Characterization of the Molecular Clockwork in the Cockroach Rhyparobia Maderae

Characterization of the molecular clockwork in the cockroach Rhyparobia maderae

The core feedback loop genes period, timeless 1 and cryptochrome 2

Dissertation zur Erlangung des akademischen Grades Doktor der Naturwissenschaften (Dr. rer. nat.)

Vorgelegt im Fachbereich Mathematik und Naturwissenschaften der Universit¨at Kassel

von Achim Werckenthin

Kassel, Dezember 2013

Eidesstattliche Erkla¨rung

Hiermit versichere ich, dass ich die vorliegende Dissertation selbstständig, ohne unerlaubte Hilfe Dritter angefertigt und andere als die in der Dissertation angegebenen Hilfsmittel nicht benutzt habe. Alle Stellen, die wörtlich oder sinngemäß aus veröffentlichten oder unveröffentlichten Schriften entnommen sind, habe ich als solche kenntlich gemacht. Dritte waren an der inhaltlich- materiellen Erstellung der Dissertation nicht beteiligt; insbesondere habe ich hierfür nicht die Hilfe eines Promotionsberaters in Anspruch genommen. Kein Teil dieser Arbeit ist in einem anderen Promotions- oder Habilitationsverfahren verwendet worden.

Kassel, 13.12.2013

Achim Werckenthin

iii iv

Vom Fachbereich 10, Mathematik und Naturwissenschaften der Universitat¨ Kassel als Dissertation am 13.12.2013 angenommen.

Prufungskommission¨

1. Gutachterin: Prof. Dr. Monika Stengl 2. Gutachterin: Prof. Dr. Mireille Schafer¨ 3. Gutachterin: Prof. Dr. Charlotte Forster¨ 4. Gutacher: Prof. Dr. Friedrich Herberg Tag der mundlichen¨ Prufung:¨ 17.02.2014 Contents

Contribution statements ix

Zusammenfassung xi

1 Summary 1

2 Introduction 3 2.1 Rhyparobia maderae andcircadianresearch ...... 5 2.1.1 Thecentralnervoussystem...... 6 2.2 Molecularrhythmgeneration ...... 11 2.2.1 Transcriptional-translational feedback regulation and adjacent pathways in D. melanogaster ...... 12 2.2.2 The circadian molecular clockworkinnon-Drosophilidinsects ...... 16 2.2.3 Period...... 20 2.2.4 Timeless ...... 22 2.2.5 Cryptochrome ...... 23 2.2.6 Posttranslationalregulation...... 26 2.3 Thecellularbasisforcircadianrhythmicity ...... 27 2.3.1 Clock neurons in Drosophila melanogaster ...... 27 2.3.2 The circadian pacemaker controlling locomotor rhythms in R. maderae . 29 2.3.3 Circadian oscillators outside the clock neurons and in peripheral tissues . 31 2.4 Photoperiodism ...... 33 2.4.1 Photoperiodandcircadianrhythm ...... 34 2.4.2 Circadianclockgenesandphotoperiodism ...... 36 2.4.3 Photoperiodismincockroaches ...... 37 2.5 Aimsofthisstudy...... 39

3 Materials and methods 41 3.1 Animalrearing ...... 41 3.2 Molecularcloning...... 41 3.2.1 RNAextraction...... 42 3.2.2 cDNAsynthesisforPCRusingdegenerateprimerpairs ...... 42 3.2.3 Poly-A+ puriﬁcation ...... 43 3.2.4 cDNAsynthesis...... 44 3.2.5 Adaptorligation ...... 47 3.2.6 Primers ...... 48 3.2.7 InitialPCR ...... 49 3.2.8 RACE-PCR...... 50 3.2.9 Electrophoresis ...... 51

v vi CONTENTS

3.2.10 Gelextraction...... 51 3.2.11 Cloning...... 54 3.2.12 Sequencingandsequenceanalysis ...... 58 3.2.13 Motifprediction ...... 58 3.2.14 Phylogeneticanalysis...... 58 3.2.15 Solutions ...... 59 3.3 Quantitativerealtimepolymerasechainreaction...... 62 3.3.1 Tissuesampling...... 62 3.3.2 RNAextraction...... 62 3.3.3 cDNAsynthesis...... 64 3.3.4 qPCR...... 65 3.3.5 Dataanalysis ...... 67 3.4 Antibodygeneration ...... 70 3.5 Westernblotting...... 72 3.5.1 Tissuesamplingandhomogenization ...... 72 3.5.2 ProteinassayPierce660nm ...... 72 3.5.3 Polyacrylamidegelelectrophoresis ...... 73 3.5.4 Blotting...... 76 3.5.5 Detection ...... 78 3.5.6 Preadsorptionoftherabbitanti-rmPERantibody ...... 79 3.5.7 Stripping ...... 79 3.5.8 Solutions ...... 81 3.6 Immunohistochemistry ...... 83 3.6.1 Primaryantibodies ...... 83 3.6.2 Secondaryantibodies...... 83 3.6.3 Sections...... 84 3.6.4 Whole-mounts ...... 86 3.6.5 Liquidphasepreadsorption...... 90 3.6.6 Imaging...... 91 3.6.7 Imageprocessing...... 92 3.6.8 Quantiﬁcation...... 92 3.6.9 Solutions ...... 93 3.7 Runningwheelassays...... 94

4 Results 95 4.1 Molecularcloning...... 95 4.1.1 Period...... 95 4.1.2 Timeless1 ...... 99 4.1.3 Cryptochrome2 ...... 99 4.2 Putativecorefeedbackloopgeneexpression ...... 102 4.2.1 Putative core feedback loop gene expression in equinox light-dark cycles 102 4.2.2 Putative core feedback loop gene expression in constant darkness . . . . 105 4.2.3 Putative core feedback loop gene expression in different photoperiods . . 106 CONTENTS vii

4.3 Circadianoscillationsinproteinabundance ...... 111 4.3.1 PERIOD ...... 111 4.3.2 TIMELESS1...... 112 4.3.3 CRYPTOCHROME2 ...... 113 4.4 AntirmPER-immunoreactivityinthenervoussystem ...... 115 4.4.1 AntirmPER-immunoreactivityinthe central nervoussystem ...... 115 4.4.2 AntirmPER-immunoreactivityintheventralnervecord ...... 120 4.5 Runningwheelactivityindifferentphotoperiods ...... 123 4.5.1 Lowlightactivity...... 123 4.5.2 Brightlightactivity...... 124

5 Discussion 127 5.1 Core feedback loop genes in R. maderae ...... 127 5.1.1 Period...... 127 5.1.2 Cryptochrome ...... 128 5.1.3 Timeless1 ...... 129 5.2 Corefeedbackloopgeneexpressionrhythms ...... 129 5.2.1 DailyexpressionrhythmsinLDcycles ...... 129 5.2.2 DailyexpressionrhythmsinDD ...... 130 5.2.3 CorefeedbackloopgeneexpressionintheAMe ...... 130 5.2.4 CorefeedbackloopgeneexpressioninMalpighiantubules ...... 131 5.2.5 The total expression level compared to actin ...... 131 5.2.6 Photoperioddependentplasticity...... 132 5.3 Circadianrhythmsonproteinlevel ...... 133 5.3.1 PERIOD ...... 133 5.3.2 TIMELESS1...... 135 5.4 DistributionofPER-ircellsinthecentralnervoussystem ...... 136 5.4.1 PER-irindifferentZeitgebertimes...... 137 5.5 Locomotoractivityindifferentphotoperiods ...... 138 5.6 Conclusionandfutureperspectives...... 140

6 Acknowledgements 169

A Appendix 171 viii CONTENTS Contribution statements

The contributions of the author for each part of this thesis will be stated clearly as follows.

Section 3.1: Molecular cloning • Designed the experiments. • Conducted the experiments. Generation of the adaptor-ligated RACE cDNA, initial cloning and RACE-PCR of rmPer, rmTim1 and rmCry2 was done partly in collaboration with Dr. Christian Derst at the Institute for Integrative Neuroanatomy, Charit´eBerlin. The author conducted the majority of the experiments. The initial primers that were used for cloning of rmPer were based on a rmPer sequence previously obtained by Julia Fischer and Jeff Hall. • Sequence and phylogenetic analysis. • Wrote the manuscript. • The results of this section are publishedin: WERCKENTHIN, A., DERST, C. & STENGL, M. Sequence and Expression of per, tim1, and cry2 genes in the Madeira cockroach Rhy- parobia maderae. Journal of Biological Rhythms 27(2012), 453–466. • Figures 4.1, 4.2, 4.4 and 4.5 also appeared in a similar form in the publication cited above, but were modiﬁed and re-rendered for this thesis.

Section 3.2: Putative core feedback loop gene expression • Designed the experiments. • Conducted the experiments. Quantitative PCR was done in collaboration with Dr. Chris- tian Derst at the Institute for Integrative Neuroanatomy, Charit´eBerlin. The author conducted the majority of the experiments. Experiments for subsection 4.2.2 and constant darkness experiments of subsection 4.2.3 were conducted by the author at the University of Kassel and the Max-Planck-Institute for Terrestrial Microbiology in Marburg. • Analyzed the data. • Wrote the manuscript. • The results of this section are published in: WERCKENTHIN,A., DERST, C. & STENGL, M. Sequence and Expression of per, tim1, and cry2 genes in the Madeira cockroach Rhy- parobia maderae. Journal of Biological Rhythms 27(2012), 453–466. • Figure 4.21 also appeared in the publication cited above.

Section 3.3: Circadian oscillations in protein abundance • Designed and conducted all experiments.

ix x CONTRIBUTION STATEMENTS

Section 3.4: Anti rmPER-immunoreactivity in the nervous system • Designed the experiments. • Conducted the experiments except the preadsorption experiments of the anti rmPER (rabbit) antibody, which were conducted by Andreas Arendt. The majority of the immuno- staining experiments of subsection 4.4.1 were conducted by Azar Massah, under guidance of the author. • Imaging of all dissections, except dissections of the anti rmPER (rabbit) antibody preadsorption experiments, which were scanned by Andreas Arendt. • Analyzed the data.

Section 3.5: Running wheel activity in different photoperiods • Designed, conducted and analyzed all experiments.

Unless stated otherwise, all parts of this thesis were implemented by the author. Zusammenfassung

Einleitung

Im Laufe der Evolution haben Organismen die Fähigkeit entwickelt, sich an rhythmisch wieder- kehrende Veränderungen des Lebensraumes anzupassen. Die Fähigkeit, derartige periodische Umweltveränderungen vorauszusehen, ist für das Uberleben¨ eines Organismus entscheidend. Einer der prominentesten Rhythmen ist der durch die Rotation der Erde bedingte Hell-Dunkel- Wechsel. Bereits im 18ten Jahrhundert konnte Jacques d’Ortous de Mairan in einem der ersten chronobiologischen Experimente zeigen, dass die Pflanze Mimosa pudica selbst in voll- ständiger Dunkelheit in einem täglichen Rhythmus ihre Blätter faltet und somit ein endogener Schrittmacher vorhanden sein muss [56]. In den 50er und 60er Jahren des letzten Jahrhun- derts wurde der Grundstein der modernen chronobiologischen Forschung schließlich von Jürgen Aschoff (Menschen) [7], Erwin Bünning (Pflanzen) [28] und Colin S. Pittendrigh (Insekten) [241] gelegt. Sie konnten zeigen, dass die täglichen Rhythmen eines Organismus auf endogenen Rhythmen mit einer Periodenlänge von etwa 24 Stunden basieren (folglich ”circadiane Rhyth- men” genannt) und dass diese endogenen Rhythmen an externe Umweltrhythmen ankoppeln können. Unter den Insekten zählten Schaben zu den ersten Organismen, bei denen die innere Uhr untersucht wurde. Bereits seit den 1960ern ist Rhyparobia maderae (Synonym: Leucophaea maderae) Gegenstand der cicadianen Forschung und es war diese Art, bei der erstmals der circadiane Schrittmacher lokalisiert werden konnte [209]. Die Forschung zu den molekularen Grundlagen der inneren Uhr begann ebenfalls mit einem Insekt, nämlich mit der Fruchtfliege Drosophila melanogaster. Um 1970 konnten Konopka und Benzer erstmals zeigen, dass ein einzelnes Gen das Verhalten stark beeinflusste. Mittels chemischer Mutagenese wurden Mu- tanten erzeugt und dann isoliert, die Anderungen¨ in ihrem circadianen Rhythmus zeigten: perl (Periodenlänge größer als 24 Stunden), pers (Periodenlänge kürzer als 24 Stunden) und per01 (arhythmisch) [145]. Nach period wurden etliche weitere ”circadianen Gene” entdeckt, die an der Erzeugung circadianer Rhythmen direkt beteiligt sind. Diese Rhythmen werden durch soge- nannte ”Rückkopplungs-Schleifen” (feedback loops) erzeugt. Hierbei inhibiert das Protein eines Genes seine eigene Transkription, so dass sowohl Transkript- als auch Protein-Konzentration rhythmisch schwanken. Per, Timeless 1 (Tim1), Clock (Clk) und Cycle (Cyc) bilden die Kompo- nenten für den ersten beschriebenen feedback loop, den sogenannten ”core feedback loop”. CLK und CYC heterodimerisieren, binden an einen enhancer und aktivieren die Transkription von per und tim1. PER und TIM1 dimerisieren dann im Cytoplasma, wandern in den Zellkern und inhi- bieren die CLK:CYC vermittelte Transkription. Verschiedene Proteinkinasen und Phosphatasen regulieren dabei Abbau, Kernwanderung und Interaktion der beteiligten Proteine. Die dadurch zu Stande kommende Oszillation der Genprodukte mit einer Periodenlänge von etwa 24 Stunden bildet die molekulare Grundlage der inneren Uhr. Die Ankopplung an den Hell-Dunkel-Wechsel

xi xii ZUSAMMENFASSUNG

der Umwelt wird bei D. melanogaster durch das photosensitive Protein CRYPTOCHROME 1 vermittelt, jedoch sind auch die Komplexaugen und extraretinale Photorezeptoren an der Vermit- tlung des Lichteinganges beteiligt. Zusätzlich zu den am core feedback loop beteiligten per und tim1, aktiviert das CLK:CYC Heterodimer auch die Expression von drei weiteren Genen, die ebenso einen feedback loop, den sogenannten ”clk loop” bilden: vrille (vri), par domain protein 1 (pdp1) und clockwork orange (cwo). Während VRI die clk Expression inhibiert, aktiviert PDP1 dessen Expression etwa vier Stunden später. CWO inhibiert ebenso die CLK:CYC vermittelte Inhibition, scheint jedoch eine komplexere Rolle bei der Regulation circadianer Expression zu besitzen. Der core feedback loop läuft allerdings auch ohne den clk loop, jedoch nicht umgekehrt. Vri, pdp1 und cwo scheinen vor allem für die Regulation von Amplitude, Phasenlage und Aus- gang des core feedback loops wichtig zu sein (Review: [3, 92]). Die grundsätzlichen Mechanismen und beteiligten Gene unterscheiden sich zwar relativ wenig zwischen den einzelnen Insektengruppen, doch gibt es im Detail teilweise erhebliche Unter- schiede. Der auffälligste Unterschied betrifft das Vorhandensein oder die Abwesenheit von tim1 und cry1 beziehungsweise cry2. Uberraschenderweise¨ exprimieren Insekten ein weiteres ”mammalian like” cry, wie zuerst bei dem Monarchfalter Danaus plexippus gezeigt wurde [367]. Mit der Sequenzierung des Genoms der Honigbiene Apis mellifera und der anschließenden Unter- suchung der circadianen Gene [269] wurde schließlich klar, dass verschiedene Insekten unterschiedliche Typen von molekularen Uhrwerken besitzen. Im Fall von A. mellifera fehlen sowohl tim1 als auch cry1 und CRY2 wirkt als Transkriptionsrepressor, ählich wie im circadianen Sys- tem der Säugetiere. Bei anderen Arten, beispielsweise D. plexippus und anderen Lepidopteren, sind sowohl tim1 als auch cry1 und cry2 vorhanden. Weitere Unterschiede betreffen den Aufbau und die Funktion einzelner Gene. Ursprünglich scheinen Insekten sowohl tim1 als auch cry1 und cry2 besessen zu haben. Verschiedene Insektenarten scheinen dann sekundär einzelne Gene ver- loren oder dupliziert zu haben [359, 366, 367]. Während das molekulare, circadiane System bei D. melanogaster sehr gut untersucht ist, fehlen bei vielen anderen Insektengruppen immernoch umfassende Untersuchungen. Zwar sind inzwischen mehrere Spezies vollständig sequenziert, doch vor allem hemimetabole Arten sind bis heute kaum untersucht worden. Obwohl nach der Sequenzierung des Genomes der Blattlaus Acyrthosiphon pisum eine Untersuchung der circadianen Gene dieser Art veröffentlicht wurde, stellte sich heraus, dass das circadiane System dieser Art wiederum eine Ausnahme bildet und nicht typisch für hemimetabole Insekten zu sein scheint [46]. Bis zum heutigen Tage ist kein Genom einer Schabenart veröffentlich worden. Lediglich Blattella germanica per [166], und Teile von per und tim1 von Periplaneta americana wurden kloniert. Die zellulären Grundlagen der inneren Uhr sind ebenfalls bei D. melanogaster am besten charakterisiert. Etwa 75 Zellen je Gehirnhemisphäre exprimieren die circadianen Gene und sind für das circadiane Laufverhalten wichtig. Die Zellen wurden vor allem nach ihrer Lage kategorisiert: Zwei Ubergruppen¨ finden sich im dorsalen (dorsal neurons mit 3 Untergruppen) und lateralen (lateral neurons mit 4 Untergruppen) Protocerebrum des Gehirns. Auch funktionell scheinen sich die Uhrneurone in mehrere Gruppen unterteilen zu lassen. D. melanogaster zeigt ein typ- isches bimodales Aktivitätsmuster, mit Aktivität gegen Morgen und Abend, und entsprechend scheinen zwei unterschiedliche Zellgruppen jeweils eine der Aktivitätsphasen zu steuern. Diese so genannten morning- und evening-cells stimmen allerdings nicht mit jeweils einer der anatomis- xiii chen Gruppen überein (Review in: [355]). Bei R. maderae wurden die circadianen Schritt- macherzellen die die Laufaktivität steuern ebenfalls im Gehirn lokalisiert. Mittels Läsions und Transplantationsexperimenten konnte gezeigt werden, dass ein kleines Neuropil im optischen Lobus, die akzessorische Medulla (AMe) mit assoziierten Zellen, die circadiane Laufaktivität steuert. Die einzelnen Zellgruppen konnten anhand ihrer Neuropeptid-Expression näher charakterisiert werden und es stellte sich heraus, dass Peptid-basierte Neuromodulation eine wichtige Rolle bei der Funktion der AMe und damit des Schrittmacherzentrums der Aktivitätsrhythmen spielt. Es ist bei R. maderae im Gegensatz zu D. melanogaster aber nur wenig darüber bekannt, welche neuronalen Zellgruppen die jeweiligen circadianen Gene exprimieren. Die Expression der circadianen Gene beschränkt sich nicht nur auf das Nervensystem. Verschiedene neuronale und nicht-neuronale Zellen und Gewebe, so beispielsweise die Malpighischen Gefäße, exprimieren Uhrgene und unterhalten somit jeweils eigene Schrittmacher. Je nach Gewebe und artspez- ifisch koppeln diese entweder an den neuronalen Schrittmacher im Gehirn an oder unterhalten eigenständige Rhythmen. Zusätzlich zum täglichen Hell-Dunkel-Wechsel, der durch die Rotation der Erde um ihre Achse hervorgerufen wird, bewirkt die Rotation der Erde um die Sonne auf Grund der Neigung der Erdachse jahreszeitliche Anderungen.¨ Uber¨ das Jahr und abhängig vom Breitengrad ändert sich so die Tageslichtlänge, aber auch andere Umweltfaktoren, wie Temperatur und Nieder- schlag. Alleine die Anderung¨ der Tageslichtlänge hat einen Einfluss auf das Verhalten und die Physiologie vieler Insekten. Zum einen muss sich das circadiane System an die sich kontinuier- lich ändernde Tageslichtlänge anpassen können, um die tägliche Rhythmik aufrecht erhalten zu können. Zum anderen gibt es bestimmte Verhaltensweisen und physiologische Anderungen,¨ mit denen sich Insekten an die Jahreszeiten anpassen. Diese werden vor allem durch die sich ändernde Photoperiode bestimmt. Beispielsweise stoppt die Entwicklung vieler Insekten in be- stimmten Phasen abhängig von der Tageslichtlänge, was als Diapause bezeichnet wird [291]. Obwohl immer noch kontrovers diskutiert wird ob und in wie weit das circadiane System im Photoperiodismus von Insekten eine Rolle spielt, scheint es zumindest einige Arten zu geben, bei denen beide Systeme miteinander verknüpft sind [120, 311].

