosensitive phase of a single oscillator coincides with external darkness. Accordingly, a long day response (e.g. non-diapause) emerges when the photosensitive phase coincides with light. The oscillator could either be a self-sustained circadian pacemaker (Pittendrigh 1966), a dampening pacemaker (Lewis and Saunders 1987) or a slave oscillator (Vaz Nunes et al. 1991b, 1991a). In 1972, Pittendrigh proposed the alternative ‘internal coincidence model’ with two oscillators: one long period oscillator (τ > 24h) coupled to dawn and one short period oscillator (τ < 24h) to dusk. A change in photoperiod would necessarily lead to a phase angle difference between the dawn (morning) oscillator and the dusk (evening) oscillator. Evidence for both, the internal and external coincidence timing models for photoperiodic response, has been found in the hypothalamus of mammals (Dardente et al. 2003, 2010; Hazlerigg and Wagner 2006; Masumoto et al. 2010; Hut 2011), but a similar level of molecular evidence is currently missing in insects. Latitudinal variation of free-running period, polymorphisms in clock genes and diapause induction expresses strong evidence for interactive pathways between the circadian clock and seasonal timing. But these pathways, cross-links and light reception are known just partly and only for certain species or groups. Interestingly, internal and external coincidence timing models have been proposed to lead to different predictions on evolutionary adaptation on circadian period, where a latitudinal cline of circadian period will be limited towards 24h under internal coincidence timing, but may readily cross the 24-h limit under external coincidence timing (Hut & Beersma 2011). Latitudinal cline data on circadian period collected in insects thus far have not provided clear evidence for either of the two photoperiodism models (Hut et al. 2013).
Nasonia vitripennis as a model organism
Well-known members of the hymenoptera are ants, bees and wasps and comprising overall more than 115,000 described species. They show a high diversity of life styles, from socially
organized to solitary species and are abundant in most habitats. Furthermore, they are probably the most beneficial insect order to humans either as natural enemies of insect pests (parasitic wasps) or as pollinators of flowering plants (bees and wasps).
The jewel wasp Nasonia vitripennis is a parasitoid hymenoptera species that lays eggs into fly pupae, thereby providing protection and nutrients to the growing progeny. In recent times, Nasonia became a model for evolutionary and population genetics, speciation and ecology, but was studied before in the area of behavioural research and chronobiology. N. vitripennis shows a distinct photoperiodic response by diapause induction in the offspring in response to shortening of the photoperiod (Saunders 1965a, 1966a). Additionally, it expresses a strong and stable circadian activity rhythm with a clear diurnal pattern (Bertossa et al. 2013). Remarkable work was done by David Saunders by investigating photoperiodism involving the circadian system. Saunders provided evidence indicating circadian involvement in photoperiodism that seems to follow the rules of the internal coincidence model (Saunders 1962, 1965a, 1966a, 1966b; Saunders et al. 1970), although he himself recently questions that interpretation (David Saunders, personal communication, 2018).
Aim and chapter overview of this thesis
In this thesis, we used Nasonia vitripennis to investigate latitudinal adaptation in circadian light resetting and its effect on survival and photoperiodism. Together this will lead to a better understanding of the effect of photoperiod on the circadian system and the mechanisms that drive adaptation of the biological clock to the seasonal environment. Seasonal sunlight patterns change with latitude and therefore it is expected that the circadian system and the seasonal timing system show latitudinal adaptation (Hut et al. 2013). To study latitudinal adaptation, Nasonia vitripennis populations were collected from different geographical locations in Europe (Paolucci et al. 2013).
In comparative studies in two populations from extremes on the European north-south axis, we addressed the adaptive output of the clock by determining free-running period (Chapter 2) and circadian light sensitivity by measuring phase response curves under various light durations and intensities (Chapter 3). When compared to the southern line, we observed longer free-running periods and increased circadian light sensitivity in the northern line. Because the circadian system in N. vitripennis has been proposed to be involved in photoperiodic diapause responses (Saunders 1968, 1969), lengthening of intrinsic circadian free-running period might be an indication of latitudinal adaptation for photoperiodic diapause timing. To study the circadian involvement in photoperiodic diapause response, we conducted a partial Nanda-Hamner protocol (Chapter 4). Indeed, the results confirm that the critical photoperiod at which Nasonia females change from producing non-diapausing larvae to diapausing larvae is different for the northern line (longer free-running period) and the southern line (shorter free-running period than the northern line). We propose that the results can be explained by an external coincidence timing model. Circadian light resetting is critical for adaptive circadian entrainment of Nasonia and seasonal timing of diapause, but the molecular mechanism by which hymenopteran insects entrain is currently unknown.
