sign in sign up
<<
Home , Circadian rhythm, Photoperiodism

Invertebrate , 3,155-164 (1997)

Insect circadian rhythms and

D. S. SAUNDERS Institute of , and Population , University of Edinburgh, West Mains Road, Edinburgh EH9 3JT

ABSTRACT Two -controlled processes, overt circadian rhythmicity and the photoperiodic induction of dia- pause, are described in the blow fly, Calliphora vicina and the fruit fly, melanogaster. Circadian locomotor rhythms of the adult flies reflect endogenous, self-sustained oscillations with a temperature compensated . The free-running rhythms become synchronised (entrained) to daily light:dark cycles, but become arrhythmic in constant light above a certain intensity. Some flies show fragmented rhythms (internal desynchronisation) suggesting that overt rhythmicity is the product of a multioscillator (multicellular) system. Photoperiodic induction of larval diapause in C. vicina and of ovarian diapause in D. melanogaster is also based on the circadian system but seems to involve a separate mechanism at both the molecular and neuronal levels. For both processes in both species, the compound and ocelli are neither essential nor necessary for photic entrainment, and the mechanism is not within the optic lobes. The central is the most likely site for both rhythm generation and extra-optic photoreception. In D. melanogaster, a group of lateral brain neurons has been identified as important circadian pacemaker cells, which are possibly also photo-sensitive. Similar lateral brain neurons, staining for arrestin, a protein in the phototransduction 'cas- cade' and a selective marker for photoreceptors in both vertebrates and invertebrates, have been identified in C. vicina. Much less is known about the cellular substrate of the photoperiodic mechanism, but this may involve the pars intercere- bralis region of the mid-brain.

KEY WORDS: Circadian; photoperiod; diapause; photoreception; measurement

Introduction of the within the central , and their regulation at the cellular level. Insects display a large number of behavioural, develop- mental and physiological events which are controlled by endogenous clock-like processes. Among the Circadian rhythms of locomotor activity 'higher' Diptera, for example, an array of daily behav- iours such as general locomotor activity, flight, mating, Rhythms of locomotor activity in Calliphora vicina are oviposition, egg hatch, pupariation and pupal eclosion easily recorded by placing a fly in a 9 cm petri dish are governed by circadian oscillators, whereas various with a supply of sugar and water, in such a way that seasonal phenomena such as the onset of diapause, or the moving fly breaks an infra-red light beam. These larval growth rates, are governed by photoperiodic events are registered by computer and presented as clocks (Saunders, 1982). This review is concerned conventional 'double-plotted' actograms (Fig. 1). with circadian rhythms and photoperiodism in the The actograms shown in Fig. 1 illustrate four cardi- adults of two species, the blow fly Calliphora vicina, and nal properties of a (for a -active the fruit fly . Both species show insect): (1) In continuous darkness (DD) and constant rhythms of locomotor activity whose properties are temperature (20~ an adult fly presents a noise-free well-known (e.g. Kenny and Saunders, 1991; Saunders and free-running rhythm of activity and rest which may et al., 1994) and whose regulatory mechanisms are persist for up to 7 (Fig. 1A). Although close to beginning to be unravelled (e.g. Hall, 1995). In addi- 24 h, the period (*) of the rhythm varies in individual tion, the blow fly overwinters in a larval diapause flies from about 21 to 25 h, with a mean value of about (Saunders, 1987), whereas D. melanogaster is known to 22.5 h (Cymborowski et al., 1993). (2) Step-wise possess a very shallow ovarian diapause (Saunders changes in temperature, either down or up, cause phase et al., 1989, 1990) which may aid overwintering sur- shifts of the rhythm, but otherwise leave ~ unaltered, vival in some populations (Saunders and Gilbert, attesting to its temperature-compensation (Fig. 1B). 1990). The two clock-like phenomena will be com- (3) When a fly is exposed to a of light and dark- pared, and questions asked about their similarities and ness (in this case LD 16:8) the activity rhythm differences, the photoreceptors involved, the location becomes entrained to an exact 24 h with locomotion

Corresponding author: D. S. Saunders. E-mail: [email protected] 15 6 Saunders largely restricted to the light (Fig. 1C). Lastly (4), Fewer than 5 per cent of flies show spontaneous transfer from DD to continuous 'bright' light (LL) leads 'splitting' of activity in DD (Kenny and Saunders, to behavioural arrhythmicity (Fig. 1D), although LL 1991) or when exposed to a light cycle of LD 1:23 below about 0.02 Wm "z causes z to lengthen (Hong with the light pulse coming on in the middle of the and Saunders, 1994). 'subjective night' (Hong and Saunders, in press)

B Time h 14.00 02.00 14.00 02.00 14.00 23 h ..... i, !k..:,• :" a ..... A lqme h Day 1 ii lit , ,,

12.00 24.00 12.00 24.00 12.00 il6M i '~ ~'"v i"ii '...... Iil I i~Ll!'l;lli!!i':~[ Day 1 ILt~IamiMi ~I 'W ~&k'i~'g'~" LI,d ~.,~ ..... IN ill a i~ fill ili~i