Fragestellung und Ziele dieser Arbeit

Die circadiane, Laufrhythmen kontrollierende Uhr ist bei R. maderae und anderen Schaben sowohl auf Verhaltens- als auch auf zellulärer Ebene gut charakterisiert. Fast nichts ist jedoch über die molekulare, circadiane Uhr dieser Art bekannt. Um mit der Analyse des molekularen Uhrwerkes von R. maderae zu beginnen, wurden circadiane Gene kloniert und analysiert. Die Fragestellung, ob diese Gene bei R. maderae vorhanden sind, ihre Struktur und ihr Expressions- muster sollte bearbeitet werden. Die core feedback loop Gene per, tim1 und cry2 wurden kloniert und ihre Expression mit Hilfe von quantitativer PCR analysiert. Mittels Zeitreihen von isolierten Gehirnen, AMae und nicht neuronalem Gewebe (Malpighische Gefäße) wurde untersucht, wie sich die Expression der identifizierten Uhrgene im Tagesverlauf ändert. Um herauszufinden, ob die Anpassung der Uhrgen-Expression an unterschiedliche Photoperioden gelingt, wurden im Kurz- oder Langtag aufgewachsene Tiere untersucht. Auf Grundlage der gewonnenen Sequenz- xiv ZUSAMMENFASSUNG

information wurden gegen R. maderae PER, TIM1 und CRY2 gerichtete Antikörper generiert um die circadiane Oszillation der Proteinkonzentration und das zelluläre Expressionsmuster im Gewebe zu untersuchen. Zusätzlich wurden Laufradexperimente dazu genutzt, den Einfluss der in den Expressionsstudien genutzten Photoperioden auf die endogene Periodenlänge zu bestim- men.

Ergebnisse und Diskussion

Die core feedback loop Gene per, tim1 und cry2 der Madeira-Schabe Mittels degenerierten Primern und rapid amplification of cDNA ends (RACE) konnte der voll- ständige offene Leserahmen (OLR) der per und cry2 cDNA mit Teilen der flankierenden, untranslatierten Bereiche und ein Teil des tim1 OLR kloniert werden. Die cDNA des klonierten und sequenzierten R. maderae per Orthologs (rmPer) ist 4261 Basenpaare (bp) lang und umfasst ein Start-Codon an Position 133-135, den vollständigen OLR, ein Stop-Codon an Position 3961- 3963 und einen Teil der 3’ und 5’ untranslatierten Region. Das von dem OLR kodierte Protein hat eine voraussichtliche Länge von 1276 Aminsäuren (AS) und ein errechnetes Molekulargewicht von 138,7kDa. Eine phylogenetische Analyse gruppierte rmPER gemeinsam mit weiteren PER Proteinen anderer Schabenarten und innerhalb weiterer hemimetaboler Insekten. Die meisten funktionellen Domänen wie sie für D. melanogaster PER beschrieben wurden, sind bei rmPER konserviert. Dies deutet darauf hin, dass rmPER eine ähnlich Funktion in der inneren Uhr hat wie bei D. melanogaster und den meisten anderen Insekten. Anstelle des TG-repeats findet sich bei rmPER ein kurzer repeat von Serin-Glycin Resten von unbekannter Funktion. Zusätzlich findet sich eine c7-Domäne bei rmPER. Diese wurde zuerst bei B. germanica beschrieben und danach bei vielen weiteren Insektenarten gefunden. Obwohl die Funktion der c7-Domäne unbekannt ist, liegt die Vermutung nahe, dass sie ähnlich wie eine ebenso Carboxy-terminale (C-terminale) Re- gion bei Vertebraten, Interaktionen von PER mit CRY2 vermittelt. Eine phylogenetische Analyse gruppierte rmPER gemeinsam mit Schaben PER Proteinen innerhalb der PER Proteine weiterer hemimetaboler Insekten. Die cDNA des klonierten und sequenzierten R. maderae cry2 Orthologs (rmCry2) ist 2221 bp lang und umfasst ein Start-Codon an Position 117-119, ein Stop-Codon an Position 1875-1877, den vollständigen OLR und einen Teil der 3’ und 5’ untranslatierten Region. Es handelt sich um das erste bei Schaben bekannte cry Gen. Das vom OLR codierte Protein hat eine voraussichtliche Länge von 586 AS mit einem errechneten Molekulargewicht von 67,3kDa. Funk- tionelle Domänen, wie sie in Mäusen gezeigt für die CLOCK:BMAL1 (BMAL, Brain and Mus- cle ARNT-like 1, ist das Säugetier Homolog von CYC) Interaktionen notwendig sind, sind in rmCRY2 konserviert. Dies deutet darauf hin, dass rmCRY2, ebenso wie in anderen Insekten und Säugetieren gezeigt, als Transkriptionsrepressor innerhalb des core feedback loops wirkt. Eine phylogenetische Analyse gruppierte rmCRY2 gemeinsam mit CRY2 anderer hemimetaboler In- sekten innerhalb der CRY2 Proteine. Das Fragment des R. maderae tim1 Orthologs (rmTim1), das kloniert werden konnte, umfasst einen 591bp langen Teil des OLR von rmTim1. Ein alignment mit D. melanogaster Tim1 zeigte, xv

dass dieses Stück für einen Teil des Proteins C-Terminal der PER-1 Bindedomäne und einen Teil der PER-2 Bindedomäne kodiert. Die erhaltene Sequenzinformation war ausreichend für das Er- stellen spezifischer quantitativer Polymerase-Kettenreaktion (qPCR) Primer und zur Erzeugung eines Antikörpers.

Expression von rmPer, rmTim1 und rmCry2 in verschiedenen Lichtrhyth- men

Mittels qPCR wurde als nächstes untersucht, ob die Expression der zuvor klonierten Gene circadian schwankt. Zunächst wurden neuronales Gewebe (AMe und Gehirngewebe ohne AMe) und Malpighische Gefäße von Tieren aus 12h Licht, 12h Dunkelheit Zyklen (LD12:12) untersucht. Die mRNA Level von rmPer, rmTim1 und rmCry2 schwanken in Gehirn und isolierter AMe-Region im Tagesverlauf mit Maxima in der ersten Hälfte der Nacht. Dieses Ergebnis ist vergleichbar mit den Resultaten bei D. melanogaster [93, 302], D. plexippus [366] und A. mellifera [269] und deutet darauf hin, dass die Expression von denselben Elementen gesteuert wird, wie bei diesen Insekten. Da die Expression auch in Gewebeproben von Gehirnen ohne den bekannten circadianen Schrittmacher der AMe und ihrer assoziierten Neurone schwankte, muss es weitere Schrittmacher im Gehirn von R. maderae geben. Mögliche Kandidaten hierfür sind die PDF-immunreaktiven Neurone der Lamina. Die Amplitude der Oszillation von rmPer, rmCry2 (<2-fach) und rmTim1 (∼2,5-fach) war ähnlich wie bei anderen bisher untersuchten hemimetabolen Insekten wie Pyrrhocoris apterus [105, 319], Riptortus pedestris [117] und A. pisum [46] im Vergleich zu vielen holometabolen Insekten (bis zu 7-fach bei A. mellifera cry2 [269]) vergleichsweise niedrig. Eventuell ergibt sich die niedrige Amplitude aus verschiedenen, nicht phasengleichen Schrittmachern innerhalb der untersuchten Gewebe (beispielsweise Glia und Neurone). Im Gegensatz zu D. melanogaster [81], und vergleichbar mit Gryllus bimaculatus [336], konnten in den Malpighischen Gefäßen von R. maderae für keines der drei untersuchten Uhrgene Unterschiede im Expressionslevel zwischen den untersuchten Zeitpunkten festgestellt werden. Dies deutet darauf hin, dass bei G. bimaculatus und R. maderae diese Uhrgene eine gewebespezifische Rolle übernehmen. Am ersten Tag konstanter Dunkelheit (DD) war die Expression von rmPer, rmTim1 und rm- Cry2 rhythmisch. Allerdings war das Ergebnis für rmPer nicht signifikant, obwohl in der subjektiven Nacht höhere Expressionslevel als am subjektiven Tag gemessen wurden. Durch Ent- kopplung individueller Schrittmacher eines Tieres, oder der gemessenen Tiere untereinander (je Probe wurden die Gewebe von zehn Tieren gemeinsam gemessen), dämpfte die schon im Hell- Dunkel-Wechsel geringe Amplitude im DD wahrscheinlich so weit, dass keine signifikanten Un- terschiede mehr gemessen werden konnten. Obwohl auch die Expression von Genen, die durch die Uhrgene reguliert werden, im DD rhythmisch schwankt, weist die rhythmische Expression von rmPer, rmTim1 und rmCry2 im DD dennoch darauf hin, dass es sich bei diesen Genen auch bei R. maderae um Uhrgene handelt. xvi ZUSAMMENFASSUNG

Photoperiodische Plastizitat¨ In von LD12:12 abweichenden Photoperioden (LD6:18; LD18:6), zeigte sich, dass die Ex- pressionsmaxima aller drei Gene immer in der Dunkelphase lagen, mit dem Beginn dieser als Phasenreferenzpunkt. Ahnliches¨ wurde auch für D. melanogaster [173, 246] und zwei weitere Fliegenarten, Sarcophaga bullata und Protophormia terraenovae [86, 195] berichtet und ist eine Voraussetzung für das externe Koinzidenzmodel [288], ein Model für die Perzeption von unterschiedlichen Photoperioden. Laufradversuche zeigten, dass Tiere, die bei höheren Licht- intensitäten (1000lx) im Kurztag (LD6:18) gehalten wurden, unter freilaufenden Bedingungen eine signifikant längere Periodenlänge aufweisen. Die photoperiodische Plastizität zeigte sich also auch im Verhalten. Die Periodenlänge unter freilaufenden Bedingungen scheint vor allem von der Nachtlänge abzuhängen, wohingegen die Lichtintensität die Stärke der Periodenlängen- änderung zu beeinflussen scheint.

Circadiane Rhythmen der rmPER und rmTIM1 Konzentration und im- munhistochemisches Farbungsmuster¨ von rmPER Basierend auf der jeweiligen putativen Aminosäuresequenz wurden Antikörper gegen rmPER (zwei Antikörper, aus Mäusen und Kaninchen), rmTIM1 und rmCRY2 hergestellt. Beide rm- PER und auch der rmTIM1 Antikörper erkannten sehr wahrscheinlich spezifisch die jeweiligen Proteine in western blots. Der rmCRY2 Antikörper erkannte wahrscheinlich kein CRY2 spe- zifisches Epitop, weder in western blots noch in immunohistochemischen Färbungen. Mittels Zeitreihen von isoliertem Gehirngewebe wurde untersucht, ob die Proteinkonzentrationen von rmPER und rmTIM1 in LD12:12 circadian schwanken. Obwohl einige Experimente leichte Un- terschiede zwischen den Zeitgeberzeiten (ZT) mit Maxima in der Dunkelphase aufwiesen, konnte bei wiederholten Versuchen keine Schwankung festgestellt werden. Falls die Konzentration von rmPER und rmTIM1 circadian schwankt, war entweder die Amplitude zu gering, um circadiane Unterschiede zu detektieren, oder Schrittmacher in unterschiedlichen Phasenlagen maskierten vorhandene Rhythmen. Von den vier generierten Antikörpern erzeugte nur der anti rmPER (rabbit) Antikörper ein Sig- nal in der Immunhistochemie. Offensichtlich alle Zellkerne im Gehirn, Suboesophageal-, ebenso wie in den Thorakal- und Abdominal-Ganglien waren PER-immunreaktiv. Zu allen untersuchten ZTs wurde nur eine Kernfärbung beobachtet. Eine weit verbreitete Zellkern-Färbung wurde auch in P. americana beobachtet, allerdings schwankte hier die Konzentration in LD12:12 circadian [371]. Falls rmPER die gleiche Funktion innerhalb des circadianen Systems hat wie bei D. melanogaster, würde man circadiane Unterschiede in der subzellulären Lokalisation erwarten. Entweder erkennt der anti-rmPER Antikörper nicht alle Formen von rmPER, oder die Funktion von rmPER unterscheidet sich teilweise von der PER Funktion in D. melanogaster. In dieser Studie konnten die erstmals molekularen Daten der circadianen Uhr von R. maderae erfasst und dargestellt werden. Die Uhrgene per, tim1 und cry2 werden in dieser Schabenart exprimiert und ihre Domänenstruktur sowie das circadiane Expressionsmuster ähneln dem hypothetischen ursprünglichen Insektenuhrwerk, welches der circadianen Uhr von Vertebraten nahesteht. Das molekulare Uhrwerk von R. maderae kann sich an unterschiedliche Photoperio- xvii den anpassen, und diese Anpassungen manifestieren sich im Expressionsprofil der untersuchten Uhrgene ebenso wie im Verhalten. xviii ZUSAMMENFASSUNG 1 Summary

The change of day and night generates a daily rhythm of various environmental cues, most importantly light and temperature. Almost every organism examined to date exhibits endogenous circadian timekeeping mechanisms, and this ”circadian clock” enables an organism to anticipate external rhythms. Circadian rhythms are endogenous and persist under constant conditions. They are based upon molecular feedback mechanisms that generate rhythms in gene expression and protein abundance of approximately 24 hours. Although the basic principles and components of this clockwork are essentially the same in all insects, it became obvious in the last decade that there are various interesting variations of the scheme. Whereas the molecular circadian clockwork in the fruitfly D. melanogaster is well studied, most other insect groups lack comprehensive investigations. Almost nothing is known about the molecular basis of circadian rhythms in the cockroach R. maderae, despite its long use as a circadian model organism. To analyze the molecular circadian system of R. maderae, the structure and expression profile of the core feedback loop genes per, tim1 and cry2 were analyzed. Using degenerate primers and RACE, the full length coding sequences of rmPer and rmCry2, and a fragment of rmTim1 were obtained. Phylogenetic analysis grouped rmPER and rmCRY2 with other hemimetabolous insects, and the proteins’ domain structure suggests a similar function in the molecular clockwork as in other insects. Using quantitative PCR, it was shown that the transcripts of rmPer, rmTim1 and rmCry2 cycle in phase in different light regimes. The phase adjusts relative to the onset of the scotophase, with peak levels in the first half of the night. In constant darkness, cycling of rmTim1 and rm- Cry2 continues, but no significant differences between ZTs could be observed for rmPer, due to low amplitude cycling. Based on the acquired sequences, antibodies directed against rmPER, rmTIM1 and rmCRY2 were generated. In western blots antibodies against rmPER and rmTIM1 specifically detected protein bands in the expected molecular weight range. Time series using brain tissues did not detect any obvious cycling in protein abundance of rmPER and rmTIM1, perhaps because of low amplitude cycling. To gain information about the cellular expression of the core feedback loop genes, fluorescent immunohistochemistry was employed. Only the rm- PER antibody, derived in rabbits, yielded a signal above background. These stainings suggested a widespread expression of rmPER in the CNS without circadian oscillation, similar to the results obtained with western blots. The staining was exclusively nuclear at all times and almost every cell in the central nervous system was PER-ir. Running wheel assays were used to examine the locomotor activity of the short and long day photoperiods used in the quantitative PCR experiments. There was no significant difference in the free-running periods of animals kept at 100lx illumination, however at 1000lx animals kept in short days exhibited a highly significant prolonged free-running period compared to animals kept in long days. In summary, the first molecular data of the R. maderae circadian clock was obtained. The core feedback loop genes per, tim1 and cry2 are expressed in the cockroach, and their structure and circadian expression profile suggests a function similar to the ancestral insect and mammalian clock. The molecular clockwork of R. maderae has the ability to adapt to different photoperiods and these adaptations manifested on the molecular and behavioral level.

1 2 CHAPTER 1. SUMMARY 2 Introduction

In the course of evolution, life acquired the ability to adapt to rhythmic environmental changes. Being able to anticipate these changes can crucially improve the ability of an organism to survive. One of the most prominent and regular changes is the daily rhythm of day and night caused by earth’s rotation around its axis. One might suspect that organisms merely react to environmental changes instead of having an internal timekeeping mechanism. However, already the first known chronobiological experiments, performed by Jean Jacques d’Ortous de Mairan in 1729, showed that in constant darkness without environmental cues, the plant Mimosa pudica continues to spread and fold its leaves in a daily rhythm [56]. In the 50s and 60s of the last century the basis of modern circadian research was established by Jürgen Aschoff [7] (humans), Erwin Bünning [28] (plants) and Colin S. Pittendrigh [241] (insects). They showed, that daily rhythms are based on endogenous clocks generating a rhythm of approximately, but not exactly 24 hours, and that this endogenous rhythm can couple to external environmental cues. These rhythms are called ”circadian rhythms”, for ”circa” meaning approximately and ”dies” meaning day. Almost every organism examined to date possesses an endogenous clock mechanism, from cyanobacteria to vertebrates. A circadian rhythm has three major characteristics:

1. It must persist in constant conditions without any periodic stimulus. The period length τ (tau) in constant conditions is called the ”free-running period” and has a period length of approximatly 24 hours.

2. It must have the ability to synchronize to external rhythmic stimuli (”Zeitgeber”). This synchronization is called ”entrainment” in chronobiology. ”Zeitgeber time” (ZT) refers to the time of the Zeitgeber. For example, in a light-dark cycle of 12h light and 12h darkness, ZT0 refers to lights on.

3. A circadian clock must be temperature compensated. Most biochemical reactions are greatly contingent on temperature, with a temperature coefﬁcient Q10≃2, which means that the reaction rate roughly doubles every 10 °C of temperature increase. Nevertheless, circadian rhythmicity can be maintained in a wide range of temperatures and the circadian clock does not run faster at higher temperatures or slower at low temperatures [20, 75]. This is especially important for poikilotherm animals like insects, that are subject to sig- niﬁcant temperature changes.

As described above, one prerequisite for a circadian clockwork is the ability to entrain to environmental cues. An animal’s circadian clock can be entrained by various external rhythms like light, temperature or even social cues. The possibly evolutionary oldest Zeitgeber is temperature [241]. Although it seems contradictory to temperature compensation at ﬁrst, temperature rhythms can set phase and period [48, 82, 330] of a circadian pacemaker. However, an ambient increase or decrease of temperature does not speed up or slow down the clock, in accordance with temperature compensation.