To identify and locate the circadian light resetting mechanism of N. vitripennis clock work, we tested for immediate early gene induction of canonical clock genes by light (Chapter 5). In addition, we measured opsin gene expression levels to identify possible mechanism by which the north-south difference in light sensitivity might be explained. Latitudinal adaptation in Nasonia diapause may have resulted in lengthening of the circadian period at higher latitudes (Chapter 3 & 4), but the circadian resonance hypothesis predicts fitness costs in terms of reduced survival due to larger daily phase shifts needed for daily circadian entrainment. Consequently, the expectation is that the more the circadian period deviates from 24-h, the more longevity is reduced. We tested this longevity hypothesis in a range of environmental light-dark periods (T-cycles) and photoperiod combinations (Chapter 2). Higher longevity is expected to occur in T-cycles close to free-running period, but such a relationship was not found in our northern and southern lines. In the conclusion & discussion (Chapter 6), we hypothesize that Nasonia optimized its circadian light sensitivity to strong resetting (type 0, Chapter 3) to maintain a fixed phase angle of behavioural entrainment, independent of intrinsic circadian period. This fixed phase angle of entrainment might be the reason that survival is not compromised when large deviations between circadian period and environmental period may evolve to optimize photoperiodic adaptation.
Chapter 2
Lifespan is unaffected by size and direction of daily
phase shifts in Nasonia, a hymenopteran insect with
strong circadian light resetting
Abstract
Most organisms have an endogenous circadian clock with a period length of approximately 24 h that enables adaptation, synchronization and anticipation to environmental cycles. The circadian system (circa = about or around, diem = a day) may provide evolutionary benefits when entrained to the 24-h light-dark cycle. The more the internal circadian period (τ) deviates from the external light-dark cycle, the larger the daily phase shifts need to be to synchronize to the environment. In some species, large daily phase shifts reduce survival rate. Here we tested this ‘resonance fitness hypothesis’ on the diurnal wasp Nasonia vitripennis, which exhibits a large latitudinal cline in free-running period with longer circadian period lengths in the north than in the south. Longevity was measured in northern and southern wasps placed into light-dark cycles (T-cycles) with periods ranging from 20 h to 28 h. Further, locomotor activity was recorded to estimate range and phase angle of entrainment under these various T-cycles. A light pulse induced phase response curve (PRC) was measured in both lines to understand entrainment results. We expected a concave survival curve with highest longevity at T=τ and a reduction in longevity the further τ deviates from T (τ/T<>1). Our results do not support this resonance fitness hypothesis. We did not observe a reduction in longevity when τ deviates from T. Our results may be understood by the strong circadian light resetting mechanism (type 0 PRC) to single light pulses that we measured in Nasonia, resulting in: (1) the broad range of entrainment, (2) the wide natural variation in circadian free-running period, and (3) the lack of reduced survival when τ/T ratio’s deviates from 1. Together this indicates that circadian adaption to latitude may lead to changes in circadian period and light response, without negative influences on survival.
Theresa S.E. Floessner1,Floor E. Boekelman1, Stella J.M. Druiven1,2, Maartje de Jong1, Pomme M.F. Rigter1, Domien G.M. Beersma1, Roelof A. Hut1
1Chronobiology unit, Neurobiology expertise group, Groningen Institute for Evolutionary Life Sciences,
University of Groningen, the Netherlands.
2Current address: Department of Psychiatry, University Medical Center Groningen, University of
Groningen, the Netherlands.