~o ~s m ~s a~ 7 il lii i F I~" Period {h} \: ,ilii 23.3 h [ 1 4 ll/lu i~. i Im~ /!13.3 h 7 = ~ I~* ..ilila, i a r ~--~- b ii i illiL J a n ,i ~- .I,; ..... lllll ,e,iu~ c 14: i i i ua e

II lit i ,liiJ Jill, blij /~ ~o ~s 2o as ao Penod {h} Period #'4 14|I t, ",i'u'~ I, i ~..~,~ '~ ,i.~lltU~jinl ~ ,.,.,u,kak~.,. , i~ it,. i~ild li-"~" .i :--""', 23.3 h 21 ~ .~ .....

t.*aka&L_

21 , ,,,h ,;....,~...~. ~. 28: Period (h)

C Time h ~me h 12.~ 24.~ 12.~ 24.~ 12.~ 22.7 h D ~5.oo 3.oo ~5.oo 3.00 15.00 Day 22.7 h Day a! r ~,,. wu, I" i~, a tiiJl ~Jilll 'i ~ J/lllli[ J|llJli Jnl a a AiJ~, hJ dmlJl I n dillil I ,,~/i I r e 7 JdUI it i=i ~ltlgL ,~lK~lki, iiM~ / Pe[iod (h) Jh/i t J ~ ~lIL t.,, ...... ,,,,,, ...... a 7 |kkMlmlJ hbJJ~d, ,l~.u .... J, a 10 15 2o 25 3c r [i~ail~, .L,~U...... ~, .a, Jv Penc

L a n

e

Period (h) o h ' --. a,.

_ ,id,,~ .~ , i,.,,,~ =, ' d J c

10 15 20 25 30 L .... dL~J .,~I t , ,,,a.,~t,. ,.-~ ,~, la|i P~nod (hl 23.3 h #hdkd v Period (h} r

Jl ~' i i~ c a i I~ ill , n .... iMt , ~l ..... c o ,iJa .~all~ 10 15 20 ~ 30 28 ~#i Penc

Fig. 1. CaUiphora vicina. Actograms showing circadian locomotor activity rhythms of single females. A - 'free running' rhythm in continuous darkness (DD), 20~ B - ditto, with step-wise temperature changes from 20~ to 15~ on day 8 (arrow) and from 15 ~ to 25~ on day 18 (arrow). C - fly initially in DD, exposed to an entraining cycle of LD 16:8 from days 8 to 14, and then back to DD. Arrow shows 'rebound' of activity shortly after light.off. D -fly initially in DD, then exposed to continuous bright light (0.024 Wm "2) from day 8 to 18, and then back to DD. Periodogram analyses for appropriate sections of the records are shown alongside each paneL Insect circadian rhythms and photoperiodism 15 7

(Fig. 2). Interpretation of these observations is that expressed rhythms with a period (,) close to 19 h, a long circadian rhythmicity is a multi-oscillator (and proba- period mutant (perL) a period close to 29 h, and a third bly multicellular) phenomenon. class of mutants (per~ which were apparently arrhyth- mic. This important and seminal study described the first 'clock' mutants in any organism and set the scene for the considerable subsequent successes in unravelling Time h the molecular genetics of the circadian system in Drosophila (see Hall, 1995, 1996). In our hands 12.00 24.00 12.00 24.00 12.00 (Saunders et al., 1994) the mean circadian period of pers Day ifllJfll !~trl[!rl[_ 1LI jr! i /,It/ _ ~_i/~L was found to be 19.70 + 0.57 h, per+ 23.72 + 0.90 h and perL2 29.10 + 3.64 h (in all cases, N = 50, mean + SD). i~ ili~lb i ,iiiili~ii pers was characterised by a shortened active phase, and i JiiliLL~ ~Jillll|,il a perL2 by a lengthened rest phase. Representative exam- lllli, ii ,~llil~ll ples of these mutant rhythms are shown in Fig. 3. ,~sllliill iJ il i Mutation at a second 'clock' locus, (tim) il il i li ,lllil, has been found to abolish circadian rhythmicity in 7i i.i ili,~, itl li!i i -- i D. melanogaster (Vosshall et al., 1994). The proposed ' hi ii~ , tii~ ill interactions between the PER and TIM proteins in the it, i li~i Q.! [~.I@| | generation of circadian rhythmicity is discussed later. ,, [,i Photoperiodic induction of larval diapause ,i t/C,t,u', i ~ ,,t ,Jill,i,/! b in Calliphora vicina 141 i . i illil,i~ ~ i ,, mill Lt I Adult females of C. vicina, entrained to daily light- dark cycles, use a photoperiodic 'clock' measuring ,,i Jll ,,_ ..... i,r,i daylength (or night length) to regulate the onset of an ,, , ..... , overwintering larval diapause. Thus flies exposed to I the long days (or short nights) of summer produce a I u F~ succession of continuously developing generations 21[l ~. li l[.i, .... ~a4 whilst days remain long, whereas those exposed to the il ,.,ia ., ~jJ,,i .... =i,l~ i, short days (or long nights) of autumn lay qualitatively i i C different eggs giving rise to larvae which burrow into the soil after completion of feeding, to pass the winter !,6 ..a..-.al ,I,, ,~'-~ ,, i i i in diapause (Vinogradova and Zinovjeva, 1972; il i Saunders et al., 1986). The maternal critical daylength i,l .~.L, i, l, . _,a.,,,.,,,, ,,,,[ 281 separating the diapausing and nondiapausing pathways h j is about 14.5 h/24 for a population of flies isolated near Edinburgh (Saunders, 1987) (Fig. 4), although the Fig. 2. C. vicina. Actogram of a female fly initially in DD and critical daylength is longer in flies from more northerly then exposed to LD 1:23 from day 8 to 21, and then back to DD. Light pulse commencing near middle of subjective night leads to an locations, and shorter in flies from the south 'internal desynchronization' into four components (I - 4). (McWatters and Saunders, 1996). Activity rhythms of the very much smaller D. A diapausing larva of C. vicina cannot be distin- melanogaster have been routinely recorded by a similar guished morphologically from a non-diapausing larva, method, but using a short length of glass tube rather but is recognised by its failure to pupariate. It is char- than a petri dish (e.g. Hamblen et al., 1986; Saunders et acterised by a low titre of circulating ecdysteroids al., 1994). Adult rhythms of a large number of wild type caused by the non-release of the cerebral neuropep- and mutant strains have now been studied. Over 25 tide, PTTH, and a developing inability of the ring ago, however, Konopka and Benzer (1971) iso- gland to respond to PTTH stimulation (Richard and lated three behavioural mutants (called period) with Saunders, 1987). In a population of diapause destined altered circadian clocks controlling both adult locomo- larvae maintained at 1 I~ ring gland refractoriness tor rhythmicity and pupal eclosion. Whereas wild type develops between days 26 and 30 post oviposition. All flies (Canton-S) showed free-running circadian periods larvae that have not pupariated by day 30 are therefore (in darkness) close to 24 h, a short period mutant (pers) regarded as being in diapause. 158 Saunders