3 4 CHAPTER 2. INTRODUCTION

The most prominent Zeitgeber to entrain the 1 A B C 2 * circadian clock is light. In nature, the rhyth- 3 * 4 * mic change of light and darkness due to earth’s 5 * 6 ...* 7 rotation entrains most organisms to the solar 8 9 day. The period length (T) of the solar day is 10 11 almost exactly 24h. The endogenous clock is days 12 13 influenced by light in two ways. First, the pe- 14 15 riod length depends on light intensity: in con- 16 17 stant light, the period is lengthened in noctur- 18 19 nal animals if light intensity is increased and 20 shortened in diurnal animals, which is known D C +4 as ”Aschoff’s rule” [238]. Second, in constant +2 A darkness or during the dark phase of the cy- 0 ∆Φ/h cle, it is important at which point during the -2 B cycle a light stimulus is applied. Light stimuli -4 0 4 8 12 16 20 24 at different circadian times can have different CT/h effects. Naturally, only in constant conditions the en- Figure 2.1: Schematic diagram of a basic circadian dogenous circadian rhythm can be observed. experiment. The locomotor activity (black bars) of The period length in constant conditions, in a nocturnal animal was recorded. In constant darkness, light pulses (yellow bars) at different circadian terms of illumination be it constant darkness times are given. The circadian time (CT) is deter- or constant light, is called the ”free-running” τ mined based on the phase reference point (CT 12, period ( ). The free-running period is de- green asterisk). Depending on the circadian time, termined by measuring the recurrend interval (A) the phase remains either unchanged (CT 6) (B) between the occurance of a certain reference is being delayed (CT 16) or (C) advanced (CT 20). point (”phase reference point”), for example, (D) Phase response curve (PRC) according to light the onset of activity. It is species specific and pulses given in the examples A-C. usually longer than 24h in diurnal and shorter in nocturnal animals in compliance with Aschoff’s rule. The subjective circadian day (with length=τ) is thus shorter or longer than 24h and so is the circadian hour being 1/24th of the circadian day. The circadian hour is the unit of the subjective circadian time (CT) the animal experiences and in nocturnal animals, the onset of activity as phase reference point is set per def- inition to CT12. External stimuli like light and temperature can ”reset” the circadian clockwork by causing phase shifts, which can be seen in constant conditions as a shift of the phase reference point. The phase can either be advanced (which means the clock has been temporarily ”accelerated”, moving the phase reference point to an earlier CT) or delayed (the circadian clock ”slowed down”, moving the phase reference point to a later CT). Importantly, the direction as well as the size of a phase shift directly depends on the phase of the circadian day. As an example, light pulses of sufficient length and intensity have no or only little effect during the subjective day, but evoke clear phase delays in the early subjective night, and phase advances in the late subjective night (Fig. 2.1). This is similar in diurnal as well as nocturnal animals and explains how these animals entrain to the solar day. 2.1. RHYPAROBIA MADERAE AND CIRCADIAN RESEARCH 5

The dual oscillator model In 1976, Pittendrigh and Daan proposed a dual oscillator model as a basis for behavioral circadian rhythms to explain the ”splitting” phenomenon in morning and evening activity peaks of mammals and birds [242]. Splitting is the phenomenon of a single activity band dissociating into two distinct bands in certain constant conditions, and this behavior cannot be explained with a single oscillator evoking activity rhythms. Thus, they proposed a dual oscillator model involving two coupled oscillators: each oscillator is independently entrained by another environmental cue, or phase locks to a divergent reference point, yet the two oscillators inﬂuence each other. The reference points proposed were sunrise (morning) and sunset (evening), and the oscillators were thus termed morning (”M”) and evening (”E”) oscillators. The model assigns the following characteristics to the oscillators:

1. The overall pacemaker period is different from the individual M or E oscillator period.

2. How the oscillators inﬂuence each other depends on the phase relationship.

3. Integrating Aschoff’s rule into the model, light intensity has diametrical effects on the M and E oscillator: light delays the E oscillator, while advances the M oscillator, so τM is assumed as a negative and τE as a positive function of light intensity. 4. Mutual coupling of the oscillators depends on environmental cues. This can also explain the splitting phenomenon: under unnatural conditions like constant light, prolonged dim- light or in certain mutants M and E oscillators uncouple and splitting of locomotor activity rhythms occurs.

2.1 Rhyparobia maderae and circadian research

Cockroaches have been widely used in chronobiological research due to their size, toughness and easy maintenance. Most species are rather primitive and unspecialized insects, which makes them interesting model organisms compared to higher specialized, derived insects. The Madeira cockroach Rhyparobia maderae (Fig. 2.2) in particular has been used in circadian research since the 1960s, and its circadian behavior has been studied excessively. In fact it was this animal, where a circadian pacemaker could ﬁrst be localized [209], and during the past 20 years it has become a model organism for behavioral, neuropeptidergic and cellular circadian research.

Rhyparobia maderae is a member of the cockroach family Blaberidae (subfamily Oxyhaloinae), originally described from Madeira by Fabricius [70], but probably originates from western africa1 [248]. Its nomenclatoric history is somewhat complicated, with the synonym Leucophaea maderae still in use by some researchers until recently. However, as stated by McE Kevan [133], the

1”From the evidence of its relatively general occurrence in tropical Africa, and its much more localized distribution in many other portions of the tropics, taken with the associated presence of closely related species in portions of West Africa, I feel maderae had its origin in that territory, possibly in that portion usually spoken of as Upper Guinea.”- James A. G. Rehn[248] 6 CHAPTER 2. INTRODUCTION correct name is clearly Rhyparobia maderae and will therefore be used in this work. R. maderae is a large cockroach of 40-50mm length and one of approximately 50 cockroach species that occur in human made environments. It is a pest insect of economic importance that was intro- duced to many tropical and subtropical countries. R. maderae is an ovoviviparous cockroach and embryogenesis occurs in the egg case within the uterus. After eclosure, the nymphs (also called larvae) go through seven (male) or eight (female) larval stages (=instars), and the development to the adult cockroach takes four to ﬁve months, depending on external factors like temperature, food and population density [45].

A B

C Styli

Cerci 2mm

Figure 2.2: Rhyparobia maderae, male. (A) Dorsal and (B) ventral view. (C) The last abdominal segments with cerci and styli. Only male cockroaches possess styli, and this distinction was used to identify specimens for experiments. Scalebar: 1 cm.

2.1.1 The central nervous system The central nervous system (CNS) of R. maderae and cockroaches in general consists of the cephalic nervous system and the ventral nerve cord with the thoracic and abdominal nervous systems. It is not as centralized as the nervous system of other insects like most Diptera, Hy- menoptera, Lepidoptera and Coleoptera and is subdivided into altogether eleven ganglia: the cephalic nervous system consist of two, the thoracic nervous system of three, and the abdominal nervous system of six ganglia. The outer hull of the ganglia is formed by the nerve sheath, which surrounds the CNS. The ganglia themselves consist of a peripheral part, also called cortex, and an inner part called the neuropil. The nerve sheath consists of the neurolemma and glial cells 2.1. RHYPAROBIA MADERAE AND CIRCADIAN RESEARCH 7 and restricts penetration of substances, serving as ”blood-brain barrier”. The cortex contains the cellbodies of the neurons, whereas the neuropil only contains neuronal processes and is the site of synaptic interaction. In addition to the neuropil, neuronal processes can also be organized in tracts, in which the processes are bundled to a common direction and usually do not form synapses. The ganglia are joined by connectives, and the connections between the bilaterally fused hemiganglia are called commissures [74]. The cephalic nervous system contains two ganglia: the supraesophageal ganglion (which is also termed ”brain” in insects) and the subesophageal ganglion, which is connected via the circumesophageal connectives with the supraesophageal ganglion. The supraesophageal ganglion can be further divided into three regions, the proto-, deuto- and tritocerebrum (Fig. 2.3). The protocerebrum processes sensory input from the complex eyes via the optic lobes and from the ocelli. The optic lobes contain three neuropils: the lamina, the medulla with the accessory medulla (AMe) and the lobula (Fig. 2.3). Also located in the protocerebrum are the pars intercerebralis (PI) and pars lateralis (PL), the mushroom bodies and the central body. PI and PL serve endocrine function and innervate the corpora cardiaca (CC), a neurosecretory organ. The mushroom bodies are important for memory and olfactory learning in insects and mainly contain projections from the densly packed dorsally located Kenyon cells. The central body is a higher integration center for sensory input overseeing motor control. The deutocerebrum receives input from the antennae and contains the antennal lobe with its olfactory glomeruli, which are connected to the mushroom bodies of the protocerebrum by projection neurons. The tritocerebrum connects the cephalic nervous system to the stomatogastric (or stomodeal) nervous system, which controls the inner organs, and receives sensory input from the labrum and the mouth parts. From the tritocerebrum the circumesophageal connectives emerge which connect the supraesophageal with the subesophageal ganglion. The subesophageal ganglion controls the mandibles, maxillae and the labium and developed from the ganglia of these parts. It is also involved in neuronal control of the salviary glands (Fig. 2.3). The thoracic nervous system consists of three ganglia, the pro- meso- and metathoracic ganglion. They control locomotor activity and integrate sensoric input of their respective segments. The abdominal nervous system of cockroaches contains six ganglia, with the sixth ganglion also called the terminal abdominal ganglion (TAG). Motoneurons in the abdominal ganglion control posture, circulation and respiration of their segments as well as the genitals and thus copulation and egg laying [74] (Fig. 2.4).

Circadian locomotor activity and the effects of light R. maderae is a nocturnal animal that hides and shows little to no activity during the day. Thus, if exposed to a 12h light, 12h dark cycle (LD12:12, period (T)=24h), the activity onsets around the light-dark transition and the animals are active throughout the dark phase, with higher activity at the beginning of the night. When raised and kept in LD12:12, the animals free-running period in constant darkness (DD) is generally less than 24h (on average 23.72 in males and 23.84 in females [222]). Interestingly, even when the animals have never been exposed to a light-dark cycle, one third of ﬁrst-instar nymphs express a free-running cycle [222]. The free-running period can be permanently altered: if nymphs develop in non 24 h periods or DD, adult cockroaches free-run with different τ: the longer the period the larvae have been exposed to, the longer the adult τ. 8 CHAPTER 2. INTRODUCTION

B Protocerebral bridge Central body

Proto- cerebrum

Supra- esophageal ganglion Deuto- Lamina cerebrum Medulla Calyx Accessory medulla Alpha lobe Mushroom body Trito- Lobula Beta lobe cerebrum Antennal lobe Lateral accessory lobe

Subesophageal ganglion

Figure 2.3: The cephalic nervous system of R. maderae. (A) Opened head capsule with exposed cephalic nervous system. Part of the head capsule, fat body, tracheae and muscles have been removed. (B) Schematic drawing of the cephalic nervous system with the most prominent neuropils. Scalebars: 500 µm. 2.1. RHYPAROBIA MADERAE AND CIRCADIAN RESEARCH 9

Brain

Cephalic nervous system

SOG

TG1

Thoracic nervous system

TG2

TG3

AG1

AG2

AG3

AG4 Abdominal nervous system

AG5

TAG

Figure 2.4: The central nervous system of R. maderae. Abbreviations: AG, abdominal ganglion; SOG, subesophageal ganglion; TAG terminal abdominal ganglion; TG, thoracic ganglion. Scalebar: 1 mm. 10 CHAPTER 2. INTRODUCTION

In contrast, entrainment of adult cockroaches to different photoperiods has little to no permanent effect on the free-running period [13, 220, 221]. There seems to be a critical period of minimal 60 days in the second and/or third instar that permanently sets the endogenous τ of the animals (Page unpublished, mentioned in [225]). However, this does not mean, that the free-running period in adult cockroaches is independent of light entrainment, since there are ”after-effects” of light entrainment in adults. After-effects describe the long lasting effect of period changes in constant conditions after previous entrainment to a speciﬁc LD cycle. In R.maderae, τ is smaller in animals that were previously entrained to T22 (LD8:14) than to T26 (LD8:16). However, these changes are weaker than the developmental effects. They are not permanent and decay after approximately 40 days in constant conditions. Most interestingly, after-effects apparently not result from a phase change in circadian oscillators, since phase resetting by cold-temperature pulses does not inﬂuence τ, and temperature entrainment does not induce after-effects [225].

0 Δ ϕ /h

-4

-8

-12

4 8 12 16 20 24 CT/h

Figure 2.5: (A) R. maderae light PRCs of white light pulses with different lengths and intensities in T 24. Blue: 700 lx, 6 h; circles, dotted line: 50.000 lx, 6 h; triangles, solid line: 50.000 lx, 8 h; ﬁlled circles, solid thick line: 80.000 lx, 12 h. (B) White light (6 h duration, 700 lx) PRCs of animals raised in T22 (green) and T26 (red). Both plots were redrawn and modiﬁed, after Wiedenmann 1977 [347] (A) and Page and Barrett 1989 [221] (A, blue and B)

The effects of light on the circadian system can be determined by exposing the free-running animal in DD to light pulses. The light PRC has a typical shape in most animals, with a phase delay peak in the early subjective night and a phase advance peak in the late subjective night. During the subjective day, light pulses usually have little or no effect. Winfree [351] differentiates two types of light PRCs: pulses of shorter duration or lower intensity lead to low amplitude ”type 1” PRCs, whereas pulses of long duration or higher intensity lead to high amplitude or ”type 0” PRCs. Type 0 PRCs have a typical shape with a ”gap” between maximum delay and advance, and ultimately light intensities and/or durations that result in a type 0 PRCs mean a full phase reset to 2.2. MOLECULAR RHYTHM GENERATION 11 a constant phase regardless of the endogenous circadian time [351]. To generalize, light intensity and the duration of the light pulse have mostlythe same effects in entrainment. In R. maderae, the light PRC has the typical shape as described above, but compared to other insects, the response to white light is rather low: pulses of 700lx and 6h duration [221] as well as pulses of 2000lx and 12h duration resulted in a maximum phase delay of approximately two hours, with an up to two fold stronger phase delay than advance. Much higher white light intensities and durations are necessary to provoke a nearly complete phase reset in R. maderae: light pulses of 80000lx with a duration of 12h resulted in type 0 PRC, whereas 50000 lx with a duration of 6h or 8h strongly increase especially the delay portion of the PRC (4-5fold higher than the advance portion), but still result in a type 1 PRC [346] (Fig. 2.5, A). In contrast, white light pulses of only 1h and 300lx pulse duration are sufﬁcient for a type 1 PRC with up to 4h phase delay and advance in Drosophila melanogaster, and 6h white light pulses with the same light intensity already result in type 0 PRC and a near complete reset of the phase [289]. The light PRC however is not static for a given light intensity and pulse duration and depends, for example, on the ZT period length (T) the animal experiences during development. If light pulses of 6h duration and approximately 700lx are used in different total T cycles in R. maderae, the delay portion of the PRC is signiﬁcantly reduced (with some animals even showing phase advances in the early night) in animals raised in T<24h (T=22), whereas the advance portion is reduced in animals raised in T>24h (T=26) [221] (Fig. 2.5, B).

The light entrainment pathway Photoreceptors are apparently the sole site for light entrainment in cockroaches, at least for R. maderae and P. americana. Neither the ocelli nor other photic input seems to be involved: ablation of the ocelli had no effect, whereas covering only the complex eyes with black lacquer results in free-running behavior, as well as surgical removement of the eyes or cutting the optic nerves [208, 261]. Two photoreceptors with maxima at 365 nm and 507 nm have been conﬁrmed in P. americana [194], and it seems to be essentially the latter that entrains the circadian clock, since these cockroaches’ entrainment sensitivity peaks at 495 nm [193]. Each optic lobe receives light input from the ipsi- as well as from the contralateral side [216, 223], but there appears to be no direct input from the histaminergic photoreceptors to the AMe [169].

2.2 Molecular rhythm generation

Molecular studies on the circadian system of animals started in 1971, when Konopka and Benzer [145] generated the ﬁrst behavioral D. melanogaster mutants using the chemical mutagen ethyl methane sulfonate. These mutants were named pers (short period), perl (long period) and per01 (arhythmic), because they showed different eclosion and adult locomotor activity rhythms. The mutations were mapped to a single locus on segment 3B1-2 of the X chromosome and subsequently to mutations of single nucleotides on the same gene, the period (per) gene [357]. These discoveries had great impact, since this was the ﬁrst time where it was shown that a single gene strongly affects behavior. After cycling of per in LD and in DD was shown on mRNA [93] and 12 CHAPTER 2. INTRODUCTION

protein [307, 361] level, with the effects of the per mutations also manifesting on a molecular level, Hardin et al. proposed a per ”feedback loop” mechanism to generate a circadian rhythm [93]. The simplest model of such a transcriptional-translational feedback regulation (TTR) consists of a gene’s transcription and translation, whereupon the protein represses its own transcription, resulting in a rhythmic accumulation of mRNA and protein. Since D. melanogaster is by far the best studied insect for the molecular basis of circadian rhythms, the basic principles for rhythm generation, entrainment, output and its cellular basis will be explained in the fruitﬂy in the following sections.

2.2.1 Transcriptional-translational feedback regulation and adjacent pathways in D. melanogaster In the last two decades, there has been excessive research on transcriptional-translational feedback regulation (TTR) in D. melanogaster. A complex regulation on transcriptional, posttranscriptional and post-translational levels was revealed (Review: [3, 92]). Several genes were identiﬁed that are directly involved in feedback regulation. mRNA and protein levels of timeless (tim1), the second ”clock gene” discovered [301], cycle in the same phase as per in LD 12:12 [302]. PER and TIM1 heterodimerize [79, 360], and their expression is regulated by the same E- box elements (CACGTG) in their respective promoters [54, 91]. E-boxes were already known to bind bHLH transcription factors [197]. Thus dCLOCK (CLK, Circadian Locomotor Output Cy- cles Kaput), a bHLH-PAS transcription factor, promotes the transcription of per and tim1 [4]. At the same time it was also shown, that CYCLE (CYC), which is also a bHLH-PAS transcription factor, is necessary for circadian rhythmicity [270]. Clk cycles in antiphase to per and tim1 [10], whereas cyc is constitutively expressed [270]. Interaction of these two transcription factors was ﬁrst shown in mammals, where their respective homologs, mCLOCK and BMAL1 (Brain and Muscle ARNT-like 1) form heterodimers [106]. CLK:CYC driven per and tim1 expression in turn is inhibited by PER:TIM1 [10, 156], which explains how the proposed feedback inhibition of per (and also tim1) works.