Introduction
The circadian system regulates temporal organisation of physiological and behavioural processes. Cyanobacteria, fungi, plants, insects and vertebrates have an internal circadian system that oscillates with a period length of about 24 h (τ, circadian free-running period). The fact that circadian systems are present in many taxa indicates that synchronization to the environmental 24-h light-dark cycle is an evolutionary adaptation. The circadian period usually deviates slightly from 24 h, which may be a way to optimize phase (Hut et al., 2013; Pittendrigh and Daan, 1976; Chapter 1). Larger deviations of the circadian period from 24 h may carry fitness costs by reducing intrinsic survival rates due to increased daily phase shifts required for entrainment to 24 h (Pittendrigh and Bruce 1959). Understanding possible fitness costs for daily phase shifts becomes particularly relevant for modern societies where increasing occurrence of jetlag, social jetlag and shift work form health threats (Rosa 1995; Folkard and Lombardi 2006; Wittmann et al. 2006; Su et al. 2008; Chen et al. 2010; Levandovski et al. 2011; Roenneberg et al. 2012; Gan et al. 2015; Plano et al. 2017).
The circadian resonance hypothesis, originally proposed by Pittendrigh and Minis (1972), states that organisms with a circadian clock resonating with the environmental cycle have better internal synchrony and longer life span. Phrasing it differently, organisms that need larger daily phase shifts for circadian entrainment may have reduced fitness. This ‘resonance fitness hypothesis’ is thought to apply for entrainment to external light-dark cycles (LD cycle) with a range of different period lengths (T; Zeitgeber period) within the range of entrainment (Chapter 1). Using different T-cycle durations, it can be tested whether survival is reduced the further τ deviates from T (τ/T ratio deviating from 1) for both, daily delays, where the internal circadian period (τ) is shorter than the environmental period (T), and daily advances, where τ>T. As a result, the survival curve as a function of the Zeitgeber period is expected to have a concave shape (inverted U-shaped; Fig. 1).
Figure 1 Circadian resonance fitness hypothesis: Survival as a function of T-cycle. Highest survival is reached
when the circadian period is in resonance with the external cycle. A) When T > τDD, the system needs daily phase
delays to entrain to longer environmental T-cycles, when T < τDD, the system needs daily phase advances (daily
phase shift = τ – T). The more T deviates from τ, the larger the entraining daily phase shifts have to be. B) Survival decreases when the circadian system has to entrain to the environmental cycle by large daily delays (T>>τ) or advances (T<<τ).
Supporting the resonance fitness hypothesis, early studies found that longevity of flies is reduced the further T-cycles deviate from 24 h (Aschoff et al. 1971; Pittendrigh and Minis 1972a). Furthermore, Drosophila flies with mutations of the period gene (per) causing different free-running periods (perL τ≈29 h and perT τ≈16 h), have about 10-20% reduction in life span under T=24 h compared to wild type animals (Klarsfeld and Rouyer 1998). However, not all results of their study are in line with the resonance fitness hypothesis, since survival of wild-type and perL flies does not differ significantly between T=24h and T=16h. Similarly, in mixed
bacterial solutions with wild type and kaiC mutants cyanobacteria strains, the strain with a free-running period closest to the environmental period length outcompeted the strains with a circadian period deviating from the environmental period (Ouyang et al. 1998).
The impact of repeated phase shifts on health was tested with a chronic jetlag protocol in mice (Davidson et al. 2006). It was shown that phase shifts of 6-h advances or delays applied every 7 days cause reduced life span. The authors argue that internal desynchrony between central and peripheral clocks may cause shorter life span. Remarkably, in Davidson et al., 2006 phase advances seemed more detrimental than phase delays, possibly because the phase response curve in mice shows stronger delays than advances. This suggests that entrainment to an advancing jetlag protocol might be less feasible for the mouse circadian system than entrainment to a delaying protocol. In addition, mice carrying the τ-mutation (casein kinase 1ɛ
mutation) with circadian periods much shorter than 24 h (Meng et al. 2008) had reduced survival and reproduction under semi-natural conditions (Spoelstra et al. 2016). Thus, fitness effects of light synchronization have been observed in bacteria, insects and mammals, although the range of tested species is restricted.