, , i I.... , per+

A~22.8

22"9 ~~/j,g,,#/ iv{

B 12 24 36 t",allw48 utNru = t~ =.t* ~tu'~l.s ...... ', ..2,i ] pe

4 ! ~..~'"' t '" ' I 19.3 "' 6L, ...... ,,.. I ,' | l ' [ l 19-8\ '1 't ,/

~11~ L~-1~11 11 fEB Ptllc41T( I|l(11"~$

C )

[. .'"xe~,, a~, i _ ~..4u.: .... j

27 6_..-7 TM

1 -9 I II ii

(

xV.~t/ "@':2

PlJlO0 t|Nrill IN I|N KIINI| Tml(l~at5

Fig. 3. Drosophila melanogaster. Locomotor activity rhythms of period mutants with appropriate periodograms. A - wild-type, per +, B - short-period mutant, per s, C - long-period mutant, per L2.

Whether or not a larva produced by a short-day pausing offspring, and this process is affected by the tem- female enters diapause depends on a complicated perature of the maternal environment and the availabil- sequence of events, some acting on the maternal genera- ity of protein necessary for ovarian maturation tion and some on the larva itself. In addition to being (Saunders, 1987). A larva subsequently hatching from a exposed to short days, a fly has to register 10 or more diapause programmed egg will only enter diapause if the such cycles to be programmed for the production of dia- temperature of the larval environment, particularly Insect circadian rhythms and photoperiodism 159

Fig. 4. C. vicina. Induction 100 of larval diapause by mater- nally operating photoperiods. Critical daylength (CDL) is at 14.5/24 h. 8O