Posttranslational modiﬁcation and subcellular localization of PER and TIM1

That PER is phosphorylated in a daytime-dependent manner was ﬁrst shown by Edery et al. [63]. Phosphorylation of PER is under control of DOUBLETIME (DBT), an ortholog of the mammalian casein kinase Iε [139]. DBT expression is constitutive and not clock controlled [140], however the subcellular localization changes similar to PER [140] with which it forms complexes [139]. In addition to DBT, PER is also phosphorylated by CASEIN KINASE 2 (CK2) [2, 167] which promotes nuclear translocation. In a similar manner, SHAGGY (SGG), an ortholog of the mammalian glykogen synthase kinase-3 (GSK3) and also CK2 [186] phosphorylate TIM1 and promote its nuclear translocation [178]. Dephosphorylation events also inﬂuence stability and subcellular localization: PER is dephos- phorylated by PROTEIN PHOSPHATASE 2A (PP2A) [278] and TIM1 by PROTEIN PHOS- PHATASE 1 (PP1) [71], both having effects opposing those of DBT, CK2 and SGG. 2.2. MOLECULAR RHYTHM GENERATION 13

Interestingly, PER and TIM1 accumulate approximately 7 hours after TYF their mRNA’s peak. This delay is PER very important to maintain a circa- TIM1 dian rhythm, since activation of transcription, transcription itself, trans- DBTDBT lation and repression of transcription would usually be much faster CK2 SGG than 24h. Initially, it was pro- P PER DBTDBT posed that this delay results from TIM1 DBT mediated phosphorylation and PP2A PP1 subsequent degradation of PER, as Cytoplasm well as light induced degradation of Figure 2.6: The core feedback loop of D. melanogaster. TIM1 [140]. Although this surely PER and TIM1 inhibit their own expression by inhibiting the has an impact on protein accumu- CLK:CYC heterodimer, thus generating daily oscillations in lation, there is strong evidence that Per and Tim1 mRNA and protein concentrations. Modifica- translational control also contributes tion by various protein kinases and phosphatases finetunes this to this lag. TWENTY-FOUR (TYF) feedback loop. For details see text. Abbreviations: CLK, was shown to interact with per and CLOCK; CYC, CYCLE; DBT, DOUBLETIME; PER, PE- tim1 mRNAs as well as regulatory RIOD; PP1, PROTEIN PHOSPHATASE 1; PP2A PROTEIN components of translation (poly(A)- PHOSPHATASE 2A; TIM1, TIMELESS 1; TYF, TWENTY- binding protein, PABP) in some cir- FOUR. cadian pacemaker cells. Translation may be first repressed and thus delayed, and later translation is promoted by TYF. This mechanism might contribute to the observed gap between mRNA and protein peak levels [164].

The core feedback loop Binding of CLK:CYC results in per and tim1 expression. In LD12:12, this occurs at about ZT4 to ZT16. PER and TIM1 accumulate in the cytoplasm and peak about dusk, where a PER:TIM1:DBT complex forms. Complex formation with TIM1 protects PER from DBT dependent phosphorylation and subsequent degradation. The PER:TIM1:DBT complex translocates to the nucleus (LD12:12 at ZT18-ZT22, [49, 267]) after phosphorylation by SGG and CK2. It has been shown that PER can enter the nucleus independently of TIM1, and repressor activity of PER is enhanced when TIM1 is not bound. Thus, dissociation of PER:TIM1 may occur prior to nuclear entry and both enter the nucleus independently [189, 268]. However, deletion of the only identiﬁed NLS in TIM1 not only strongly decreases nuclear localization of TIM1, but also of PER, despite PER:TIM1 heterodimerization not being impaired [271]. Phosphorylation and dephosphorylation events regulate the stability and subcellular localization of the PER:TIM1:DBT complex which sets the period and phase of the TTR clock. CLK:CYC mediated transcription is inhibited by phosphorylation of CLK by the PER:TIM1:DBT (and/or PER:DBT) complex (In LD12:12 at ZT18-ZT4, [135, 205, 358]). In LD, TIM1 is light-dependently degraded after lights on, which allows for DBT dependent phosphorylation of PER. This phosphorylation 14 CHAPTER 2. INTRODUCTION

then leads to ubiquitinylation by the E3 ubiquitin ligase Supernumerary Limbs (SLIMB). Sub- sequently, PER is then degraded by the 26S proteasome [142]. After degradation of PER, CLK peaks, heterodimerizes with CYC and binds to the per and tim1 promoters, closing the loop (Fig 2.6).

Feedback regulation outside the core feedback loop In addition to per and tim1, the CLK:CYC heterodimer also acti-

vates E-boxes in the promoters of CWO

three feedback loop related genes: VRI

vrille (vri), par domain protein 1 PDP1 (mainly isoforms ε and δ, pdp1ε/δ) and clockwork orange (cwo). These genes dominate the regulation of amplitude, phase, and output of the CLK core feedback loop. Expression of vri [19, 50] and pdp1 [50] is activated at about ZT4 to ZT16. In phase with per and tim1 mRNA Cytoplasm and in antiphase to clk mRNA and protein, vri mRNA as well as pro- Figure 2.7: The clk loop of D. melanogaster. The protein of tein accumulate to peak levels at the CLK:CYC controlled gene VRI activates clk expression, whereas the protein of the also CLK:CYC controlled pdp1 in- about ZT14 [19, 50]. In con- hibits clk expression approximately four hours later. Together, trast, PDP1 maxima occur approx- this leads to a circadian oscillation in CLK concentration. In imately four hours later, at about addition, CWO also inhibits CLK:CYC driven expression, and ZT18 [50]. Both proteins bind to thus its own expression, creating another feedback inhibition V/P-boxes in the clk promoter, but loop. For details, see text. Abbreviations: ACT, activating whereas VRI inhibits clk expression sequence; CLK, CLOCK; CWO, CLOCKWORK-ORANGE; [50, 84], PDP1ε/δ moderately ac- CYC, CYCLE; PDP1, PAR DOMAIN PROTEIN 1; UA, un- tivates clk transcription four hours known activator; VRI, VRILLE. later [17, 50, 363]. Together, VRI feedback inhibition as well as delayed PDP1ε/δ activation lead to oscillating clk mRNA levels in this ”clk loop”. Abolishing PDP1/VRI driven clk transcription in clkJRK and cyc0 mutants leads to constitutively high clk levels. Apparently a to date unknown activator constitutively activates clk transcription independent of feedback regulation [83] (Fig. 2.7). The bHLH orange-domain repressor CWO inhibits CLK:CYC driven expression by repressing E-box activity [128, 163, 180], but can also activate circadian expression [256], suggesting a more diverse role of CWO action on circadian expression. In vivo CWO seems to mainly activate per, tim1, vri and pdp1ε expression, but repression seems to prevail on cwo expression itself, thus creating another feedback inhibition loop [256] (Fig. 2.7). The core feedback loop and the clk loop are interconnected in two ways. First, CLK activates the transcription of all E-box regulated genes, namely per, tim1 as well as vri and pdp1. The 2.2. MOLECULAR RHYTHM GENERATION 15

phase and amplitude of clk mRNA oscillation however has little effect on Per/Tim1 oscillation and locomotor rhythms [134]. On the other hand, the PER:TIM1:DBT complex not only inhibits CLK:CYC E-box mediated per/tim1 expression but also CLK:CYC E-box mediated expression of vri. Simpliﬁed, the core feedback loop runs without the clk loop, but not vice versa, indicating that the core feedback loop ”dominates” the clk loop.

Light entrainment pathways In D. melanogaster, the loss of signaling from opsin-based photoreception like CRY1 compound eyes and ocelli does not lead h·f to a loss of light entrainment of locomotor rhythms [112, 212, 227]. On molecu- CRY1* TIM1 lar level, it was shown that degradation of P TIM1 is induced by light [115, 198, 360], although TIM1 itself is not light sensi- TIM1 JET tive. A cryptochrome (CRY1) was identi- P fied to directly transmit light information JET CRY1* COP9 to the clock in a cell-autonomous manner [66, 310]. In D. melanogaster, cry1 is mainly a bluelight photoreceptor, regulated like clk by VRI/PDP1, resulting in circadian oscillations in phase with CLK Figure 2.8: Direct light input into the circadian clock of [50]. Flies overexpressing cry1 are hy- D. melanogaster. Light activates CRY1, which then binds persensitive to light pulses [66], in con- TIM1. Both proteins are thereafter bound by JET, which trast to cryb mutant flies that do not react leads to degradation. For details see text. Abbreviations: to short light pulses [310]. Also, cryb mu- COP9, COP9 signalosome; CRY1, CRYPTOCHROME 1; tant flies still show locomotor rhythms in f, frequency; h, Planck constant; JET, JETLAG; TIM1, constant light [67] whereas wildtype flies TIMELESS 1. are arhythmic [146]. However, cryb mutant flies can be entrained to temperature cycles, show robust cycling of clock genes in some neurons in constant darkness and are behaviorally rhythmic in LD12:12 [310], albeit light entrainment is impaired [67]. In summary, CRY1 appears to mediate direct light input to the moleular oscillator, but is not exclusively responsible for light entrainment. Only loss of all known light input pathways (compound eyes, ocelli, H-B eylet and cry1), renders D. melanogaster unentrainable to light [98]. In addition to CRY1, two factors have been identified to be important for transmitting light information to the core feedback loop: the F-box protein JETLAG (JET) [144] and the COP9 signalosome (CSN) [141]. Upon activation by light, CRY1 binds to TIM1 [26] and becomes a target for JET binding [230]. TIM1 protects CRY1 from JET mediated degradation, as TIM1 bound by activated CRY1 is a higher affinity target for JET. Binding of CRY1 to TIM1 also promotes phosphorylation of TIM1 and subsequent binding of JET, which mediates degradation of TIM1 [230]. After binding of JET, proteasomal degradation of TIM1 depends on the CSN [141] (Fig. 2.8). The reduction of TIM1 levels then leads to time-dependent phase shifts: during the 16 CHAPTER 2. INTRODUCTION

early night, TIM1 levels are increasing. Thus, TIM1 degradation delays accumulation which results in a phase delay. During the late night, TIM1 levels are falling and degradation accelerates the reduction in TIM1 levels, resulting in a phase advance. During day or subjective day, TIM1 levels are very low, and further reduction has hardly any effect on the circadian phase.

Clock output pathways In contrast to feedback regulation, the output pathways downstream of the feedback loops are much less understood. Microarray studies identified hundreds of cycling transcripts with little overlap [31, 39, 168, 183, 333] and re-evaluation of these datasets revealed 214 cycling transcripts, including the genes directly involved in feedback regulation like per, tim1, cry1, clk, pdp1, vri and cwo [132]. The amount of cycling transcripts is tissue-specific. In clock neurons, the proportion of cycling mRNAs is much higher than in other tissues [150]. One of the major clock output factors is CLK, since CLK not only regulates expression of the TTR loop genes, but directly regulates expression of approximately 1500 genes. Most of these genes also bind CYC and PER, indicating a regulation similar to that of the feedback loop genes [1, 39]. Although most cycling mRNAs also have cycling nascent RNA, a lot of mRNAs cycle without their respective nascent RNA showing circadian expression. This indicates that output is not only mediated by transcriptional, but also to a significant part by posttranscriptional regulation [262]. Genes identified to be important for clock output include ebony [316], lark [318], neurofibro- matosis 1 [349] and takeout [16] as well as voltage gated ion channels that mediate neuronal output by regulating membrane excitability. Slowpoke (SLO), a voltage gated Ca2+-activated potassium channel, acts downstream of the clock gene expressing cells, since slo mutants are arhythmic, and rescueing slo expression in these cells does not fully restore rhythmic behavior. SLO and Slowpoke binding protein (SLOB), a putative regulator of SLO, are both rhythmically expressed [31, 55, 126, 125]. Slo expression is also regulated by dyschronic (dysc), an ion channel regulator that seems to be specifically required for locomotor output [127] while being regulated by CLK [1]. Other ion channels important for rhythmic behavior are narrow abdomen (na) and shaw. Flies with a mutation in the na gene show reduced circadian rhythms which can be rescued by targeted expression in pacemaker cells [155], regulated shaw expression in pacemaker neurons is important to maintain normal rhythmicity [104]. The molecular clock remains unaltered upon manipulation of both genes, hinting to a role of these channels in the output pathway.

2.2.2 The circadian molecular clockwork in non-Drosophilid insects D. melanogaster is by far the best studied insect for the molecular basis of circadian rhythms, and multiple genetic tools and mutants allow for a detailed dissection of the mechanisms underlying circadian rhythms. Single clock genes have been cloned in several other species, and thanks to genome sequencing projects complete sets of clock genes have been identiﬁed in some. Although most of the clock genes are conserved, it turned out that there is quite some variability amongst insect species. The most striking difference between insect clocks concerns Tim1 and Cry. The fact that insects also express a second, ”mammalian-like” cry2 was ﬁrst reported for Danaus 2.2. MOLECULAR RHYTHM GENERATION 17

plexippus [367]. Several other insects with different sets of core feedback loop genes followed. This led to the assumption that there is an ”ancestral” clock in insects, where both crys and tim1 are expressed as shown for D. plexippus. Several different clocks in which some of these genes are missing may have derived from this hypothetical ancestral clock [359, 366, 367] (Fig. 2.9). However, these are not the only differences as will be shown in the following with some examples.

D. melanogaster A. mellifera, S. invicta Lepidoptera ("derived") ("derived, mammalian-like") ("ancestral")

CRY1* CRY1* h·f h·f ? h·f TIM1 TIM1 PER PER per, tim1 PER per,cry2? CRY2 per,tim1,cry2 CRY2 E-Box E-Box E-Box ~CLK CYC CLK CYC~ ?~CLK CYC

TIM1 PER PER PER CRY2 CRY2

Figure 2.9: Examples of insect clock core feedback loop models. For details, see text. Except for D. melanogaster, functional analysis for some core feedback loop genes are missing in each model. The tilde indicates which of the two genes’, clk or cyc, protein concentration cycles. Abbrevitaions: CLK, CLOCK; CRY1, CRYPTOCHROME 1; CRY2, CRYPTOCHROME 2; CYC, CYCLE; f, frequency; h, Planck constant; Per, PERIOD; TIM1, TIMELESS 1.

The ”mammalian-like” clockwork in social hymenopterans Within the non-Drosophilid holometabolous insects, hymenopteran and lepidopteran species have been studied most intensively. After D. melanogaster, Apis mellifera was the second species whose complete set of clock genes was studied [269]. Recently, also an analysis of the clock gene expression in the ant Solenopsis invicta was published [121]. These holometabolous, social hymenopteran insects share several clock gene features, which clearly distinguish them from D. melanogaster and there is quite some resemblance to the mammalian circadian clock (Fig. 2.10). First, no orthologs of tim1 and cry1 are encoded in the genome, but an ortholog of cry2 [121, 269]. Tim1 and cry1 are absent similar to mammals. Furthermore, as in mammals, CRY2 is a transcriptional repressor [359]. In A. mellifera and S. invicta, CRY2 oscillates in phase with PER and contains all domains necessary for nuclear localization and transcriptional inhibition [121, 269]. Second, in these two species clk is constitutively expressed and cyc oscillates in antiphase to per, in contrast to D. melanogaster, where clk oscillates and cyc is constitutively expressed. Also, in these insects CLK lacks the C-terminal region responsible for transcriptional activity in dmCLK, and CYC contains a C-terminal transactivation domain also found in the mammalian CYC homolog BMAL1, but not in D. melanogaster [121, 269] (Fig. 2.9). Third, expression patterns of vri and pdp1 are also different from those observed in D. melanogaster. Whereas vri and pdp1 expression is controlled by the CLK:CYC 18 CHAPTER 2. INTRODUCTION

heterodimer, and thus mRNA oscillates in phase with per in D. melanogaster, vri in S. invicta possibly cycles in antiphase to per and pdp1 appears to be constitutively expressed [121]. Fourth, cwo expression in these hymenopteran insects peaks similar to its mammalian orthologs [121, 263], PE R1/2 but not D. melanogaster’s, although CRY1/2 CWO domain structure is more similar within these insect species than between hymenopterans and mammals [121]. CK1α CK1δ CK1ɛ

Two cryptochromes in butterflies PER1/2 CRY1/2 PP1 Several lepidopteran species have PP5 been examined for circadian genes, Cytoplasm like Antheraea pernyi [32, 33], Bom- Figure 2.10: The core feedback loop in the suprachiasmatic nu- byx mori [124] and D. plexippus cleus (SCN) of mammals. Transcription of per1, per2, cry1 and [359, 365, 366]. Per and tim1 cry2 is activated either by a CLK:BMAL1 or NPAS2:BMAL1 mRNA levels cycle in these species [57] heterodimer. Although there are three per genes, only with maxima in darkness phase [76, PER1 and PER2 appear to play a role in the circadian system 124, 255, 366] and cyc is constitu- of the SCN [229]. PER1/2 and CRY1/2 (both type 2 CRYs) tively expressed in B. mori, simi- heterodimerize in the cytoplasm and heterodimerization, stabil- lar to D. melanogaster [177]. How- ity and nuclear translocation are controlled by posttranslational ever, at least in A. pernyi, CYC modification, for example by protein kinases and phosphatases. but not CLK contains a transactiva- In the nucleus, CLK/NPAS2:BMAL1 mediated transcription is tion domain, similar to A. mellifera inhibited by PER1/2:CRY1/2 [226]. Abbreviations: BMAL1, and other insects [33]. Interestingly, Brain and Muscle ARNT-like 1; CLK, CLOCK; CRY1/2, there seems to be no lag between per CRYPTOCHROME 1 or CRYPTOCHROME 2; NPAS2, NEU- mRNA and protein in A. pernyi in RONAL PAS DOMAIN-CONTAINING PROTEIN 2; PER1/2, PERIOD 1 or PERIOD 2; PP1/5, PROTEIN PHOSPHATASE contrast to D. melanogaster [279]. 1/5. All butterflies examined so far express cry1 and cry2. Distinct functionality of the two Crys was shown in D. plexippus: CRY1 appears to mediate light induced degradation of TIM1 and light input to the core feedback loop as in D. melanogaster. However, in contrast to D. melanogaster, CRY2 is apparently the major repressor of CLK:CYC mediated transcription and not PER. CRY2 translocates to the nucleus and potently inhibits CLK:CYC mediated transcription [366], whereas PER could not yet be detected immunocytochemically in the nucleus of clock gene expressing cells in D. plexippus [366], A. pernyi [279] or B. mori [300] using antibodies (Fig. 2.9).