In addition, some of these studies were performed with clock gene mutations that may carry pleiotropic effects besides a clock phenotype. The mammalian τ-mutation, for example, has reduced fitness in mice under natural conditions (Spoelstra et al. 2016) but also causes an upregulation of metabolism by about 20%, (Oklejewicz et al. 1997; Meng et al. 2008). We tested the circadian resonance hypothesis in the parasitic jewel wasp Nasonia vitripennis (hymenoptera), which has a broad global and latitudinal distribution and is expected to exhibit large natural variation in circadian period length and circadian light sensitivity. We used a range of light-dark cycle periods and photoperiod combinations to gain better insight into the potential evolutionary trade-off between survival and circadian entrainment phase shifts.
Materials & Methods Experimental lines
All experiments were performed with two Nasonia vitripennis isofemale lines, one originating from Oulu, Finland (65.01°N; northern line) and the other from Corsica, France (42.04°N southern line; Paolucci et al. 2013). These lines were reared in a temperature and humidity controlled climate chamber (20 ±1°C, 50-55% RH), under a daily light-dark cycle of 16 h of light and 8 h of darkness (LD 16:8). The wasps used during these experiments were offspring from individually housed females, supplied with Calliphora spp. pupae as hosts.
Longevity assessment
Three to five days after emerging from their host puparia, males and mated females were separated and transferred to the experimental setup to assess longevity. Wasps were kept in polystyrene tubes (6 x 1 cm) in groups of 10, with food (0.5ml, 30% sucrose, 1.5% agar, 0.15% nipagin) and placed in ventilated light-tight boxes (23 x 14 x 32 cm) in a climate controlled room at 18 °C (±1 °C). Each box was equipped with one LED light source (Neutral White 4000K PowerStar, Berkshire, UK) providing 2.1 · 1015 photons·cm-2·s-1 at the level of the animals. Different groups of wasps were exposed to a range of continuous light-dark cycles with different T-cycles, generated by changing either the light or the dark phase duration (Table 1). Throughout the entire assessment period, the light-tight boxes were opened each day during the light phase and dead animals were counted. To assess differences in mean longevity, a Kaplan-Meier survival analysis was performed in SPSS (IBM Corp. Released 2016. IBM SPSS Statistics for Windows, Version 24.0. Armonk, NY, USA: IBM Corp.). Differences in the shapes of the survival curves were tested by a Tarone-Ware test. Average survival differences between sexes, strains and conditions were tested with a general linear model regression analysis (Statistix 8, Analytical Software, USA), including a quadratic τ/T term to test for the expected concave relationship between τ/T ratio and longevity (Fig. 1).
Table 1 Applied external light-dark cycles. The duration of the light-dark cycle (T-cycle) was changed either
by changing the duration of the light phase while keeping the dark phase constant at 8 h (∆L:D), or by changing the duration of the dark phase, while keeping light phase duration constant at 16 h (L:∆D). Number (n) of animals per light condition
Entrainment assay & determination of circadian free-running period (τ) and phase
Locomotor activity of individual wasps was recorded to observe their ability to entrain to the different T-cycles and photoperiods (DAMS, Drosophila Activity Monitoring System; TriKinetics, Waltham, USA). Furthermore, we determined τ in constant darkness after entrainment to LD 16:8 in 2.1 · 1015 photons·cm-2·s-1. Three to five day old mated females and males from the northern and southern lines were individually placed in activity tubes (6.5 x 0.5 cm) filled one quarter with agar food (30% sucrose, 1.5% agar, 0.15% nipagin) and closed with an air permeable plug. Tubes were placed in 32-slot recording monitors, with each slot containing an infrared light beam. The monitors were placed into the light-tight boxes (see above) at 18°C (±1 °C) with the following light regimes: T-20 h (LD 16:4), T-24 h (LD 16:8), and T-28 h (LD 16:12). The activity of each individual wasp was recorded as the number of light beam interruptions per minute by the DAMSystem. Actograms were created with the ImageJ plugin software ActogramJ (Schmid et al. 2011). Average activity profiles were calculated using data from day 10 to 15 for each individual. Following this, the data of each experimental group and condition were smoothed using a moving average of 11 bins and normalized to the highest value as described in Schlichting and Helfrich-Forster, 2015. Circadian free-running period of northern and southern females and males were determined in Microsoft Excel by using a χ²-periodogram analysis (Sokolove and Bushell 1978) as described in Hermann et al., 2012.