60 -~--14.5 h

c~ 40

~ c~

A 21f w w i ,d~ I I I I $ I I I ! | 8 12 16 20 24 Photoperiod, h/24

when the larva wanders from its food, is below about to long days or by transfer to a higher temperature. 15~ (Vaz Nunes and Saunders, 1989). In addition, Other strains, notably Oregon-R, were found to have a overcrowding of diapause destined larvae within their lower incidence of diapause under short days (Saunders food may lead to the smaller individuals side-stepping and Gilbert, 1990) whereas some with a tropical origin the diapause programme to form diminutive puparia and (e.g. Dahomey) showed ovarian maturation under all adults which may be capable of further reproduction photoperiods (Saunders, unpublished). The compara- despite the prevailing short days (Saunders, 1997). tively shallow nature of the response in D. melanogaster These important developmental 'decisions' almost cer- probably reflects the relatively recent evolutionary origin tainly provide flexibility to the onset of diapause, per- of diapause in this essentially commensal species. haps allowing a further nondiapause generation if Diapausing adults of D. melanogaster accumulated conditions allow. yolk polypeptides (YPs) in the haemolymph but very little in the ovary (Saunders et al., 1990). On the other hand, long day, nondiapausing females, and short day Photoperiodic induction of ovarian diapause females whose diapause had been broken by an upshift in Drosophila melanogaster of the temperature from 12 ~ to 25~ showed enhanced titres of YPs in the ovaries suggesting stimu- Although an ovarian or reproductive diapause is wide- lated uptake. In vitro studies of juvenile (JH) spread among Drosophila spp. (Lumme, 1978), synthesis by single excised corpora allata (CA) showed D. melanogaster was until recently regarded as day-neu- that glands from nondiapausing flies, or from flies in tral and without a diapause, probably overwintering which diapause had been broken, produced about four (where it needed to) in a state of quiescence. Some more JH than those from diapausing flies. strains of D. melanogaster, such as Canton-S, however, Diapause was also terminated under short days by a were shown to enter a shallow ovarian diapause when topical application of JH at a concentration of about newly emerged flies were transferred to short days at a 0.5 ~ug per fly. These data suggest that ovarian diapause relatively low temperature (< 14~ (Saunders et al., in D. melanogaster occurs as a result of a block to the 1989; Saunders and Gilbert, 1990). Females transferred JH stimulated uptake of YPs from the haemolymph, to long days (LD 16:8) at the same temperature com- caused by a reduced rate of synthesis of juvenile hor- menced a slow cycle of ovarian maturation. The short mones by the CA. Photoperiodic regulation of this day induced diapause was shown to persist for up to 6 response is probably mediated by brain regulation of weeks at 12~ but was terminated rapidly by exposure CA activity (Saunders et al., 1990). 160 Saunders

The circadian basis of photoperiodic time length measurement. measurement There are, therefore, a number of formal similari- ties between the regulation of overt rhythmicity and Although at first sight dissimilar, overt circadian photoperiodic time measurement: both have their base rhythms and photoperiodic time measurement are in the circadian system, both use light for entrainment functionally related, at least at the formal level of the or photoinduction and, in C. vicina and D. 'clock'. In a wide range of taxa, from flowering melanogaster, both operate simultaneously in the adult to (Follett and Follett, 1981), it has been female fly. Here, however, the similarities seem to end. shown that circadian rhythmicity provides the 'clock- In C. vicina, the free-running period for locomotor work' for day- or nightlength measurement, as postu- activity (Fig. 1) is about 22.5 h (Cymborowski et al., lated by Biinning (1936) over 60 years ago. 1993) whereas that for photoperiodic induction, as Experimental evidence for this functional relationship judged from the inter-peak intervals in Nanda- comes from photoperiodic experiments in which the Hamner experiments (Fig. 5), is closer to 24 h. In the organism is exposed to light-dark cycles whose overall former, the rhythm persists undamped for several period differs from that of the solar day. Most com- weeks, whereas for the latter there is evidence suggest- monly used is the so-called Nanda-Hamner experi- ing a rather heavily damping oscillator (Vaz Nunes mental design in which different groups of organisms et al., 1990). For D. melanogaster the evidence for the (flies in this case) are raised in cycles with a constant separate nature of overt rhythmicity and photoperiodic photophase coupled with variable dark periods such timing is more compelling, since the critical day that the total cycle length ranges over several multi- lengths for the period mutants (pers and per L2) are the ples of the presumed circadian period (for example, 18 same as that for wild type (Saunders, 1990), and flies to 84 h). Figure 5, for the Calliphora case, shows 'peaks' carrying per ~ or those devoid of the per gene (per') are of high diapause incidence at approximately 24 h apparently still capable of distinguishing short days intervals (24, 48 and 72 h) as the dark period is from long days, although with a shortened critical day extended, interspersed with 'troughs' of low diapause length (Fig. 6). It is concluded that the per gene and incidence at about 36 and 60 h (Saunders, 1992). A its product, although part of the feed-back loop gov- similar, but rather weaker result was obtained for erning overt circadian rhythmicity (see below) has lit- D. melanogaster (Saunders, 1990). It is concluded that tle to do with photoperiodism, the latter probably the circadian system is somehow involved in night being regulated by a separate system.

T 3T 2T 100 /,

8O

d 60 r~

~.. 4o

20 / ~ i 149 1 I 12 24 36 48 6O 72 84 Duration of light/dark cycle, T (h)

Fig. 5. C. vicina. Larval diapause induction in Nanda-Hamner photocycles (see text), each consisting of 13 h of light coupled with various h of darkness to give photocycles (T h) ranging from 18 to 80 h. r, 2r and 3r: peaks of high diapause incidence occuring at circadian inter. vals as darkness is extended. Insect circadian rhythms and photoperiodism 161

Cymborowski, 1996). Since bilateral lobectomy sepa- rates the compound eyes from the brain and, of course, A completely eliminates both optic lobes, this operation demonstrates that the compound eyes are not the essen- ...,'~176 801-f L tial photoreceptors, and that the optic lobes are similarly P not essential for rhythm generation or time measure- ",,< ment. This focuses attention on the central brain as a / likely site for both functions. of} 40 t~ Possible cellular substrates for extra-optic photore- O. ception in C. vicina were indicated by immunocyto- i5 20 chemical studies using an antibody raised against bovine S-antigen (= arrestin), a protein in the phototransduc- tion 'cascade' and a selective marker for ocular and 2 4 6 8 10 12 14 16 18 20 22 24 extra-ocular photoreceptors (Cymborowski and Korf, 1995). Arrestin-like immunoreactivity was found in reti- nal photoreceptors and in various groups of neurons B bilaterally distributed in the optic lobes and brain. Since