The clockwork in hemimetabolous insects All insects mentioned so far are holometabolous insects. Unfortunately, circadian clock genes and their function in hemimetabolous insects still lack comprehensive investigations. Circa- 2.2. MOLECULAR RHYTHM GENERATION 19

dian genes of crickets and some heteroptera have been published and the only complete set of circadian genes based on a genome sequencing project so far is known from the pea aphid Acyrthosiphon pisum. Only few functional studies have been published, mostly based on gene silencing by RNA interference (RNAi). Unfortunately, there is no comprehensive functional and genomic study of at least the core feedback loop genes in a basic hemimetabolous insect so far. In Gryllus bimaculatus, per and tim1 cycle with maxima in the subjective night [53, 191]. Knock- down of per and clk via RNAi stops circadian oscillations of mRNA levels as well as rhythmic locomotor activity indicating that these genes are essential for circadian rhythmicity as in D. melanogaster and other insects [191, 192, 190]. However, knockdown of tim1 did not abolish locomotor rhythms in constant conditions, although rhythmic expression was eliminated [53]. In contrast, D. melanogaster tim1 knockout mutants are arhythmic. Similar to A. mellifera and to the mammalian clock, clk is rather constitutively expressed and cyc oscillates. Surprisingly, cyc RNAi induced cyclic expression of clk, which indicates that both genes may be rhythmically expressed in G. bimaculatus [334]. This has not been observed in any other insect so far, and hints at yet another regulatory mechanism specific for some insects only. The bean bug Riptortus pedestris is another hemimetabolous insect that has been studied for clock genes. Expression of per, cyc, vri and cry2 has been shown in this species [117]. Clk expression was previously shown in heteroptera in the linden bug Pyrrhocoris apterus [319]. Rhythms in clk gene expression are very low in long day and absent in short day photoperiods [319], pointing to a constitutive expression of clk in heteroptera as in A. mellifera, mammals and in certain conditions in G. bimaculatus. Knockdown of cry2 using RNAi in R. pedestris suspends circadian cuticle deposition and induces per expression, indicating that CRY2 works within the core feedback loop as previously described for other insect species [116]. The pea aphid A. pisum is the only hemimetabolous insect where the complete set of circadian clock genes has been analyzed [46]. A. pisum expresses per, tim1, cry1, two cry2 genes (cry2-1 and cry2-2), clc, cyc, vri and pdp1. Per, tim1 and cry2 (primers did not distinguish between both cry2) oscillate with maxima around dusk and early night as in other insects. Clk mRNA levels did not oscillate, but cyc levels did. This goes along with CYC having a transactivation domain in A. pisum. Several characteristics indicate that the A. pisum clock is a highly derived clock: the core feedback loop genes clk, cyc, both cry2 and especially per and tim1 appear to have co-evolved at an accelerated speed, being highly modified compared to other insect species. This high divergence seems to be specific for the A. pisum clock, since other hemimetabolous insects clock gene sequences did not share this feature. Also, domains important for protein interaction and nuclear import of PER were either missing or highly modified. In contrast, CRY2-1 and CRY2-2 domains necessary for nuclear translocation and transcriptional repression were conserved. This indicates that CRY2 may also be the main repressor in the core feedback loop of A. siphon. An- other difference seems to concern direct light input. Jet, in D. melanogaster described as a gene necessary for direct light input, is absent from the genome of A. pisum, despite the presence of tim1 and cry1. In other insects that also lack jet, like A. mellifera and T. castaneum, either tim1 or cry1 is absent. Cockroaches haven’t received much attention so far. The per cDNA sequence of Blattella germanica has been published, the structure of the protein has been analyzed and the circadian expression of per was examined. Per mRNA peaked during the night at ZT22 in LD16:8 cy- 20 CHAPTER 2. INTRODUCTION cles, and cycling continued in constant darkness [166]. In addition to B. germanica, the per sequence of Blattella bisignata, and partial sequences of Periplaneta americana per and tim1 (GeneBank accession number AY116671) have been published. Unfortunatly, no Blattodean genome is available to the scientific community to date. In summary, there are striking differences between insect molecular clocks. The main differences seem to concern expression of cry1, cry2 and tim1, and hence main transcriptional repression and light input. cry2 is expressed in all insects so far examined for this gene with Drosophilid species being the only exception, and CRY2 may be the main factor for transcriptional repression in most insects as it is in mammals. The role of CLK and CYC seems to be ”reversed” in many insects compared to D. melanogaster, with CYC being the main transcriptional activator in a hypothetical ancestral insect circadian clock. Although more species need to be examined in detail, such a hypothetical ancestral clock may have contained one copy of cry1, cry2 and tim1, and single genes have been lost or duplicated in some species. Since in this thesis, circadian genes and proteins were investigated in the hemimetabolous cockroach R. maderae, the structure of per, tim1 and cry2 will be described in the following more detailed.

2.2.3 Period Period (Fig. 2.11) was first cloned from D. melanogaster in 1984, and was initially proposed to be a proteoglycan due to a dipeptide threonine-glycine repeat (TG-repeat) [247]. However, later it was shown to share PER-ARNT-SIM-domains (PAS-domains) with several bHLH transcription factors [202] and subsequent studies revealed that these PAS domains are protein dimerization domains, first demonstrated for PER [114]. To date the PAS domain region is the best characterized structure of D. melanogaster PER, and crystal structures are available. It is located in the N-terminal half of PER and contains two PAS domains (PAS-A and PAS-B) and two C-terminal helices [137, 353]. PER-TIM1 interaction has been mapped to the PAS domains [79, 101, 272] as well as binding of PER homodimers. The classic perl mutation for example prevents homodimerization, which results in delayed nuclear entry [137, 153] and impeded temperature compensation [113]. Interaction of the αF helix C-terminal to the PAS-B domain is essential for this homodimerization and the αF helix may provide a conformational switch between PER homodimerization and PER:TIM heterodimerization [137, 353]. In addition to TIM1, PER also interacts with the CLK:CYC heterodimer, DBT and CRY1. Inhibition of the CLK:CYC heterodimer is mediated by the CLK:CYC inhibition domain (CCID), which is the most C-terminal functional domain identified [33]. There are two sites of possible DBT interaction: the N-terminal end of PER [139] and a motif in the conserved C3 region of the CCID, termed dPER-DBT binding domain (dPDBD) which may be the main DBT interaction site [136]. The second conserved stretch within the CCID, the C4 region has been identified to be necessary for CLK binding, and thus was termed the CLK binding domain (CBD). Although CLK binding is impaired when this domain is missing, transcriptional repression is not, most probably because TIM1 can mediate PER-CLK interaction [317]. PER probably also interacts with CRY1 with the C-domain [113, 265]. This may be important especially in some peripheral oscillators, where CRY1 is necessary for core feedback loop function in D. melanogaster [149, 161]. 2.2. MOLECULAR RHYTHM GENERATION 21

C1 C2 C3 C4 C5 C6

V243D(perl) S589N (pers)

A PAS B CLD C-Domain TG dPDBT CCID CBD αF αE 512 755 809 977926 73 77 238 390 452 524 585 600 685 764 813 840 1034 NLS S/M NLS DBT? CRY1 DBT CLK TIM1

Figure 2.11: Structure of D. melanogaster PERIOD, Canton-S strain. Relative length of the domains are up to scale. Abbreviations: A, PAS-A domain; αF, αF helix; αE, αE helix; B, PAS-B domain; C1-C6, conserved sequence blocks 1-6; CBD, CLK binding domain; C-Domain, PAS C-terminal domain; CCID, CLK:CYC inhibition domain; CLD, cytoplasmic localization domain; CRY1, CRYPTOCHROME 1 binding site; DBT, DOUBLETIME binding site; dPDBT, dPER-DBT binding domain; NLS, nuclear localization signal; PAS, PER-ARNT-SIM domain region; perl, per long mutation site; pers, per short mutation site; S/M, short mutable region; TG, threonine-glycine repeat; TIM1, TIMELESS 1 binding site.

The subcellular localization is mediated by three putative sites: The cytoplasmic localization domain (CLD), a region C-terminal to the PAS-B domain inside the PAS domain region, contains a sequence that prevents nuclear translocation [272]. Nuclear localization is mediated by at least two nuclear localization signals (NLS), one basic monopartite near the N-terminus [14] and another, functional bipartite within the CCID [33]. The extended dipeptide TG-repeat [357] seems to be specific for Drosophilid species. It is of variable length and lies within a variable region that can also contain other peptide repeats. In D. pseudoobscura, the repeat consists of five degenerate peptides with more than 30 copies, whereas in D. virilis the dipeptide repeat consists of only two pairs [42, 228]. The D. melanogaster TG- repeat consists of 14-24 dipeptide pairs [292] and is important for the species specific courtship song cycle [264, 345] and for temperature compensation. Allels with different variable length regions have different abilities of temperature compensation. The circadian clocks of flies with higher numbers of repeats are better buffered against temperature changes [264, 292] and it was shown that the number of dipeptide repeats correlates with the latitudinal distribution, with the longer allels being predominant at higher latitudes [47, 293]. In other insects, for example A. mellifera or B. germanica, a short serine-glycine repeat of unknown function is found instead of a TG-repeat [166].

PER function in other insects There are hardly any functional analyses of PER domains in non-Drosophilid species, but it is generally accepted that conserved domains probably have similar functions. Most of the domains described above have been found to be conserved in PER proteins of several other species, though with exceptions. For example, PER is poorly conserved compared to other species in the derived circadian clock of A. pisum. The NLSs identiﬁed in D. melanogaster are missing in A. pisum, and the PAS domain region as well as CCID are highly divergent [46]. The conserved blocks of sequences (c1-c6) deﬁned by Colot et al. [42] from the comparison of some Drosophila species are also found in other non-Drosophilid species, but the PER proteins of most other insects ex- 22 CHAPTER 2. INTRODUCTION

amined also contain another conserved block of sequence C-terminal to c6, called c7 [166]. This motif is especially interesting, since most insects with c7 motif also express CRY2 and a similar C-terminal domain in mammals was shown to interact with CRY2 [331].

Phosphorylation of PER Also in D. melanogaster it was ﬁrst shown that PER is phosphorylated progressively in a daytime dependent manner [63]. The PER phospho-status is regulated in a complex manner by several kinases (DBT [245], CK2 [2, 166] and probably also SGG [143]) as well as phosphatases (PP1 [71] and PP2A [278]). Basically, de novo synthesized PER is phosophorylated by DBT and highly unstable in the cytoplasm if not stabilized by TIM1. DBT [12, 51] and CK2 [2, 167, 165] mediated phosphorylation then leads to nuclear import. There, PER is progressively phosphorylated further. This results in increased repressor activity [204] and ultimately hyperphosphorylated PER is SLIMB-dependently degraded in the proteasome [38, 142]. DBT alone phosphorylates multiple sites, and PER contains phosphorylation sites of different functionality [138]. The phosphorylation of one region can gate the phosphorylation of other regions [38]. For example, an inhibition of the DBT mediated phosphorylation of the short mutable region (S/M) leads to in- stability of PER and to increased repressing power, whereas additional inhibition of a region C-terminal to the S/M region results in increased stability and little repressor activity [138]. In summary, various events regulate stability, subcellular localization and repressor activity and hence the properties of PER, which ﬁnetunes the speed of the circadian clock.

2.2.4 Timeless Timeless 1 (Fig. 2.12) was the second gene shown to be involved in circadian rhythmicity [301]. In contrast to per, timeless 1, mostly just called timeless, has been much less studied. TIM1 does not contain PAS domains, but binds PER with two functionally identiﬁed regions (PER binding site 1, PER-1 and PER binding site 2, PER-2) [272]. Located within the ﬁrst region lies a bipartite NLS important for nuclear TIM1 and PER import [199, 271], and a region important for cytoplasmic localization was mapped to the last 160 residues [272].

P243L(timPL)

AD 1 2 CLD

505 541 553 578 715 914 1238 NLS PER PER

Figure 2.12: Structure of D. melanogaster TIMELESS 1, S-TIM1398 isoform. Relative length of the domains are up to scale. Abbreviations: 1, PER binding site 1; 2, PER binding site 2; AD, acidic domain; CLD, cytoplasmic localization domain; NLS, nuclear localization domain; timPL, tim proline to lysine mutation site.

Light dependent degradation of TIM1 is mediated by interaction with CRY1 [30] and JET [144, 231, 230]. The exact CRY1 and JET binding sites of TIM1 however are unknown. Similar to 2.2. MOLECULAR RHYTHM GENERATION 23

PER, TIM1 is progressively phosphorylated day-time dependently, although to a lesser degree compared to PER [9]. Phosphorylation of TIM1 by SGG renders it more sensitive to light- mediated degradation and to enhanced nuclear translocation [178]. The exact sites involved in both actions are unknown [9]. In addition to SGG, phosphorylation by at least one more kinase seems to enhance the light induced degradation of TIM1 as has been shown by pharmacological studies [200]. In D. melanogaster strains examined in the lab, there are two major allels of tim1: ls-tim and s-tim. The tim1 mRNA contains two start codons, but whereas in ls-tim two isoforms of TIM1 are produced, L-TIM1421 and S-TIM1398, a G deletion in s-tim results in a stop codon before the second start codon and hence in only S-TIM1398 and a short 19 aa peptide being produced. L- TIM1421 is more stable, probably due to impaired light-dependent CRY1 binding, and combined TIM1 levels are higher in ls-tim ﬂies [277]. This explains the reduced photoresponsiveness observed in ls-tim ﬂies [231], which more readily enter diapause [327]. In nature, there is a clear correlation between the frequency of the ls-tim allele and geographic distribution, and the ls-tim allele might have developed as an adaptation to temperate environments [327].

TIM1 function in other insects Some non-Drosophilid insects have been examined for TIM1 structure and function. In the primitive insect Thermobia domestica, PER-1, PER-2 dimerization domains, NLS and CLD are conserved and TIM1 is essential for circadian locomotor rhythmicity, similar to D. melanogaster [129]. In G. bimaculatus TIM1 however, these domains are also conserved suggesting a similar function, but here TIM1 is expendable for rhythmicity and might only have a ”ﬁne tuning” role: TIM1 knockdown in G. bimaculatus leads to shorter free-running periods [53], but not arhythmicity.

Timeless 2 (Timeout) The gene timeless 2 (tim2), also called timeout is a paralog of tim1. Tim1 probably separated from tim2 around the cambrian explosion [269]. In contrast to TIM1, TIM2 is most probably not part of the circadian clock in animals. In D. melanogaster, TIM2 is required for chromosome stability. It also plays a marginal role in circadian photoreception, which might explain how TIM1 became an important factor in the circadian clock in the course of evolution [18]. ”timeless” in mammals is more closely related to timeout than the insect tim1. Like tim2 in insects it possesses hardly any function in the circadian clock.

2.2.5 Cryptochrome The flavoprotein photoreceptor group of cryptochromes (crys) has first been identified in the plant Arabidopsis thaliana, where it is an important factor in growth and development. Cryp- tochromes belong to the large family of photolyases/cryptochromes, that can be divided into three major subfamilies: the cyclobutane-pyrimidine dimer (CPD) photolyases, the 6-4 pyrimidine- pyrimidone adduct (6-4) photolyases and the cryptochromes. Photolyases are important for DNA 24 CHAPTER 2. INTRODUCTION repair and use the absorbed light energy to repair UV damage. Cryptochromes share highly conserved chromophore binding domains in the N-terminal region (photolyase homology domain, PHD) with the photolyases, but in addition to these they have an extended C-terminal region (C-terminal tail, CTT) which is divergent amongst cryptochromes and missing in photolyases. Cryptochromes evolutionary descend from the photolyases, and are functionally defined by not being primarily DNA repair proteins [36]. There are two major types of cryptochromes: the cry- DASH cryptochromes (named after their occurence in Drosophila, Arabidopsis, Synechocystis, and Homo) evolved from a CPD photolyase and do not share a common ancestor with the other cryptochromes. They have ssDNA repair functionality, but most probably also an additional role in signaling. The second group are the ”classic” plant and animal cryptochromes without DNA repair functionality. Plant cryptochromes descend from CPD photolyases, whereas animal cryptochromes descend from 6-4 photolyases [36, 170]. The animal cryptochromes as such consist of the light receptive, ”insect-like” type 1 crys which act as circadian photoreceptors, and ”mammalian-like” type 2 crys which act as transcriptional repressors and do not respond to light [36]. Due to historical reasons, the nomenclature of these cryptochromes is somewhat confusing. ”Cry1” in mammals is also a type 2 cryptochrome (mammals lack type 1 cryptochromes), and the ”mammalian-like” cryptochromes can as well be found in insects. Plant cry1 and cry2 form a separate group, and are not related to animal cry1 or cry2, respectively. In this thesis, animal, ”insect-like” type 1 cryptochrome will just be called ”cry1”, and ”mammalian-like” type 2 will be termed ”cry2”.

Cryptochrome 1 In D. melanogaster, CRY1 (Fig. 2.13) was first shown to be an ultraviolet-A/bluelight photoreceptor that directly feeds light information into the circadian clock [66, 67, 310]. Also, CRY1 proteins of other insect species exhibit light sensitivity [359]. Like all cryptochromes, CRY1 has two major domains, the PHD and the CTT. Most members of the photolyase/cryptochrome family contain two chromophores: reduced FAD (FADH-) as catalytic cofactor (binding at the FAD-B), and an additional chromophore that serves as an ”antenna” and transfers its excitation energy via Förster resonance energy transfer (FRET) to FADH-. The PHD of CRY1 contains an FAD-binding domain (FAD-B), but seems to lack a second chromophore [303, 369]. The CTT mediates light dependent interaction with TIM1: CRY1 lacking the C-terminal 20 aminoacids is ”constitutively active” and much less stable, binding TIM1 (and PER) in light and dark [26, 59, 100, 265]. CRY1 is either stabilized by its CTT binding to the PHD (which ”mimics” DNA binding of photolyases), or to a lesser extent by binding of TIM1 to the PHD. Conformational changes in the CTT depending on the electronic state of FAD thus mediate light induced TIM1 binding [369]. CRY1 does not only integrate bluelight information into the circadian clock, but is also magnetosensitive. The electomagnetic field with a frequency of 10Hz shows a circadian oscillation and can serve as a Zeitgeber [62, 68, 344]. It is perceived by CRY1 wavelength-dependently, probably by singlet-triplet transition in the photopigment according to the radical pair hypoth- esis [259]. It has been shown experimentally, that D. melanogaster behaviorally responds to a magnetic field of 10Hz in a wavelength and field strength dependend manner and that the pe- 2.2. MOLECULAR RHYTHM GENERATION 25 riod length is influenced. This effect is probably mediated by CRY1, since it is absent in cry1 knockout flies [77, 356]. The behavioral response itself though seems to be independent of the circadian clock and CRY1-TIM1 interaction [77, 78]. In a similar clock independent way, light- activated CRY1 can also directly modulate ion channel activity and neuronal firing rate in the supraesophageal ganglion of D. melanogaster [73, 304]. Interaction of CRY1 with the visual transduction complex via interaction of PDZ domains in the CRY1 CTT with inactivation no afterpotential D (INAD) suggest an additional role of CRY1 in D. melanogaster vision [181].

D410N(cryb) crym PHD FAD C

225 513 522 Figure 2.13: Structure of D. melanogaster CRYPTOCHROME 1, Relative length of the domains are up to scale. Abbreviations: C, C-terminal tail (CTT); cryb, crybaby mutation site; crym, crym mutation site; PHD, photolyase domain; FAD, FAD binding domain.

In addition to CRY1’s function in signaling, it also appears to be necessary for the clock oscillator function itself in some, but not all peripheral tissues. Rhythms in the amplitude of elec- troantennogram (EAG) responses to stimuli in the antennae of cryb mutant ﬂies are lost because of disruption of the core feedback loop [149, 161]. CRY1 is also required for oscillator function in forelegs, Malpighian tubules and eyes, where it can function as a repressor of CLK:CYC activated expression [41, 123, 161].

Cryptochrome 2 Insect CRY2 (Fig. 2.14) is expressed in all non-Drosophilid insect species examined so far. It is a potent transcriptional repressor, and transcriptional repression was proven experimentally with CRY2 from several insect species, like A. mellifera (Hymenoptera), T. castaneum (Coleoptera), D. plexippus (Lepidoptera) and Riptortus pedestris (Hemiptera) [119, 359]. Interestingly, D. plexippus CRY2 has been shown to be magnetosensitive like CRY1 in a light-dependent manner [78] and the FAD binding pocket is highly conserved in M. musculus CRY1 (a type 2 CRY), probably able to bind FAD [52]. Several studies however did not detect any light responsiveness of insect CRY2 [73, 359], but there is indication for light responsiveness of H. sapiens CRY1 (a type 2 CRY) in cell cultures [103].

PHD 2a 1 2b FAD C CT

184 215 230 282 308 215 230 293 297 NLS Figure 2.14: Structure of Apis mellifera CRYPTOCHROME 2. Relative length of the domains are up to scale. For details, see text. Abbreviations: 1, repressive domain 1 (RD1); 2a, repressive domain 2a (RD2a); 2b, repressive domain 2b (RD2b); C, coiled-coil region; CT, C-terminal tail (CTT); NLS, nuclear localization signal; PHD, photolyase domain; FAD, FAD binding domain. 26 CHAPTER 2. INTRODUCTION

Functional domains have not been defined in insect CRY2 yet, but insect CRY2 and vertebrate CRY2 aminoacid sequences are overall highly conserved. Some domains defined in vertebrates are well conserved in insect CRY2. Repressive domains (RDs), which have been defined using chimaeric Danio rerio CRY2, are well conserved in CRY2 proteins. They have been shown to be necessary for transcriptional repression and nuclear localization. An NLS within the RD2b domain could also be identified [102, 269]. A small region essential for M. musculus CRY2- PER2 interaction is conserved in insect species [215]. This region is almost identical with a coiled-coil region that has been shown to mediate nuclear translocation in M. musculus CRY1 [37]. Two residues (Arginine 501 and Lysine 503) within this region, whose substitution leads to a loss of CRY2-PER2 interaction [215] are conserved in all insect CRY2 proteins. Deletion of the CTT in M. musculus CRY1 leads to a loss of CLK:BMAL1 mediated repression and plant photolyases gain repressor function upon fusion of the CTT, indicating that the CTT is important for transcriptional repression as well [37]. In the derived clock of A. pisum, CRY2 is not as well preserved as in other insect species and there are two copies of the gene in the genome [46].