To determine phase-period relationship, phase determination was conducted in ChronoShop (Spoelstra et al. 2004) by determine centre of gravity per individual, averaging of
LD-cycle (h) T-cycle (h) north (n) south (n) north (n) south (n)
∆L:D 12:8 20 99 100 101 92 ∆L:D 14:8 22 99 100 100 103 ∆L:D 16:8 24 100 101 101 100 ∆L:D 18:8 26 99 99 100 100 ∆L:D 20:8 28 100 99 96 100 L:∆D 16:4 20 20 40 38 29 L:∆D 16:6 22 20 40 50 30 L:∆D 16:8 24 20 40 50 20 L:∆D 16:10 26 20 40 50 20 L:∆D 16:12 28 20 40 50 19 female male
five consecutive cycles of stable entrainment (around day 10 to 15) and further averaging within the experimental group.
Phase response curve
To understand the light resetting capacity of the circadian system, a circadian phase response curve was established to single light pulses at different phases of the circadian cycle for both sexes from the northern and southern line. We applied the Aschoff type II PRC protocol (Aschoff 1965), using LD 16:8 entrainment followed by free-run in continuous darkness (DD), where the light stimulus was applied on the third day of release in DD. Each group received a single 8-h light pulse of 2.1 · 1015 photons·cm-2·s-1 at twelve different time points separated by 2 h across the circadian cycle. The phase shift was quantified by recording the free-running activity rhythm in constant darkness after the light pulse. The phase of the light pulse was calculated relative to the light dark cycle, while applying free-run phase correction in DD before the light pulse was applied. Individual phase shifts were determined with ChronoShop software methodology (Spoelstra et al. 2004) using circular centre of gravity.
Results
Survival curves
Overall, females survived longer than males (mean survival females 16.6 days (SEM 0.71), males 10.7 days (SEM 1.03); T=4.75, df=38, p<0.0001). Females lived longer in L:∆D than ∆L:D conditions (Fig. 2, F1,17=13.07, p<0.002), both in the northern and southern strain (Fig. 2, F1,17=0.02, p<0.91). Females under LD cycles with different periods showed significant differences in survival only when the duration of the light phase was changed to modify the period of the LD cycle (Fig. 2A). Survival seemed independent of the duration of the dark phase (Fig. 2B).
Figure 2 Survival curve for females from northern and southern lines in five different T-cycles. Different
curves indicate proportion surviving wasps over time (in days of 24 h) LD cycle length (T) equals 20 h (black), 22 h (blue), 24 h (green), 26 h (orange), 28 h (magenta). A Kaplan-Meier survival analysis was performed and a Taron-Ware test was applied using SPSS. A) ∆L:Dnorthern line: p<0.01, χ²=97.77; ∆L:D southern line: p<0.01, χ²=57.91; B) L:∆Dnorthern line: p=0.08, χ²=8.49; L:∆D southern line: p=0.14, χ²=6.87. During the light phase survival has been assessed of each individual.
Figure 3 Survival curve for males from northern and southern lines in five different T-cycles. Different
curves indicate proportion surviving wasps over time (in days of 24 h) LD cycle length (T) equals 20 h (black), 22 h (blue), 24 h (green), 26 h (orange), 28 h (magenta). A Kaplan-Meier survival analysis was performed and a Taron-Ware test was applied using SPSS. A) ∆L:Dnorthern line: p<0.01, χ²=87.39; ∆L:D southern line: p<0.01, χ²=111.35; B) L:∆D northern line: p<0.01, χ²=14.93; ∆L:D southern line: p=0.05, χ²=9.57.
Males also lived longer in L:∆D than in ∆L:D conditions (Fig. 3, F1,17=105.37, p<0.0001), while males from the southern strain had higher survival than males from the northern strain (Fig. 3, mean south 11.6 (SEM 1.8) days, mean north 9.8 (SEM 1.1) days; F1,17=4.91, p<0.05). Survival rate differed between T-cycles in both ∆L:D and L:∆D experiments (Fig. 3A, B).