8O the eyes and optic lobes are not essential, this work focused attention on four groups of cells in the brain G) (J 6o some of which might be equivalent to the so-called lat- t~ \ eral neurons thought to have clock function in Drosophila (see below). On the other hand, expression of ~ 40 c-fos protein, which may be evidence for circadian regu- ~5 lation by photic stimuli, was observed in neurons of the 2O pars intercerebralis where photic stimulation of c-fos expression seemed to be associated with phase shifting of i 1 I a I I IT "~''O I I J per 2 4 6 8 10 12 14 16 18 20 22 24 the locomotor activity rhythm (Cymborowski and King, Photophase, h light/24 1996). In another series of experiments, injections of Fig. 6. D. melanogaster. Photoperiodic induction of adult (ovar. arrestin antibody directly into the brain of C. vicina were ian) diapause. A- photoperiodic response curve for Canton.S (wild found to reduce sensitivity to light in both entrainment type), pers (replicated) and per L2, showing essentially the same of the locomotor activity rhythm to light cycles, and to critical daylength in all strains (at about 14/24 h.). B- ditto for an arrhythmic mutant perOl , a 'double deletion of period (per') and the arrhythmicity generating effects of constant 'bright' an attached-X strain, XX; per~ , all showing shortened CDLs. light (Cymborowski et al., 1996). In D. metanogaster, the possible roles of the com- Photoreceptors, clock locations, and outputs pound eyes and optic lobes were investigated using a range of mutants including sine oculis (so) in which com- Circadian activity rhythms and the photoperiodic induc- pound eyes are lacking, small optic lobes (sol) in which the tion of diapause both require a photoreceptor and a medulla and lobula are reduced to about 50 per cent, and mechanism to measure the passage of time (a 'clock'). a double mutant sol;so in which the optic lobes are fur- What little is known about the anatomical locations of ther reduced to less than 5 per cent of normal (Helfrich these components, coupled with the brain-centred and Engelmann, 1983; Helftich, 1987). None of these endocrine responses outlined above, strongly indicates mutants showed an incidence of arrhythmic behaviour that higher neural centres are involved. In , in constant conditions greater than wild type, suggesting, for example, the compound eyes are the sole photorecep- as in C. vicina, that the compound eyes are not essential tors for rhythm entrainment, and the circadian clock(s) for photoreception, and that the optic lobes are not the are housed within the optic lobe(s) (Page, 1984). sole location for the circadian pacemaker. However, the In C. vicina, complete bilateral optic lobectomy, and mutant flies presented a greater incidence of'complex' or simultaneous removal of the ocelli, left the circadian 'split' rhythms, supporting the view that coupling locomotor activity rhythm intact and able to entrain to a between constituent (and possibly cellular) oscillators light-dark cycle (Cymborowski et at., 1994). Similarly had been weakened. lobectomised flies were also able to distinguish diapause- Immunocytochemical studies using a polyclonal anti- inducing short days from diapause-averting long days in body raised against the period gene product revealed the regulation of larval diapause (Saunders and staining in diverse tissues during several developmental 16 2 Saunders stages of D. melanogaster (Siwicki et a/., 1988), but most maker cells. The level of PER oscillates with a circa- significantly in certain lateral neurons in the brain dian (Hardin, 1994; Vosshall et al., 1994) which have been implicated as putative pacemaker cells. and is thought to regulate its own transcription nega- The same cells may also be photoreceptive (Wheeler et tively; step-wise post-translational phosphorylation of al., 1993; Vosshall and Young, 1995), and possibly equiv- PER, together with subsequent coupling with TIM alent to the arrestin-positive cells of C. vicina. PER pro- may provide a time delay in the loop which is essential tein is also expressed in glial cells (Ewer et al., 1992), for the system to oscillate (Edery et al., 1994; Vosshall particularly over the surface of the medulla. Helfrich- et al., 1994). The TIM/PER complex facilitates F6rster (1995, 1996) demonstrated two groups of PER- translocation of PER to the nucleus, and TIM appears containing lateral neurons lying between the lateral to be involved in the light induced resetting of the margin of the brain and the medulla, a dorsal clump oscillation because it is rapidly degraded when the fly (LNd) of about 3 to 7 cells, and a ventral clump (LNv) is exposed to light (Hunter-Ensor et al., 1996; Myers of about 4 to 10 cells. Using an antiserum raised against et al., 1996; Lee et al., 1996; Zeng et al., 1996). No the crustacean pigment dispersing hormone (PDH) she comparable data are available for C. vicina. showed that the LNv cells, which co-express PER and Computer models based on this molecular story PDH, had extensive arborization patterns over the (Goldbeter, 1995; Lewis et al., 1997) provide excellent medulla and further into the brain, possibly indicating simulations of the system and its properties, although details of the neural architecture of the fly's circadian certain elements such as temperature compensation and system, both in its input and output pathways in relation how different time delays are generated in the short and to the pacemaker cells. long period mutants are still obscure. The proposed mol- Flies with the brain structure mutation disconnected ecular model is well suited to describing circadian oscil- (disco), in which optic lobes are not connected to the lations at the cellular level, but is still too simple; work central brain, are generally arrhythmic (Dushay et al., now needs to move towards unravelling the complex 1989), although even a single LN remaining was suffi- multicellular neuronal architecture of the system and its cient to generate locomotor rhythmicity (Helfrich- internal cell-cell couplings (Helfrich-F6rster, 1996; F6rster and Engelmann, 1995). It was concluded that Meinertzhagen and Pyza, 1996). the ventral lateral neurons (LNvs) were the most Compared with overt rhythmicity in D. important cellular pacemakers, and that the fly uses melanogaster, the study of photoperiodic time measure- multiple photoreceptors. Normal flies may use their ment in any insect species is in its infancy. However, compound eyes but these were not essential; in their an essentially similar computer model based on a absence, photoreception in the brain (possibly LNvs) damping circadian oscillator (Lewis and Saunders, was sufficient. 1987; Saunders and Lewis, 1987a,b) provides excellent Next to nothing is known about the roles of eyes simulations of diapause induction, and of a wide vari- and optic lobes in the photoperiodic induction of ety of photoperiodic responses, across a range of insect ovarian diapause in D. melanogaster except that the examples. This suggests that the circadian 'clock' gov- mutants sine oculis, small optic lobes, disconnected and erning photoperiodic time measurement may rely on a minibrain all show a photoperiodic response curve with similar molecular feedback loop to that described for increased diapause induction under short days Drosophila activity rhythms. The two clock mecha- (Saunders, unpublished) suggesting that, as with loco- nisms (overt rhythmicity and photoperiodism), how- motor rhythmicity, brain-centred photoreceptor(s) ever, possess different formal properties suggesting that and clock are sufficient. they are operationally distinct. PER protein is expressed in the ovary, but does not enter the nucleus, or cycle (Hall, 1996) and therefore is presumably not Circadian rhythms and photoperiodism: part of a feedback loop. Anyway, in D. melanogaster feedback oscillations evidence suggests that per is not causally involved in photoperiodic time measurement (Saunders, 1990), Transcription of the per and tim genes and the actions and the ovary is merely seen as the target organ, with of the proteins they express are thought to be, accord- photoperiodic regulation being a brain-centred mecha- ing to current dogma, crucial elements in a negative nism controlling JH synthesis by the corpora allata feedback loop generating the circadian oscillation in (Saunders et al., 1990). Clearly there is need for mole- D. melanogaster (Hall, 1995; but see Hall, 1996). PER cular analysis of insect photoperiodic responses, protein and its mRNA are found in the nucleus and although the obvious organism of choice, D. cytoplasm of cells in the and, more significantly melanogaster, may not be ideal because of its extremely here, the lateral neurons which are regarded as pace- shallow diapause. Insect circadian rhythms and photoperiodism 163