2.2.6 Posttranslational regulation Transcriptional-translational feedback regulation (TTR) was the first model explaining how circadian rhythms on cellular level are being generated [93]. Posttranslational events like protein modifications, interactions and nuclear translocation can greatly influence period length and amplitude of the feedback oscillation, however the central dogma of transcriptional feedback- inhibition is a prerequisite for circadian rhythm generation in this model. Although some form of TTR was observed in many organisms and although it is undoubtedly a basic mechanism whose failure often leads to arhythmicity, it is important to note that molecular circadian rhythms are also generated in other, TTR independent ways. A circadian rhythm can often be observed when the expression of genes involved in TTR is held constant or even missing. Therefore, posttranslational regulation (PTR) can be sufficient for circadian rhythm generation. For example, the rhythmic KaiC autophosphorylation in the cyanobacterium Synechococcus elongatus can generate a self-sustained, temperature compensated circadian rhythm [110, 201]. In the eukaryotic alga Ostreococcus tauri, rhythms of peroxireduxin oxidation are sufficient to drive at least one additional cycle of timed protein accumulation when transcription and translation is inhibited [214]. This was observed also in vertebrates, where a circadian rhythm can be generated in the complete absence of transcription: in human red blood cells, oxidation of peroxireduxin dimers also occurs in a circadian manner. These rhythms continue in constant conditions and are entrainable, showing all properties sufficient for circadian oscillators [213]. However, although PTR is sufficient for a circadian oscillation, in Synechococcus elongatus TTR seems to be necessary for oscillator synchrony and stability [328]. In Ostreococcus tauri, it depends on the time of treat- ment, if transcriptional and translational inhibition has an impact on the circadian clock [214]. Taken together this indicates, that the entirety of circadian rhythms on a molecular level can not be reduced to either TTR or PTR, but that there is an interplay of transcriptional-translational and posttranslational oscillations. 2.3. THE CELLULAR BASIS FOR CIRCADIAN RHYTHMICITY 27

2.3 The cellular basis for circadian rhythmicity

2.3.1 Clock neurons in Drosophila melanogaster In D. melanogaster, approximately 75 core clock gene expressing neurons per hemisphere in the dorsal and lateral brain maintain behavioral rhythms (Fig. 2.15). These ”clock neurons” are sufficient and corequisite for rhythmic locomotor activity [111, 130]. The clock neurons were categorized by their position and size, so this classification does not reflect ho- mogenous groups: neurons within each group can differ in terms of clock gene, neuropeptide and neurotransmitter expression, indicating specific functions. Basically, the clock cells are categorized into two major groups, the dorsal neurons (DN) and the lateral neurons (LN). The dorsal neurons are subdivided in three groups, DN1-3. The DN1 are further subdivided into two groups: a pair of anterior, larger neurons (DN1a) and approximately 15 smaller posterior cells (DN1p). The DN2 Figure 2.15: Clock neurons and their projections in consist of only two cells, the DN3 of approx- the brain of D. melanogaster. For details see text. imately 40 cells. All DN project to the dor- After Helfrich-Förster et al. 2007 [97]. Abbrevia- sal protocerebrum, an output region connected tions: Al, antennal lobe; AMe, accessory medulla; to locomotor output and neuroendocrine func- Ca, calyx; DN1-DN3, dorsal neuron groups 1-3; l- LN , large ventral lateral neurons; LN , dorsal lat- tion, with overlapping terminals. Some DN1 v d eral neurons; LPN, lateral posterior neurons; Me, and DN3 also project to the AMe. The lateral neurons consist of four subgroups: the small medulla; MB, mushroom body; s-LNv, small ventral lateral neurons. (s-LNv, five cells including the fifth s-LNv) and large (l-LNv, four cells) ventral lateral neurons, the dorsal lateral neurons (LNd, six cells) and the lateral posterior neurons (LPN). The PIGMENT DISPERSING FACTOR immunoreactive (PDF-ir) s-LNv project to the dorsal protocerebrum and to the AMe and receive light information from the H-B eyelet. The fifth s-LNv only projects to the AMe [97]. In the dorsal protocerebrum, the PDF-ir s-LNv project to the vicinity of the DN2, the pars intercerebralis (PI, a cluster of neuroendocrine cells) and the calyx of the mushroom bodies, which have been shown to regulate sleep [237]. The terminals of the s-LNv in the dorsal protocerebrum overlap with those of the DN and the LNd, which also project to the dorsal protocerebrum. Furthermore, the LNd project to the ipsilateral AMe [97] (Fig. 2.15). The PDF-ir l-LNv are the only clock neurons with known arborizations that do not project to the dorsal protocerebrum. They arborize ipsilateral on the surface of the medulla and in the AMe and send fibres to the contralateral medulla’s surface via the posterior optic tract, mediating light input from the complex eyes [73] 28 CHAPTER 2. INTRODUCTION and probably coupling both hemispheres’ pacemaker centers [97] (Fig. 2.15).

Morning and evening cells D. melanogaster has a bimodal activity pattern with two typical activity bouts around dawn (ZT0, lights-on) and dusk (lights-off). Activity already increases before dawn and dusk, showing that the circadian system anticipates these events. To understand how the neuronal network and in particular the clock neurons evoke the bimodal locomotor activity, various mutants, driver lines and the GAL4-UAS [25] and GAL80 [157] systems were used, allowing for a more detailed analysis on cellular level (Review: [99, 210, 355]). A model has been proposed that categorizes the clock cells in two types in line with Pittendrigh and Daan’s dual oscillator model: the ”M-cells”, responsible for morning anticipation, and the ”E-cells”, responsible for evening anticipation [89, 314]. Based on various experiments, the PDF positive s-LNv and CRY-positive DN1p have been assigned to be M- cells, while the fifth PDF negative s-LNv, three or four LNd and Figure 2.16: Cells of the probably the CRY-negative DN1p have been assigned to the E- cells (Fig. 2.16). In constant light (if CRY1 mediated light in- D. melanogaster circadian system that were designated morn- put is compromised), the M-cells alone cannot drive rhythmicity ing (M) and evening (E) cells. but are necessary for rhythmicity in constant darkness, as the E- Blue: E-cells, Red: M-cells. cells alone cannot maintain rhythmicity in DD. It seems to de- Modified after Helfrich-Förster pend on the light regime which cell group sets the clock: in et al. 2007 [97]. Abbreviations: short days and darkness the M-cells determine phase and am- see Fig. 2.15; DN1p, posterior plitude of the rhythm, and in long days and light the E-cells dorsal neuron group 1. [196, 236, 257, 258, 315, 313, 362]. Analysis of split rhythms also confirmed the dual oscillator model: the short period component, which sometimes (but not always) originates from the M-peak and seems to rely on the M-cells, shortens upon increasing light intensities, whereas the long period component originating from the E peak and probably from the E-cells is prolonged [61, 257, 354]. However, the assignment of the clock cells to one of the two groups is not without contradictions and can depend on the experimental setup. The two oscillator model might be too simplified to explain all experimental data, and there is probably significant flexibility in the network with regard to M- or E-cell designation [304]. Moreover, the rhythmic locomotor output generated by the M- or E-cells cannot simply be coupled to the phase of the TTR clock genes: tim1 mRNA and PER protein levels are in phase in LNv and LNd [8, 304, 314]. 2.3. THE CELLULAR BASIS FOR CIRCADIAN RHYTHMICITY 29

2.3.2 The circadian pacemaker controlling locomotor rhythms in R. maderae In the 1960s the circadian pacemaker controlling locomotor activity in cockroaches was located in the optic lobes [209], after it was previously suggested that it is located in the subesophageal ganglion [94, 95]. Removal of the optic lobes permanently abolishes rhythmic locomotor behavior in R. maderae [209, 219] and other cockroach species, like P. americana [260] and Blaberus craniifer [171]. The location of the pacemaker controlling rhythmic locomotor behavior as well as rhythmic changes in complex eye sensitivity [43, 350] was subsequently narrowed down with microlesion experiments to the cortex in a ventral region between the medulla and the lobula [309]. Transplantation experiments showed that the mutual coupled [217] pacemakers indeed reside in the optic lobes, since host animals regained rhythmicity and adopted the free-running period of the donor animals [218, 219]. Later, further lesion and transplantation experiments located the pacemaker in neurons associated with a small neuropil, the AMe [234, 252, 312].

The accessory medulla The AMe itself is a pear-shaped neuropil of approximately 90µm length at the anterior-proximal edge of the medulla. It is non-retinotopically organised and consists of a loose shell neuropil, a core neuropil and the anterior neuropil which is connected to the shell. The core neuropil is composed of an irregular formed nodular core of dense neuropil that is embedded in an intern- odular neuropil [250, 253]. Overall, approximately 300 cells are associated with the AMe: 250 in a group anterior, medial and ventral of the AMe, and 50 in a posterior group. The anterior group contains six morphologically distinct main groups named after their location relative to the AMe: the distal and medial frontoventral (DFVNe and MFVNe), the medial (MNe) and ventral (VNe), the ventromedial neurons (VMNe), and the ventroposterior (VPNe) neurons [250, 253]. In addition to these groups, a seventh group is associated with the accessory medulla called the anterior neurons (AN) [321] (Fig 2.17). The AMe contains large amounts of dense-core vesicles, which have been shown to be involved in storage and transport of peptides within neurons. Sev- eral peptides have been identiﬁed in the AMe: allatostatin and allatotropin [234], baratin [211], corazonin [234], FMRFamid related peptides like leucomyosuppressin and short neuropeptide F [234, 320, 321], gastrin/cholecystokinin [234], leucokinin [234], several myoinhibitory peptides [299], orcokinin and pigment dispersing factor (PDF) [234, 250, 253]. The structure and the large amount of different peptides in the AMe indicate, that peptide based neuromodulation plays an important role in the function of the AMe and hence the circadian pacemaker.

PDF-immunoreactive neurons PDF-ir neurons near the AMe were originally proposed as circadian pacemakers in R. maderae due to their morphological characteristics and lesion experiments [234, 312]. The cell bodies of some PDF-ir neurons were located at the exact position where previous lesion experiments [309] predicted the circadian pacemaker. In addition, the projection pattern ﬁts well to predicted pacemaker functions: PDF-ir neurons project to the midbrain, where information could be passed 30 CHAPTER 2. INTRODUCTION

to downstream neurons controlling locomotor activity, the contralateral optic lobe allowing for the previously shown mutual coupling of pacemakers [217] and ﬁnally the lamina, where they could control circadian modiﬁcation of the visual system [350]. Functional analysis using PDF injections near the AMe showed that this peptide has indeed a prominent role in the circadian clock, since injections in the vicinity of the AMe provoke time-dependent phase-shifts [233]. Thus, and since no clock genes have been known in R. maderae, the analysis of pacemaker sites focused primarily on PDF-ir neurons.

Figure 2.17: (A) Main projections of the largest (dark red) and medium sized (blue) PDFMe in R. maderae.(B) Accessory medulla associated neurons of R. maderae. The number of gray circles does not reﬂect the group size. Colored circles represent PDFMe and the number of each color represents the mean cell number of the respective PDFMe group. After S¨ohler et al 2011 [322]. Abbreviations: AMe, accessory medulla; Al, antennal lobe; AOC, anterior optic commissure; Ca, Calyx; Cb, central body; DFVNe, distal group of frontoventral neurons; ILP, inferior lateral protocerebrum; dPDFLa, dorsal PDF-ir lamina neurons; dSLP, dorsal lateral protocerebrum; La, lamina; Lal, lateral accessory lobe; Lo, Lobula; Mb, mushroom body; Me, medulla; MFVNe, medial group of frontoventral neurons; MNe, medial neurons; PDFMe, PDF-ir medulla neurons; Pb, protocerebral bridge; POC, posterior optic commissure; POTu, posterior optic tubercles; SLP, superior lateral protocrebrum; VNe, ventral neurons; VMNe, ventromedial neurons; vPDFLa, ventral PDF-ir lamina neurons.

There are three main sites of PDF-ir cell bodies in the cephalic nervous system: two located dorsally (dPDFLa) and ventrally (vPDFLa) to the lamina and one near the AMe (PDFMe). Ap- proximately twelve AMe associated neurons are PDF-ir and can be further subdivided into three groups containing four cells each: the large (one being considerably larger), medium and small aPDFMe. The large and the medium aPDFMe belong to the VNe, whereas the small aPDFMe are assigned to the DFVNe AMe neurons. In addition, there are two to six PDF-ir neurons in a more posterior located group (pPDFMe) [250, 253]. The pPDFMe however are not a physiolog- ically distinct group. pPDFMe differ in size and count between individual cockroaches and the sum of large pPDFMe and large aPDFMe is always six. Also, large posterior as well as large 2.3. THE CELLULAR BASIS FOR CIRCADIAN RHYTHMICITY 31

anterior PDFMe are both neither FMRFamid-ir nor anti-Asn13-orcokinin-ir. Taken together, this suggests that some PDFMe seem to randomly move to a more posterior position during development [322]. Not much is known about the projections of individual PDF-ir neurons so far, however there is some data on cell group projections and their putative functions. The PDFLa project into the lamina and at least partially into the AMe. Most large and medium aPDFMe project into the ipsilateral lamina, medulla, AMe, the midbrain and the contralateral optic lobe. The small aPDFMe project only into the ipsilateral optic lobe and are possibly interneurons of the AMe. Coupling of the AMae seems to be accomplished by the largest aPDFMe and three medium sized aPDFMe, which directly connect both AMae [251, 249, 322] (Fig 2.17). Interestingly, the PDFMe and their ﬁbre network exhibit plasticity depending on period length and photoperiod. For example, the number of medium aPDFMe increases in animals that were raised in longer Zeitgeber periods, and in T24 the number of medium and large aPDFMe as well as pPDFMe increases in animals raised in longer photoperiods [340]. A similar effect could be shown when adult R. maderae were transferred for at least 6 weeks to different photoperiods [131]. This indicates that these changes are no permanent developmental effects (i.e. cell death) but are transient effects that manifest in changing PDF-ir and may be related to behavioral after-effects observed earlier [225].

2.3.3 Circadian oscillators outside the clock neurons and in peripheral tissues

Circadian pacemakers are not limited to neurons in the cephalic nervous system. Whereas the central clock controls the overall rhythmic behavior of an animal, peripheral clocks govern tissue and organ specific functions (reviewed in [329]). Independent neuronal pacemakers can for example reside in neurons outside the cephalic nervous system: motoneurons in D. melanogaster exhibit circadian changes in the size of synaptic boutons, even in decapitated animals without central pacemaker [184, 185]. Also in D. melanogaster, it has been shown that glial cells have a widespread per expression. Per expression is also rhythmic in tissues outside the nervous system, like the compound eyes, antennae and various other tissues like legs, wings, proboscis [243] and Malpighian tubules. Sensory systems probably have been most intensively studied for peripheral oscillators in insects. The compound eyes have been shown to exhibit central clock independant rhythms in moths and D. melanogaster [243], whereas electroretinogram (ERG) rhythms in R. maderae seem to depend on the central pacemaker [43, 350]. Circadian pacemakers also reside in the antennae of insects. D. melanogaster exhibits circadian rhythms in clock gene expression and the amplitude of EAG responses to stimuli. The EAG rhythms depend on PER, TIM1 and CRY1 rhythms [148, 149] and specific rescue of clock gene rhythms in the olfactory receptor neurons (ORNs) is sufficient to drive EAG rhythms [326]. Thus, an independent circadian clock governing olfactory rhythms resides in D. melanogaster antennae. Circadian rhythms in EAG responses as well as clock gene expression have also been shown in various moth species: clock genes are expressed in the antennae of Manduca sexta [297], Mamestra brassicae [187], and Spodoptera littoralis, and the latter was also shown to exhibit circadian rhythms in the EAG response to pheromone stimuli as 32 CHAPTER 2. INTRODUCTION

well as rhythmic clock gene expression [188]. However, in some moths circadian rhythmicity in odor responses seems to be mediated by a central pacemaker, since it was shown that there are no circadian rhythms in EAG responses to stimuli in Agrotis segetum, yet the response to odor was clearly under circadian control [266]. In R. maderae, the amplitude of EAG responses to stimuli exhibits a circadian rhythm, but this rhythm seems to be driven by a central pacemaker [224], though individual olfactory neurons in R. maderae show rhythmic responsiveness to stimuli even after the optic lobe was removed [273]. Cuticle deposition rhythms have been shown in various insects. In cockroaches, these ryhthms seem to be completely independent from the pacemaker controlling locomotor activity. Cuticle deposition rhythms in R. maderae as well as B. craniifer continue in constant conditions and are temperature compensated with a period length of approximatly one day. However, these rhythms are unentrainable by light, the phase and period can be different from the locomotor activity, and they persist when separated from the locomotor pacemaker [172, 339, 348]. In D. melanogaster this peripheral oscillator is also independent from the central clock and here it was shown that it uses the same molecular mechanism as the central pacemaker. However, in contrast to cockroaches it is entrainable by light via CRY1 [122]. Peripheral oscillators have also been shown to reside in the digestive [34, 35], reproductive [15] and excretory system of insects. For example, Malpighian tubules, the excretory system in insects, maintain their endogenous circadian rhythms in terms of rhythmic per and tim expression if they are transplanted to another host in D. melanogaster, indicating that the clock in this tissue runs independent from the pacemaker in the central nervous system [65, 80, 81]. In contrast, the Malpighian tubules of G. bimaculatus, exhibit no circadian rhythm in per and tim oscillation, not even in LD [336]. Interestingly, the molecular mechanism of the circadian clock can differ from the central pacemaker in peripheral tissues. In D. melanogaster, cryb mutants have normal PER and TIM oscillations in the central clock neurons, but none in Malpighian tubules [123]. Also, rhythms in EAG responses are abolished in these mutants [149]. This indicates, that CRY1 may function inside the core feedback loop in some peripheral oscillators, instead of merely mediating light input. In summary, peripheral oscillators have been shown to exist in various tissues in different species, many of them exhibiting rhythms independent of the central pacemakers. Whether non neuronal tissues exhibit circadian rhythms can be species dependent and even within the same species the components of the molecular mechanism underlying circadian rhythmicity sometimes seem to adopt tissue speciﬁc functions. 2.4. PHOTOPERIODISM 33

2.4 Photoperiodism

Figure 2.18: The tilt of earth’s rotational axis is the basis for changing daylength.