Mean longevity and Zeitgeber period
Figure 4 Mean longevity in different T-cycles per light-dark approach A) ∆L:D and B) L:∆D for males and females; northern line (blue) and southern line (orange). Mean longevity was calculated from survival data; bars represent standard error.
Mean longevity values were analysed in a single ANOVA model (Fig. 4), which shows that survival is affected by strain (p<0.007), sex (p<0.0001), protocol (∆L:D or L:∆D, p<0.0001), Zeitgeber period (T-cycle, p<0.015), and the interaction terms strain-protocol (p<0.011) and sex-protocol (p<0.002). Although Zeitgeber period duration influences survival in this analysis (Fig. 4), the resonance fitness hypothesis predicts a concave dependency of survival on T or the τ/T ratio (Fig. 1). We calculated τ/T ratios using circadian free-running periods measured in each line and sex under continuous darkness after LD entrainment. Free-running circadian periods were found to be: 26.3 h (± 1.6 h) for northern females, 24.8 h (± 1.0 h) for northern males, 25.0 h (± 0.3 h) for southern females, 24.2 h (± 0.3 h) for southern males (Table 2). A concave relationship between survival and T-cycle duration is not immediately clear from the data (Fig. 4) and requires statistical testing by adding τ/T and (τ/T)2 as linear terms in a sample size weighted regression model that includes strain, sex, protocol (∆L:D or L:∆D) and their mutual interaction terms. In a backward step wise regression, the interaction terms of τ/T and (τ/T)2 with protocol (p>0.29 and p>0.27) and strain (p>0.13 and p>0.15), were
found to be not significant and were therefore dropped from the model. Also, the interaction between strain and sex was not significant and dropped from the model (p>0.19). Hence, there was no statistical evidence in the data supporting a concave relationship between τ/T ratio and survival.
Activity patterns
To determine whether the wasp lines are able to entrain to the different T-cycles, locomotor activity was recorded and actograms were generated for T-20 h (LD 16:4), T-24 h (LD 16:8), and T-28 h (LD 16:12) (Fig. 5). Activity was recorded in 1-minute activity bins and double plotted in actograms. All experimental groups did entrain to all T-cycles. Consistent with data from Bertossa et al. (2013) the wasps showed high locomotor activity throughout the light phase and nearly no activity during the dark phase. Females were active during almost the entire light phase, whereas males were active during a smaller portion of the light phase. The northern line showed overall higher activity levels than the southern line. Although a rhythmic activity pattern is evident at T-20 h in females of both lines, there is also some activity separated from the main activity bout and some individuals showed a second activity peak (Fig. 5A & Fig. 6A). Overall, the wasps are clearly able to entrain to the different light cycles.
Figure 5 Actograms of individual wasps, females and males, from northern and southern lines in L:∆D
A) T-20 h (LD 16:4), B) T-24 h (LD 16:8) and C) T-28 h (LD 16:12). Activity was recorded in 1-minute activity bins and double plotted in actograms (each line in the figure represents 48h = 2 days, where the first line represents day 1 and 2, the second line day 2 and 3 and so forth). Animals appear to be active during the light phase. In T-20 h (LD 16:4) females showed two activity bouts, in T-24 h (LD 16:8) and T-28 h (LD 16:12) females and males, as well as in T-24 h (LD 16:8), expressed a unimodal activity pattern during the light phase. Females are active longer than males; northern animals have higher activity level than the southern ones.
The actogram data were analysed by generating activity profiles under the three different light conditions. 1-minute activity bins were averaged for each individual over five consecutive
days and individual average profiles were again averaged within each experimental group and condition (Fig. 6). Standard error of the mean is based on variation between individuals. As described above, females show locomotor activity throughout almost the entire light phase whereas males compress their activity in half of the light phase in all T-cycles.Overall, northern females show higher activity levels than southern females. In T-20 h, females from both lines exhibit an increase in activity with lights-on and a short dip in activity at the end of the first half of the light phase (Fig. 6). In the middle of the light phase, activity increased again and remained for the rest of the light phase. At T-24 h and T-28 h, northern females showed activity onset slightly after lights-on whereas southern females got active right away with lights-on. At T-20 h, activity levels of males were highest in the middle of the light phase whereas in T-24 h and T-28 h activity peaked in the first half of light phase. Furthermore, activity in northern males gradually increased after lights-on, peaked in the first half of the light phase and from there decrease gradually again. They showed a small second activity increase right before lights-off whereas in the southern line high activity started immediately after lights-on, followed by a further increase of activity near the end of the dark phase.