References ence of lateral neurons in these flies. The Physiologist, 38, 6.2. Biinning, E. (1936) Die endogene Tagesrhythmik als Grundlage der Hong, S.-F. and Saunders, D. S. (1994) Effects of constant light on photoperiodischen Reaktion. Ber. Dtsch. Bot. Ges., 54, 590- the rhythm of adult locomotor activity in the blowfly, 607. Calliphora vicina. Physiol. Entomol., 19, 319-324. Cymborowski, B., Gillanders, S. W., Hong, S.-F. and Saunders, D. S. Hong, S.-F. and Saunders, D.S. (1998) Internal desynchronisation (1993) Phase shifts of the adult locomotor activity rhythm in in blow fly (CaIliphora vicina) locomotor activity rhythms: evi- Calliphora vicina induced by non-steroidal ecdysteroid agonist dence for a complex circadian pacemaker. Biol. Rhythm. Res. RH 5849. J. Comp. Physiol., 172, 101-108. (in press) Cymborowski, B., Lewis, R. D., Hong, S.-F. and Saunders, D. S. Hunter-Ensor, M., Ousley, A. and Sehgal, A. (1996) Regulation of (1994) Circadian locomotor activity rhythms and their the Drosophila protein timeless suggests a mechanism for reset- entrainment to light-dark cycles continue in flies (Calliphora ting the circadian clock by light. Cell, 84,677-685. vicina) surgically deprived of their optic lobes. J. Insect Physiol., Kenny, N. A. P. and Saunders, D. S. (1991) Adult locomotor rhyth- 40, 501-510. micity as 'hands' of the maternal photoperiodic clock regulat- Cymborowski, B., Hong, S.-F., McWatters, H. G. and Saunders, D. ing larval diapause in the blowfly, Calliphora vicina. J. Biol. S. (1996) S-antigen antibody partially blocks entrainment and Rhythms, 6, 217-233. the effects of constant light on the circadian rhythm of loco- Konopka, R.J. and Benzer, S. (1971) Clock mutants of Drosophila motor activity in the adult blow fly, Calliphora vicina. J. Biol. melanogaster. Proc. Natl. Acad. Sci. USA, 68, 2112-2116. Rhythms, 11, 68-74. Lee, C., Parikh, V., ltsukaichi, T., Bae, K. and Edery, I. (1996). Cymborowski, B. and King, V. (1996) Circadian regulation of Fos- Resetting the Drosophila clock by photic regulation of PER and like expression in the brain of the blowfly Calliphora vicina. a PER-TIM complex. Science, 271, 1740-1744. Comp. Biochem. Physiol., 115C, 239-246. Lewis, R. D. and Saunders, D. S. (1987) A damped circadian oscilla- Cymborowski, B. and Korf, H.-W. (1995) Immunocytological tor model of an insect photoperiodic clock. I. Description of demonstration of S-antigen (arrestin) in the brain of the blow the model based on a feedback control system. J. Theoret. Biol., fly CalIiphora vicina. Cell Tissue Res., 279, 109-114. 128, 47-59. Dushay, M. S., Rosbash, M. and Hall, J. C. (1989) The.disconnected Lewis, R. D., Warman, G. R. and Saunders, D. S. (1997) mutations in Drosophila melanogaster drastically Simulations of free-running rhythms, light entrainment and disrupt circadian rhythms. J. Biol. Rhythms, 4, 1-27. the light-pulse phase response curves for the locomotor activity Edery, I., Zwiebel, L. J., Dembinska, M. E. and Rosbash, M. (1994) rhythm in period mutants of Drosophila melanogaster. J. Theoret. Temporal phosphorylation of the Drosophila period protein. Biol., 185,503-510. Proc. Natl. Acad. Sci. USA, 91, 2260-2264. Lumme, J. (1978) Phenology and photoperiodic diapause in north- Ewer, J., Frisch, B., Hamblen-Coyle, M. J., Rosbash, M. and Hall, J. em populations of Drosophila. In Evolution of Insect Migration C. (1992) Expression of the period clock gene within different and Diapause, ed. H. Dingle, pp 145-170. Springer-Verlag. cell types in the brain of Drosophila adults and mosaic analysis McWatters, H. G. and Saunders, D. S. (1996) The influence of each of these cells' influence on circadian behavioral rhythms. J. parent and geographic origin on larval diapause in the blow fly, Neurosci., 12, 3321-3349. Calliphora vicina. J. Insect