In addition to the circadian rhythm which is caused by the earth’s rotation on its axis, the 23.4°tilt of the axis causes the annual seasonal rhythm as earth rotates around the sun (Fig. 2.18). Over the year and depending on latitude, the light:dark ratio (”daylength”) of the 24h solar day (Fig 2.19), but also temperature and other environmental factors like precipitation change. These changes have impact on the behavior and physiology of most insects: First, the majority of animals exhibit rhythms that are coupled to certain times of the day (i.e. they are diurnal, 24 nocturnal or crepuscular). If the animal lives 66.55 °S 22 in an environment where the relation between day and night changes, the circadian clock has 20 60 °S to adapt to changing light:dark ratios in the 18 same total period in order to maintain the daily 16 50 °S 40 °S rhythmicity. So for example, if an animal is 14 30 °S 20 °S 10 °S inactive during the day and is mainly active 12 10 °N 20 °N during dusk, the system has to reliably pre- 10 30 °N dict dusk during changing daylength. Second, Hours of daylight 40 °N 8 50 °N there are certain behaviors and physiological changes that are triggered by changing pho- 6 60 °N toperiods. This photoperiodic response is set 4 2 off when a certain light/dark ratio is reached 66.55 °N and the point where 50% of the animals ex- Jan Feb Mar Apr May Jun Jul Aug Sep Okt Nov Dec hibit this response is called the critical pho- Month toperiod (CPP). The mechanism that can de- Figure 2.19: Approximate daylength in the course tect different light/dark ratios and lead to a of the year for several latitudes. Red: northern hemi- photoperiodic response is called the ”photope- sphere; blue: southern hemisphere. riodic timer”. However, the photoperiodic response is not triggered when the CPP is reached once, but usually several successive days are necessary. The mechanism that ”counts” the days that lead to a photoperiodic response is called the ”photoperiodic counter”. Consequently, at least four elements are necessary for a simple model that explains photoperiodism [147, 291]: 1. Input pathways via photoreceptors. 34 CHAPTER 2. INTRODUCTION

2. A photoperiodic timer that can detect different light/dark ratios, distinguishing short from long days.

3. A photoperiodic counter that accumulates or ”counts” information from successive short or long days.

4. Output pathways that forward the information to downsteam effectors that trigger the photoperiodic behavior.

The most prominent response to photoperiod and thus best studied is diapause, which will mainly be used as an example in the following. In diapause, the developmental stage of the insect is ar- rested in order to survive negative environmental conditions. Usually an insect species can enter diapause only in a certain life stage and this stage varies between insect species. The direct trigger of diapause is the endocrine system: the lack of certain molting hormones triggers diapause during the larval and pupal stages, and the lack of hormones responsible for the maturation of reproductive organs triggers adult diapause [58, 294]. Photoperiod has been shown to be the main signal to mediate daylength, however little is known about how this signal is mediated to the endocrine system.

2.4.1 Photoperiod and circadian rhythm Photoperiodism and the circadian system both rely on daily environmental changes. Also, the circadian system has to adapt to a changing light:dark ratio to maintain clock output timed to a certain time of the day. Thus it is obvious to assume an involvement of the circadian clock in photoperiodism, which was first suggested by Bünning [27]. It is still controversely discussed however if the circadian system is directly involved in photoperiodism, and if it is involved, how exactly. Historically, essentially two protocols were used to assess the question whether photoperiodism is based on the circadian system in a species: the Nanda-Hamner or resonance [203], and the Bünsow [29] protocol. In experiments using the Nanda-Hamner protocol, the animals were kept in different total periods in which the length of the photophase was kept constant, but the length of the scotophase was increased each experiment. Then, the percentage of diapausing animals was plotted against the total period. Saunders [291] lists 22 species of arthropods tested with this protocol, of which 16 exhibited a positive response and six species a negative one. If the circadian system underlies the photoperiodic response, recurrent peaks of diapausing animals are observed that occur in approximately 24 h intervals. In experiments using the Bünsow protocol, the length of the photoperiod was also kept constant with a long scotophase of 48 or 72 hours. Short light pulses of one to two hours disrupted the scotophase systematically, and the time of the light pulse in the scotophase that induced diapause was then plotted against the percentage of induced diapause. Similar to the Nanda-Hamner protocol, a circadian rhythm underlying photoperiodism was suspected when alternating maxima and minima of diapausing animals were observed during the scotophase. To further dissect the mechanism underlying photoperiodism, night interruption experiments were used. The animals were kept in diapause inducing conditions, for example LD12:12 cycles, and the dark phase is then systematically interrupted with 2.4. PHOTOPERIODISM 35

short (1h) light pulses and the percentage of animals going into diapause is observed. These night interruption experiments in a 24h period length supported the external coincidence model, one of the photoperiodic clock models. More than ten models explaining photoperiodism have been proposed and most involve the circadian clock as an integral part [291]. The three most prominent models will be explained in the following.

The hourglass timer The oldest and simplest model to explain photoperiodic responses in insects is the hourglass timer model: a mechanism measures day or night length by the accumulation of a chemical sub- stance, and the response is triggered when a critical level is reached. This was suggested for insects that fail the Nanda-Hamner protocol, like the aphid Megoura viciae. In this species, the autumnal morph occurs when the night length exceeds 9.5h and the occurance remains con- stantly high without any oscillation in increasing scotophases, neglecting the involvement of the circadian system [158, 159, 287]. The basic model of an hourglass timer does not involve the circadian clock in photoperiodism. However, there are certain hints that night length measurement of the hourglass timer in some insects including M. viciae is based on dampened circadian oscillators. In these cases the circadian system underlying photoperiodism may dampen so quickly in darkness, that no oscillation is observed upon prolonged night length [286].

The external coincidence model The external coincidence model was originally proposed by Bünning in 1936 [27] and was the first proposal of a circadian clock being involved in photoperiodism. In T24 (LD12:12), short light pulses of approximatly 1h in the early as well as the late night produce diapause averting effects in several species [239, 240]. With the light PRC and the phase adjustment of the circadian clock relative to dusk in mind this can be explained as follows: a light pulse in the early night results in a phase delay, and this phase delay leads to a photoinducible phase at the end of the night being ”pushed” into the successive main light phase. On the other hand does a light pulse at the late night directly coincide with the photoinducible phase (Fig. 2.20). In both cases, diapause averting behavior is observed. In natural LD cycles this means that if the photoinducible phase falls into the scotophase in several successive days, this is being accumulated by the photoperiodic counter and this ultimately leads to diapause. In other words, the driving force of the photoperiodic timer in the external coincidence model is the measurement of night length. An example for a species that seems to confirm the external coincidence is the fleshfly Sarcophaga argyrostoma. It shows a positive response to Nanda-Hamner cycles with interval peaks approximatly 24h apart, indicating a general circadian oscillation underlying photoperiodism [280]. Night interruption experiments in the early as well as in the late scotophase resulted in diapause aversion. If the successive dark phase was prolonged, light pulses in the early night did not have an effect, indicating that light pulses in the early night indeed shifted the photoinductive phase into the following main photophase. Prolonging the dark phase after a light pulse given in the late night, did not counteract the effect of the light pulse, in accordance with directly hitting the photoinducible phase and the external coincidence model [283, 284, 286]. 36 CHAPTER 2. INTRODUCTION

0 6 12 18 0 6 12

Figure 2.20: Night interruption experiments and the external coincidence model. If a photoinducible phase falls in the scotophase, a photoperiodic counter is activated, and after a while diapause is induced (A). If light pulses at the early or late night are given, light illuminates the photoinducible phase either by phase shifting it into the main photophase (B), or by directly hitting the photoinducible phase (C). Both light pulses induce non-diapause development. The same happens, if the photoinducible phase falls in the main photophase (D). Gray line: PER,TIM1 level in Drosophila and many other insects. Orange line: light PRC in Drosophila and many other insects. Dot: photoinductive phase. Gray background: scotophase, white bars in the scotophase indicate light pulses.

The internal coincidence model In contrast to the external coincidence model in which light has a dual role as entrainment factor as well as being a photoinductor, light merely entrains the oscillators in the internal coincidence model. Based on two light entrainable oscillators, the photoperiod is detected by the phase relationship of the two oscillators [238, 240, 332]. The wasp Nasonia vitripennis is an example for Nanda-Hamner experiments revealing separate dusk and dawn oscillators, and both oscillators were shown to be involved in diapause [282]. Temperature cycles in constant darkness induce diapause with a ”critical thermoperiod” of about 13h, indicating an involvement of the temperature entrainable circadian system and suggesting some kind of internal coincidence, since no light was given [281]. In contrast, no thermoperiodic effect was observed in S. argyrostoma, further supporting an external coincidence model for this insect [285].

2.4.2 Circadian clock genes and photoperiodism Clock gene mutants in different species as well as RNAi mediated knockdown strongly effect photoperiodism, supporting a role of clock genes in photoperiodism. In the bean bug Riptortus pedestris, per and cry2 knock-down with RNAi has diapause inhibiting effects, whereas clk and cyc RNAi has diapause promoting effects [85, 118, 119, 120]. A tim1 mutant in the Drosophilid fly Chymomyza costata which has almost undetectable tim1 levels is unresponsible to photoperiods and never enters diapause [311]. A mutant of P. apterus that fails to enter diapause has similar low levels of per in long as well as in short days. In contrast, the wildtype has approximately ten times elevated per levels during short days and diapause is induced [105]. Defects in 2.4. PHOTOPERIODISM 37 photoperiodism could also be traced to changes in expression of circadian genes in other species including Sarcophaga bullata [87] and Drosophila triauraria [352]. Data from all these species indicate an involvement of single clock genes in photoperiodism, but there is also data object- ing an involvement. In D. melanogaster, circadian genes seem to be only arbitrarily involved in photoperiodism. Per01 mutants are arhythmic, but can discriminate photoperiods and enter diapause [64, 290]. Tim01 flies can enter diapause. However, a certain proportion of the flies always enters diapause independent of the photoperiod, but dependent on two tim1 allels [327]. This indicates that at least tim1 is involved in the photoperiodic timer. In another Drosophila species, D. littoralis, it has been shown that diapause and the circadian clock governing eclosion rhythms are genetically separable: the cross of a ”northern” and a ”southern” strain having different diapause CPPs and eclosion rhythms could be selected to have ”northern” eclosian rhythms and a ”southern” CPP [154]. A similar lack of correlation between photoperiodism and the circadian clockwork can be observed in the mosquito Wyeomia smithii [23, 24, 179]. The lack of correlation between the circadian clockwork and photoperiodism does not exclude a participation of the circadian clock in photoperiodism. One model explaining this lack of correlation that includes the circadian clock and the external coincidence model claims that the position of the photoinducible phase changes in various mutants. This leaves the oscillation of the clock genes and hence circadian behavior unaffected, but changes the CPP and thus diapausing behavior [323]. There is also another likely explanation: all examples mentioned before used a system wide knock-down of clock gene expression or mutations that effect all systems using these genes. That circadian clock genes can have different tissue specific functions has been shown in several studies, for example in D. melanogaster where CRY1 seems to have different roles in the peripheral and central pacemaker controlling locomotor activity. A recent paper indicates a tissue specific role of clock genes in photoperiodism: clk, cyc, cry2 and pdp1 are likely involved in a tissue specific signaling system downstream of juvenile hormone in the gut of P. apterus, which is independent from central circadian oscillations of these genes [11]. The system controlling photoperiodism may only partly recruit the same genes as the circadian clock, and these genes may assume tissue specific functions independent from the central circadian oscillator governing circadian behavior like locomotor or eclosion rhythms. Although debatable, the circadian clock has been shown to be involved in most insects photoperiodism on a behavioral level. However, it is largely unknown how exactly the circadian system is involved, and whether it is the system as a whole, or merely some of the genes adopting specific functions in the photoperiodic system independent of the circadian system, or even both. Likely, in detail there may be many differences between species, similar to the circadian system itself.

2.4.3 Photoperiodism in cockroaches Hardly any work has been done on photoperiodism in cockroaches. Experiments performed on Parcoblatta pennsylvanica by Wassmer and Page show a photoperiodic response that is different from most other insects [338]. Whereas most insects respond to changing daylength with a switch like change in their physiology that is triggered by the CPP (like diapause), P. pennsylvanica undergoes a ”graded response” in which the animals gradually gain weight and molt faster with increasing day length. Nymphs of all stages can undergo this graded response, allowing all 38 CHAPTER 2. INTRODUCTION

stages to respond to environmental changes. A circadian system seems to underlie the photoperiodic response: neither the absolute day nor night length determined the growth rate, neglecting an hourglass model. However, a T cycle experiment similar to the Nanda-Hamner protocol, with 3h light and different total period lengths ranging from T21 to T27 resulted in a trough around T24 and higher values from T21 to T23 and T26 to T27. The authors suggest an external coincidence model with a photoinducible phase, but interestingly there was never a full response achieved in the T cycle experiment. One explanation by the authors is that the duration of light (3h) was too short to produce the full effect, which would also explain how the graded response is achieved [338]. The ability to exhibit a photoperiodic response in various stages was also observed in Periplaneta japonica. This semivoltine2 species undergoes one to two diapauses. All nymphs except the first instar can go through developmental arrest of the first or early diapause, only the last instar however can enter the second or late diapause [305]. It was also shown in this species, that the photoperiodic response pattern of different populations adapted to the geographical occurance. In the Okinawa population (26°N), nymphal development is faster in LD16:8 than LD12:12. In the Hachijo population (33°N) the speed of nymphal development is inverted: nymphs develop faster in LD12:12 than in LD16:8 [365]. Several diapauses seem to be common in long-lived cockroaches. The small subtropical cockroach Symploce japonica undergoes three diapauses with very slow nymphal development: a short day diapause in midlevel instars in the first winter, a long day diapause in later instars in the successive summer, and an adult diapause in the 2nd winter to time adult reproduction to the following spring. Thus, depending on the nymphal stage, the speed of development was either decelerated by short days (winter diapause) or by long days (summer diapause) [325]. It should be noted that growth did not completely stop in diapausing nymphs, indicating that these cockroaches also exhibit a graded response as described by Wassmer and Page [338], although termed diapause by the authors [325]. The speed of nymphal development of another subtropical cockroach species, Opisoplatia orientalis, shows hardly any response to photoperiods. Furthermore, changing photoperiods did not induce adult diapause. Developmental speed and adult reproduction was rather under direct temperature control. However, the number of female instars was under photoperiodic control: the proportion of females with seven instead of six instars was significantly higher in short day animals [364]. In summary, cockroaches can exhibit a wide range of photoperiodic responses throughout all developmental stages, and the circadian clock probably underlies photoperiodism at least in one species. The complex photoperiodic responses of the subtropical species Symploce japonica, Opisoplatia orientalis and P. japonica are especially interesting in regard to R. maderae. Some populations examined were collected on Hachijo island and live in a similar latitude (33°N) as some R. maderae populations, like the one on Madeira (32°N), or even closer to the equator on Okinawa (26°N). Although a photoperiodic response primarily depends on the life cycle and climate rather than on the proximity to the equator, it shows nevertheless that the physiology of subtropical and maybe even primarily tropical species can undergo complex photoperiodic or seasonal responses.

2”Voltinism” in entomology describes the number of generations per year. Semivoltine species need more than a year to develop, in contrast to uni- bi- or tri- or polyvoltine species that can bear one, two, three or more generations per year. 2.5. AIMS OF THIS STUDY 39

2.5 Aimsof thisstudy

The circadian clock of R. maderae controlling locomotor activity is well examined on cellular and behavioral level, but nothing is known about the molecular rhythm generation. To begin analyzing the molecular clockwork of R. maderae, circadian clock genes of this species should be cloned and analyzed. The core feedbackloop genes per, tim1 and cry2 were cloned and their expression was analyzed using quantitative PCR. The daily expression pattern was determined in the supraesophageal ganglion and the AMe as well as in a peripheral tissue, the Malpighian tubules. Using animals raised in different photoperiods, the adaptation to different photoperiods was observed. Antibodies were raised against R. maderae PER, TIM1 and CRY2 and used in western blot time series and immunohistochemical stainings to determine the daily oscillation in protein abundance and the spatial expression pattern. To gain information about the circadian activity of the animals used in the photoperiodic experiments, the locomotor activity of these animals was monitored using running-wheels. 40 CHAPTER 2. INTRODUCTION 3 Materials and methods

3.1 Animal rearing

The R. maderae colonies at the Universities of Kassel and Marburg originate from animals obtained from the University of Regensburg in 1980. Mass cultures of approximately 250 imagines and countless juveniles in various stages were kept in large polyethylene boxes (65·40·40 cm, l·w·h). These boxes were illuminated for 12 hours a day from a warm white strip light attached to the room ceiling. This illumination resulted in approximately 100lx inside the cockroach boxes. The bottom of the boxes was ﬁlled with wood shavings (commercial rodent litter) and egg cartons were used as hideouts. Eight cultures were used: Most experiments (Molecular cloning, quantitative PCR, western blot experiments and running wheel assays using LD12:12) were executed on animals from four mass cultures as described: two with illumination from 8:00-20:00CET (”normal”) and two, in a seperate room, with illumination from 20:00-8:00CET (”inverse”). These cultures have been used for various experiments on R. maderae physiology and behavior [232, 249, 296, 299, 322, 341] before. Other cultures used and originating from the mass cultures were smaller colonies kept in standard aquarium glass tanks (50·30·30 cm or 40·25·25 cm, l·w·h) with the same decoration as the large mass cultures (LD6:18, LD18:6, normal and inverse illumination). These tanks were housed in wooden boxes which were lightproof, ventilated and illuminated from a 6W strip light. The temperature was 25-27°C and relative humidity ranged from 50% to 70%. The cockroaches were fed with dry dog food (Happydog, Interquell) and various vegetables (apples, potatoes and others) at least twice a week. Water was supplied ad libitum.

3.2 Molecular cloning

Tissue sampling

Brain tissue (”neuronal tissue”) and tissue from muscle and Malpighian tubules (”non-neuronal tissues”) of male R. maderae was sampled. Ten animals were dissected to gain mRNA that was subsequently used to produce adaptor ligated cDNA for RACE. Brain tissue contained the central brain and the optic lobes, without the photoreceptor cells and the neurohemal complex. Muscle tissue from the thoracic leg’s coxa and trochanter was used. Malpighian tubules were sampled as completely as possible without contamination from the hindgut. Muscle and Malpighian tubules were pooled right after dissection. The tissue was snap-frozen in liquid nitrogen and stored at −80°C until further use.

41 42 CHAPTER 3. MATERIALS AND METHODS

3.2.1 RNA extraction

Muscle and Malpighian tubules were homogenized in TRIzol® using liquid nitrogen, ceramic mortar and pestle, whereas brain tissue was homogenized using a motor driven plastic pestle in 1.5ml reaction tubes on ice. TRIzol® (Life technologies, Carlsbad, CA, USA) was used to extract total RNA. The protocol mainly followed the manufacturers instructions. RNA handling followed priniciple RNA handling guidelines.

Protocol: RNA extraction 1. use 1.5ml microcentrifuge tubes. 2. add 1ml TRIzol® to 5 brains and the approximately same amount of muscle or Malpighian tubules tissue. 3. incubate for 5min at room temperature. 4. add 0.2ml chloroform. 5. shake vigorously for 15s. 6. incubate 2-3 min at room temperature. 7. centrifuge 15min at 4°C / 12000xg. 8. transfer the supernatant to a new microcentrifuge tube. 9. add 0.5ml ethanol to the supernatant. 10. incubate 10min at room temperature. 11. centrifuge 10min at 4°C / 12000xg. 12. wash the pellet with 1ml 75% ethanol. 13. centrifuge 5min at 4°C / 7500xg. 14. remove the supernatant carefully and discard it. 15. air dry the pellet at room temperature.

16. dissolve in 50µl nuclease-free H2O.

3.2.2 cDNA synthesis for PCR using degenerate primer pairs In some cases, not adaptor ligated cDNA was used for PCR using degenerate primer pairs. Superscript™III (Life technologies, Carlsbad, CA, USA) reverse transcriptase was used in these cases according to the manufacturers instructions. The composition of the buffers was not revealed by the manufacturer. 3.2. MOLECULAR CLONING 43

Protocol: cDNA synthesis for PCR using degenerate primer pairs 1. use 0.2ml PCR tubes.