Figure 6 Activity profiles of females and males, from northern and southern lines in L:∆D A) T-20 h (LD
16:4), B) T-24 h (LD 16:8) and C) T-28 h (LD 16:12).For each experimental group and condition, activity data were averaged and normalized by its highest level. Wasps were predominantly active during the light phase; females were active during the entire light phase, males were more active only during part of the light phase. In general, the northern line has higher activity levels than the southern line.
Taken together, both in the actograms and in the activity profiles a coherent diurnal activity pattern indicates normal circadian entrainment under all conditions and strains (Fig. 5 & Fig. 6). The northern line showed higher activity levels than the southern line and females are more active than males. Females were active nearly the entire light phase, while males are mainly active in the first half of the light phase (in T-24 h and T-28 h) or in the middle of the light phase (in T-20 h). In both males and females (when considering the major activity bout) we find evidence for a phase-period rule (Pittendrigh and Daan 1976) with earlier phase angle of entrainment in longer T-cycles (Fig. 7). Phase was determined by taking the average centre of gravity over five cycles of stable entrainment (Fig. 6) within each individual. Phase angle of
entrainment was negatively correlated with T-cycle duration with different slopes for strain, but not for sex within a strain. Overall males had an earlier phase angle of entrainment than females (Fig. 7; for full statistics see figure text).
Figure 7 Phase-period relationship. Regression analysis showing that the slope of the phase-T-cycle
relationship differs between lines and sexes. The dependent variable is phase calculated as the individual centre of gravity (in hours after lights on) and the independent variables are T-cycle duration (T-cycle, 20, 24, 28h), strain (south=0, north=1), sex (female=0, male=1). Regression parameters include: constant (28.44), T-cycle (-0.88, p<0.0001), strain (-5.62, p=0.017), sex (-2.72, p<0.0001), T-cycle*strain (0.37, p=0.0002), strain*sex (-1.63, p=0.013). T-cycle*sex and T-cycle*sex*strain were found to be not significant and were dropped from the model (full model R2=0.85, F
5,80=93.68, p<0.0001).
The different levels of the phase-period regression analysis can be understood when we compare free-running circadian periods between lines and between males and females (Table 2). Indeed, in both lines males have shorter free-running periods than females, which would lead to a predicted earlier phase of entrainment. In addition, the southern line displays shorter free-running periods, which indeed would also lead to earlier phase angle of entrainment (Fig. 7).
Table 1 Free-running period and percentage of rhythmic animals of the northern and southern females and males; τ was determined after an entrainment period in LD 16:8 in constant darkness.
north south north south
n 16 15 13 9
τ ± SD 26.3 ± 1.6 25.0 ± 0.3 24.8 ± 1.0 24.2 ± 0.3
rhythmic % 100 100 69 100
The steeper slope of the phase-period relationship in the southern line, however, needs to be explained from differences in circadian phase resetting response to light. Oscillator theory would predict that stronger phase resetting would yield more shallow phase-T-cycle period relationships because a strong circadian light response would increase the range of entrainment. To answer the above relationship we evaluated circadian light responses by measuring a PRC in both lines and sexes.
Phase Response Curve
Phase response curves (PRCs) were measured to determine possible differences in circadian light response between the different lines. Light pulses of 8-h applied to different groups of animals at different times of day caused phase shifts of various magnitude, depending on the internal phase at which the pulses were applied. In both lines and both sexes the light pulses applied around ZT0, ZT24 respectively, caused large phase shifts of several hours (Fig. 8). The phase shifts were reduced towards the dead zone between ZT8 and ZT12, depending on line and sex. Because maximal phase shifts can reach up to 12-h advance or delay, these PRCs are also called type 0 or strong resetting PRCs. Although a similar shaped curve in all experimental groups is seen, individual variation is larger in the northern than in the southern line. Nevertheless, the large phase shifts indicate a complete phase resetting of the circadian system, with minor deviations from a straight line.