Physiol., 42, 72 t-726. Follett, B. K. and Follett, D. E. (1981) Biological Clocks in Seasonal Meinertzhagen, I. A. and Pyza, E. (1996) Daily rhythms in cells of Reproductive Cycles. John Wright and Sons Ltd., Bristol. the fly's optic lobe, taking time out from the circadian clock. Goldbeter, A. (1995) A model for circadian oscillations in the Trends Neurosci., 19, 285-291. Drosophila period protein (PER). Proc. Roy. Soc. Lond. B., 261, Myers, M. P., Wager-Smith, K., Rothenfluh-Hilfiker, A. and Young, 319-324. M. W. (1996) Light-induced degradation of TIMELESS and Hall, J. C. (1995) Tripping along the trail to the molecular mecha- entrainment of the Drosophila circadian clock. Science, 271, nisms of biological clocks. Trends Neurosci., 18, 230-240. 1736-1740. Hall, J. C. (1996) Are cycling gene products as internal Page, T. L. (1984) Neuronal organization of a circadian clock in the no longer the Zeitgeist of ? Neuron, 17,799-802. Leucophaea maderae. Photoperiodic regulation of insect Hamblen, M., Zehring, W. A., Kyriacou, C. P., Reddy, P., Yu, Q., and molluscan , Ciba Foundation Symposium, 104, Wheeler, D. A., Zwiebel, L. J., Konopka, R. J., Rosbash, M. 115-135. and Hall, J. C. (1986) Germ-line transformation involving Richard, D. S. and Saunders, D. S. (1987) Prothoracic gland func- DNA from the period locus in Drosophila melanogaster: Over- tion in diapause and non-diapause Sarcophaga argyrostoma and lapping genomic fragments that restore circadian and ultradian Calliphora vicina. J. Insect Physiol., 33,385-392. rhythmicity to per~ and per" mutants. J. Neurogen., 3,249-291. Saunders, D. S. (1982) Insect Clocks, 2nd edn. Oxford: Pergamon Hardin, P. E. (1994) Analysis of period mRNA cycling in Drosophila Press. head and body tissues indicates that body oscillators behave Saunders, D. S. (1987) Maternal influence on the incidence and differently from head oscillators. Mol. Cell. Biol., 14, 7211- duration of larval diapause in Calliphora vicina. Physiol. 7218. Entomol., 12,331-338. Helfrich, C. (1987) Use of Drosophila melanogaster brain mutants for Saunders, D. S. (1990) The circadian basis of ovarian diapause regu- the localization of the pacemaker of circadian locomotor activ- lation in Drosophila melanogaster: Is the period gene causally ity rhythms. J. Neurogen., 4, 137-140. involved in photoperiodic time measurement? J. Biol. Rhythms, Helfrich, C. and Engelmann, W. (1983) Circadian rhythm of the 5,315-331. locomotor activity in Drosophila melanogaster and its mutants Saunders, D. S. (1992) Insect seasonality: diapause, hormones and 'sine oculis' and 'small optic lobes'. Physiol. Entomol., 8, 257-272. the photoperiodic clock. In Circadian Clocks from Cell to Helfrich-F6rster, C. (1995) The period clock gene is expressed in Human, ed. T. Hiroshige and K. Honma, pp 21-38. Sapporo: central nervous system neurons which also produce a neu- Hokkaido University Press. ropeptide that reveals the projections of circadian pacemaker Saunders, D. S. (1997) Under-sized larvae from short-day adults of cells within the brain of Drosophila melanogaster. Proc. Natl. the blow fly, Calliphora vicina, side-step the diapause pro- Acad. Sci. USA, 92,612-616. gramme. Physiol. Entomol., 22,249-255. Helfrich-F6rster, C. (1996) Drosophila rhythms: from brain to behav- Saunders, D. S. and Cymborowski, B. (1996) Removal of optic lobes ior. Seminars in Cell & Dev. Biol., 7,791-802. of adult blow flies (Calliphora vicina) leaves photoperiodic Helfrich-F~Srster, C. and Engelmann, W. (1995) The occurrence of induction of larval diapause intact. J. Insect Physiol., 42, 807- clear rhythmic locomotor activity in individual disconnected 811. mutants of Drosophila melanogaster is accompanied by the pres- Saunders, D. S. and Gilbert, L. I. (1990) Regulation of ovarian dia- 164 Saunders

pause in the fruit fly Drosophila melanogaster by photoperiod at moderately low temperature. J. Insect Physiol., 36, 195-200. Saunders, D. S. and Lewis, R. D. (1987a) A damped circadian oscil- lator model of an insect photoperiodic clock. II. Simulations of the shapes of the photoperiodic response curves. J. Theoret. Biol., 128, 61-71. Saunders, D. S. and Lewis, R. D. (1987b) A damped circadian oscil- lator model of an insect photoperiodic clock. III. Circadian and hourglass responses. J. Theoret Biol., 128, 73-85. Saunders, D. S., Macpherson, J. N. and Cairncross, K. (1986) Matemal and larval effects of photoperiod on larval diapause in the flies CaIliphora vicina and Lucitia sericata. Exp. Biol., 46, 51- 58. Saunder, D. S., Henrich, V. C. and Gilbert, L. I. (1989) Induction of diapause in Drosophila melanogaster: photoperiodic regulation and the impact of arrhythmic clock mutations on time mea- surement. Proc. Natl. Acad. Sci. USA, 86, 3748-3752. Saunders, D. S., Richard, D. S., Applebaum, S. W., Ma, M. and Gildert, L. I. (1990) Photoperiodic diapause in Drosophila melanogaster involvves a block to the juvenile hormone regula- tion of ovarian maturation. Gen. Comp. Endocrin., 79, 174- 184. Saunders, D. S., Gillanders, S. W. and Lewis, R. D. (1994) Light- pulse phase response curves for the locomotor activity rhythm in period mutants of Drosophila melanogaster. J. Insect Physiol., 4O, 957-968. Siwicki, K. K., Eastman, C., Petersen, G., Rosbash, M. and Hall, J. C. (1988) Antibodies to the period gene product of Drosophila reveal diverse tissue distribution and rhythmic changes in the visual system. Neuron, 1,141-150. Vaz Nunes, M., Kenny, N. A. P. and Saunders, D. S. (1990) The photoperiodic clock in the blowfly, Calliphora vicina. J. Insect Physiol., 36, 61-67. Vaz Nunes, M. and Saunders, D. S. (1989) The effect of larval tem- perature and photoperiod on the incidence of larval diapause in the blowfly, Calliphora vicina. Physiol. Entomol., 14, 471-474. Vinogradova, E. B. and Zinovjeva, K. B. (1972) Maternal induction of larval diapause in the blowfly, Calliphora vicina. J. Insect Physiol., 18, 2401-2409. Vosshall, L. B., Price, J. L., Sehgel, A., Saez, L. and Young, M. W. (1994) Block in nuclear localization of period protein by a sec- ond clock mutation, timeless. Science, 263, 1606-1609. Vosshall, L. B. and Young, M. W. (1995) Circadian rhythms in Drosophila can be driven by period expression in a restricted group of central brain cells. Neuron, 15,345-360. Wheeler, D. A., Hamblen-Coyle, M. J., Dushay, M. S. and Hall, J. C. (1993) Behavior in light-dark cycles of Drosophila mutants that are arthythmic, blind or both. J. Biol. Rhythms, 8, 67-94. Zeng, H., Qian, Z., Myers, M. P. and Rosbash, M. (1996) A light entrainment mechanism for the Drosophila circadian clock. Nature, 380, 129-135. Accepted 19 November 1997