2. combine the following on ice:

• upto5µgtotalRNA • 50ng/µl random hexamer primers (Roche Hexanucleotide mix) • 1 µl annealing buffer

• to 8 µl nuclease-free H2O

3. incubate in a thermal cycler:

1: 5min - 65°C

4. incubate 1min on ice.

5. add 10µl 2x ﬁrst-strand reaction mix.

6. add 2 µl Superscript™III / RNaseOUT™ enzyme mix.

7. vortex brieﬂy.

8. centrifuge brieﬂy to collect the reaction on the bottom of the tube.

9. incubate in a thermal cycler:

1: 10min - 25°C 2: 50min - 50°C 3: 5min - 85°C 4: hold - 4°C

10. store at −20 °C.

3.2.3 Poly-A+ purification To gain highly purified poly-A+ RNA for RACE adaptor ligation, total RNA was purified using ® Oligotex (Qiagen N.V., Velden, Netherlands). This method uses dC10T30 oligonucleotides co- valently bound to the surface of polystyrene–latex particles (1.1 µm in diameter). Poly-A+ RNA is bound by these particles and can be washed and eluted in order to gain highly purified poly- A+ RNA. The purification protocol followed the manufacturers protocol. 44 CHAPTER 3. MATERIALS AND METHODS

Protocol: Poly-A+ puriﬁcation 1. use 1.5ml RNAse free microcentrifuge tubes.

2. add nuclease-free H2O to 500µl to the total RNA. 3. add 500µl buffer OBB and 30µl Oligotex® Suspension.

4. mix thoroughly by pipetting.

5. incubate for 3min at 70°C.

6. incubate at room temperature for 10min.

7. centrifuge 2min at 14000 - 18000xg.

8. remove the supernatant carefully and discard it.

9. resuspend the Oligotex®:mRNA complex pellet in 400µl buffer OW2.

10. pipette the resuspended Oligotex®:mRNA complex onto a small spin column placed in a 1.5ml microcentrifuge tube.

11. centrifuge 1min at 14000 - 18000xg.

12. transfer the column to a new RNAse free 1.5ml microcentrifuge tube.

13. add 400µl buffer OW2 to the column.

14. centrifuge 1min at 14000 - 18000xg.

15. discard the ﬂow-through.

16. transfer the column to a new RNAse free 1.5ml microcentrifuge tube.

17. add 20µl buffer OEB (70°C hot) to the column.

18. pipette up and down 4 times to resuspend the resin.

19. centrifuge 1min at 14000 - 18000xg.

20. repeat step 17-19 once.

21. store eluted puriﬁed poly-A+ RNA at -80 °C.

3.2.4 cDNA synthesis

The Marathon® cDNA ampliﬁcation kit (Clontech, Takara Bio Inc., Kyoto, Japan) was used for this step. The protocols follow the manual of the manufacturer. 3.2. MOLECULAR CLONING 45

First-strand synthesis For the first-strand synthesis, a modified lock-docking oligo(dT) primer [22] was used (5’– ® TTCTAGAATTCAGCGGCCGC(T)30VN–3’), as provided in the Marathon cDNA amplification kit (Clontech, Takara Bio Inc., Kyoto, Japan). This primer contains two degenerate nucleotides at the 3’ end, which position the primer at the beginning of the poly-A tail and thus eliminate 3’ heterogeneity.

Protocol: First-strand synthesis 1. use RNAse free 0.5ml microcentrifuge tubes.

2. combine the following:

• 1µg (1-4µl) puriﬁed poly-A+ RNA. • 1µl cDNA Synthesis Primer (10µM).

3. add nuclease-free H2O to 5 µl. 4. mix by pipetting and/or ﬂicking the tube.

5. centrifuge brieﬂy to collect the contents on the bottom of the tube.

6. incubate 2min at 70°C.

7. incubate 2min on ice.

8. centrifuge brieﬂy to collect the contents on the bottom of the tube.

9. add the following:

• 2 µl 5x ﬁrst-strand buffer. • 1µl dNTP mix (10mM). • 1µl AMV reverse transcriptase (20U/µl)

• 1 µl nuclease-free H2O. 10. mix by pipetting.

11. incubate 60min at 42°C.

12. incubate on ice until proceeding with second-strand synthesis.

Proceed with second-strand synthesis right after ﬁrst-strand synthesis.

Second-strand synthesis Second-strand synthesis followed the method published by Gubler and Hoffmann 1983 [90]. 46 CHAPTER 3. MATERIALS AND METHODS

Protocol: Second-strand synthesis 1. use the RNAse free 0.5ml microcentrifuge tube containing the reaction from the ﬁrst- strand synthesis.

2. add the following to the 10µl ﬁrst-strand reaction

• 48.4 µl nuclease-free H2O. • 16 µl 5x second-strand buffer. • 1.6µl dNTP mix (10mM). • 4µl 20x second-strand enzyme cocktail.

3. mix by pipetting.

4. centrifuge brieﬂy to collect the contents on the bottom of the tube.

5. incubate 90min at 16°C.

6. add 2µl (6U) T4 DNA Polymerase.

7. mix by pipetting.

8. incubate 45min at 16°C.

9. add 4µl EDTA/glycogen mix.

10. add 100µl phenol:chloroform:isoamyl alcohol (25:24:1).

11. mix thoroughly by vortexing.

12. centrifuge 10min at 17000xg.

13. pipette the aqueous layer to a new RNAse free 0.5ml microcentrifuge tube.

14. discard inter- and lower phase.

15. add 100µl chloroform:isoamyl alcohol (24:1) to the aqueous layer and mix thoroughly by vortexing.

16. centrifuge 10min at 17000xg.

17. pipette the aqueous phase to a new RNAse free 0.5ml microcentrifuge tube.

18. discard inter- and lower phase.

19. add half of the volumeof the aqueous phase ammonium acetate (4M) to the aqueous phase.

20. add 2.5 volumes (of the aqueous phase/ammonium acetate mixture) of 95% Ethanol (room temperature). 3.2. MOLECULAR CLONING 47

21. mix thoroughly by vortexing.

22. centrifuge 20min at 17000xg.

23. remove and discard the supernatant carefully.

24. add (overlay) 300µl 80%Ethanol to the pellet.

25. centrifuge 10min at 17000xg.

26. remove and discard the supernatant carefully.

27. air dry the pellet at room temperature (approximately 10min)

28. resolve the pellet in 10 µl nuclease-free H2O. 29. store at -20°C.

3.2.5 Adaptor ligation

Figure 3.1: Marathon cDNA adaptor and primer binding sites.

The Marathon® cDNA amplification kit (Clontech, Takara Bio Inc., Kyoto, Japan) was used for this step. Ligation of the Marathon Adaptor (Fig. 3.1) to the 5’ and 3’ ends of the purified poly-A+ RNA allows for amplification of complete 5’ as well as 3’ cDNA ends in subsequent RACE-PCR steps. Thus, full length cDNA sequences can be obtained with only a short sequence fragment of the target cDNA known.

Protocol: Adaptor ligation 1. use RNAse free 0.5ml microcentrifuge tubes.

2. combine the following:

• 5µl ds-cDNA. 48 CHAPTER 3. MATERIALS AND METHODS

• 2µl Marathon cDNA Adaptor (10µM). • 2µl 5x DNA ligation buffer. • 1µl T4 DNA ligase (400U/µl).

3. mix by vortexing.

4. centrifuge brieﬂy to collect the contents on the bottom of the tube.

5. incubate overnight at 16°C.

6. incubate 5min at 70°C.

7. dilute the reaction 1:50 and 1:250 with Tricine-EDTA buffer.

8. store at -80°C.

3.2.6 Primers Primers (Table 3.1) were designed using Primer 3 and Vector NTI. Degenerate primers (Table 3.2) were chosen by comparing amino acid sequences of the genes and choosing regions with high homology and a high content of amino acids encoded by as few codons as possible (like methionine and tryptophan).

Gene Forward Primer Reverse Primer period cctccgtggatggaagatgtgcat accagaggaaagttttgggctcgtagtg acccaagatctcgtgtacaggtatcagctg aaacctccattctgtgctcg agtagcaatgagccatttcctcagcctc atttcacccagcataaccga gtcgtgtggttttggtgtga aaggagaggtggcttgtacg cgagcacagaatggaggttt ggttggctggattggagtaa attcggttatgctgggtgaa cttccggttacaggacgttg gctcgtacaagccacctctc ttactccaatccagccaacc gcagccatgcctggtatgttgtacc ccaatacccgatgccagctc cagcctgccacagcattgtt timeless 1 tgcagttgcaaattggagcc tgcctcatcacttctactgttgc gcagtcacaacccaaatgcatcctacgg tgcccaacttgaattggaattggatcac ggatgcatttggttgtgactgc tcccacatgcatcattgtgaac cryptochrome 2 ctcttcatggtcagctgttatggcg atgatgttctggtcccgcaccc caaatatgtgtccagattccatggg atggctctggatcctcctcaaatgt

Table 3.1: RACE-PCR and PCR primers. 3.2. MOLECULAR CLONING 49

Gene Forward Primer Reverse Primer timeless 1 cayttyttytggctngtnac tcnccnccnayrtgcatcat ttytggctngtnacntaytt ccnayrtgcatcatngtraa cryptochrome 2 ggncarccngcngaygcn ccacatccacatnccngcrttnac acnttygargargayccngarccn raanacyttcatnccytcytccca

Table 3.2: Degenerate primers.

3.2.7 Initial PCR Since insects with a sequenced genome as well as insects with cloned circadian genes are mostly holometabolous insects or hemimetabolous insects not closely related to R. maderae, it was decided to use a degenerate primer strategy. An exception was period, where the known sequence of Periplaneta americana period was used to clone and sequence an initial cDNA fragment.

Protocol: Intial PCR 1. use PCR grade 0.2 or 0.5ml microcentrifuge tubes.

2. combine the following on ice:

• 1-5µl template. • 5µl 10x Advantage™ 2 PCR buffer. • 1 µl forward primer. • 1 µl reverse primer. • 1µl dNTP mix (10mM each). • PCR-grade water to 49 µl. • 1µl 50x Advantage™ 2 polymerase mix.

3. mix by pipetting and/or ﬂicking the tube.

4. centrifuge brieﬂy to collect the contents on the bottom of the tube.

5. incubate in a thermal cycler:

1. 3min - 94°C 2. 1min - 94°C 3. 1min - 37°C 4. 3min - 68°C 5. goto 2, repeat 3 times 6. 1min - 94°C 7. 1min - 40°C 8. 3min - 68°C 50 CHAPTER 3. MATERIALS AND METHODS

9. goto 2, repeat 32 times 10. 3min - 68°C 11. hold - 4°C

After the PCR, electrophoresis and subsequent steps were performed as described for RACE- PCR.

3.2.8 RACE-PCR Rapid amplification of cDNA ends (RACE) is a method to acquire full length sequences of a cDNA (mRNA respectively) of interest, including full 5’ and 3’ ends [306]. Marathon Adaptor ligated cDNA was used as template. The Advantage™ TAQ PCR system (enzyme mix and buffer, Clontech, Takara Bio Inc., Kyoto, Japan) was used for RACE-PCR reactions. Nested PCR was used to increase yield and specificity. Touch-down PCR and conventional three step PCR were employed. Two rounds of PCR were employed: In the first round, a 5’ or 3’ (depending on whether 5’ or 3’ RACE should be performed) genespecific primer (GSP) and adaptor primer (AP) 1 was used. After the PCR the reaction was diluted 1:1000 or 1:10.000 and used as template in the second, nested PCR, where a nested GSP and AP2 were used.

Protocol: RACE-PCR 1. use PCR grade 0.2 or 0.5ml microcentrifuge tubes.

2. combine the following on ice:

• 1-5µl template1. • 5µl 10x Advantage™ 2 PCR buffer. • 1µlAP1or2(10µM)2. • 1µlGSP1or2(10µM)2. • 1µl dNTP mix (10mM each). • PCR-grade water to 49µl. • 1µl 50x Advantage™ 2 polymerase mix.

1 PCR: 1-5µl 1:50 or 1:250 diluted Adaptor ligated cDNA, Nested PCR: 1µl 1:1000-1:10000 diluted PCR reaction 2 PCR: AP1, GSP1, Nested PCR: AP2, GSP2

3. mix by pipetting and/or ﬂicking the tube.

4. centrifuge brieﬂy to collect the contents on the bottom of the tube.

5. incubate in a thermal cycler: 3.2. MOLECULAR CLONING 51

1. 2min - 95°C 2. 1min - 95°C 3. 1min - 60°C 4. 1min - 68°C 5. goto 2, repeat 35 times 6. 10min - 68°C 7. hold - 4°C

Step 3 (annealing) temperature can vary depending on GSP/AP combination; Step 4 (elongation) time depends on the expected amplicon length: 1min/kb; Step 5: 25-35 cycles were used.

6. store at -20°C or continue with TA cloning.

3.2.9 Electrophoresis Biorad Sub-Cell and Mini Sub-Cell GT systems were used for PCR product analysis and sepa- ration. Gel loading buffer was added to the PCR products (5 to 20µl) and separated at 100V on a 1 % agarose gel in TAE buffer.

3.2.10 Gel extraction After Electrophoresis, bands of interest were excised with a clean scalpel and DNA from the gel piece was puriﬁed using Wizard™SV Gel and PCR Clean-Up System (Promega) or High Pure PCR Product Puriﬁcation Kit (Roche).

Gel extraction using the Wizard® SV Gel and PCR Clean-Up System (Promega) The Wizard® SV Gel and PCR Clean-Up System (Promega, Fitchburg, WI, USA) was used according to Promega technical bulletin TB308.

Protocol: Gel extraction using the Wizard® SV Gel and PCR Clean-Up System (Promega) 1. use PCR-grade 1.5ml microcentrifuge tubes. Weight tubes before starting.

2. excise the band of interest and place the gel piece in the previously weighed microcentrifuge tube.

3. weight the microcentrifuge tube and subtract the weight of the empty tube.

4. add 10µl Membrane Binding Solution per 10mg of gel piece.

5. centrifuge brieﬂy to collect the contents on the bottom of the tube.

6. incubate 10min at 50-65°C. Vortex the gel slice every few minutes during incubation. 52 CHAPTER 3. MATERIALS AND METHODS

7. transfer the dissolved gel mixture to a SV Minicolumn assembly.

8. incubate 1min at room temperature.

9. centrifuge 1min at 16000xg.

10. discard the ﬂow-through.

11. add 700µl Membrane Wash Solution to the SV Minicolumn.

12. centrifuge 1min at 16000xg.

13. discard the ﬂow-through.

14. add 500µl Membrane Wash Solution to the SV Minicolumn.

15. centrifuge 5min at 16000xg.

16. discard the ﬂow-through.

17. centrifuge 1min at 16000xg.

18. discard the ﬂow-through.

19. transfer the SV Minicolumn to a new PCR-grade 1.5ml microcentrifuge tube.

20. incubate 1min at room temperature.

21. add 30 µl nuclease-free H2O (pH7.0) directly to the center of the silica membrane, without touching it.

22. incubate 2min at room temperature.

23. centrifuge 1min at 16000xg.

24. discard the SV Minicolumn.

25. store at 4°C or -20°C.

26. continue with the ligation right after gel extraction if possible to avoid nicking of the de- oxyadenosyl overhang.

Gel extraction using the High Pure PCR Product Purification Kit (Roche) The High Pure PCR Product Purification Kit (Roche, Rotkreuz, Switzerland) was used with a slightly modified protocol (provided with the kit). 3.2. MOLECULAR CLONING 53

Protocol: Gel extraction using the High Pure PCR Product Puriﬁcation Kit (Roche) 1. excise the band of interest and place the gel piece in the previously weighed microcentrifuge tube.

2. weight the microcentrifuge tube and subtract the weight of the empty tube.

3. add 30µl Binding Buffer per 10mg of gel piece.

4. centrifuge brieﬂy to collect the contents on the bottom of the tube.

5. incubate 10min at 56°C. Vortex the gel slice every few minutes during incubation.

6. add 15µl isopropanol for every 10mg agarose gel piece.

7. mix thoroughly by vortexing.

8. transfer the dissolved gel mixture to a High Pure Filter Tube/Collection tube assembly. Do not exceed 700µl total volume. Split the volume and use seperate ﬁlter tubes if exceeding 700 µl.

9. centrifuge 1min at 14000xg - 18000xg.

10. discard the ﬂow-through.

11. add 500µl Wash Buffer to the High Pure Filter Tube.

12. centrifuge 1min at 14000xg - 18000xg.

13. discard the ﬂow-through.

14. add 200µl Wash Buffer to the High Pure Filter Tube.

15. centrifuge 1min at 14000xg - 18000xg.

16. discard the ﬂow-through.

17. transfer the High Pure Filter Tube to a new nuclease-free 1.5 ml microcentrifuge tube.

18. incubate 1min at room temperature.

19. add 30 µl nuclease-free H2O (pH7.0) directly to the center of the silica membrane, without touching it.

20. incubate 2min at room temperature.

21. centrifuge 1min at 16000xg.

22. discard the High Pure Filter Tube. 54 CHAPTER 3. MATERIALS AND METHODS

23. store at 4°C or -20°C.

24. continue with the ligation right after gel extraction if possible to avoid nicking of the de- oxyadenosyl overhang.

3.2.11 Cloning Ligation

Figure 3.2: pGEM®-T and pGEM®-T Easy vectors. Numbers in brackets behind restriction enzymes indicate the position of the restriction site within the MCS. Abbreviations: MCS, multiple cloning site; mut., mutated; DP6, SP6 RNA polymerase promoter; T7, T7 RNA polymerase promoter;

After gel extraction, PCR products were ligated into pGEM®-T or pGEM®-T Easy vectors (Fig. 3.2). PCR using the Advantage™ 2 Polymerase Mix results in products with terminal 3’ de- oxyadenosyl residue overhang, due to the Taq polymerase (Titanium TaqDNA pol.) in the enzyme mix. These overhangs can be used to clone the PCR product into linearized T vectors that carry terminal 3’ thymidine residue overhangs [108, 175], which is more efﬁcient than blunt-end cloning [175]. The pGEM®-T and pGEM®-T Easy Vector Systems kit was used for this step.

Protocol: Ligation 1. use nuclease-free 0.2 ml microcentrifuge tubes.

2. mix 2x Rapid Ligation Buffer, T4 DNA ligase thoroughly by vortexing.

3. combine the following on ice:

• 5µl 2x Rapid Ligation Buffer, T4 DNA ligase. • 1 µl pGEM®-T and pGEM®-T Easy vector (50 ng). • 3µl PCR product. 3.2. MOLECULAR CLONING 55

• 1µl T4 DNA ligase (3 WeissU/µl).

4. mix by pipetting.

5. incubate overnight at 4 °C in a thermal cycler.

Transformation

After ligation of the PCR product into pGEM®-T or pGEM®-T Easy vectors, the vectors were transformed into Subcloning Efﬁciency™ DH5α™ Competent E. coli (Life technologies, Carls- bad, CA, USA). These chemical competent cells can be used for β-galactosidase α-comple- mentation. The cells were thawed on ice and split into 50µl aliquots in nuclease-free 1.5ml microcentrifuge tubes directly after arrival. The single reaction aliquots were then stored at - 80 °C.

Protocol: Transformation 1. apply principle rules for sterile working.

2. thaw a 50µl Subcloning Efﬁciency™ DH5α™ Competent E. coli aliquot on ice.

3. add 1-5µl of the ligation reaction by gentle pipetting. Do not mix by pipetting up and down or vortexing!

4. incubate 20min on ice.

5. incubate 30s at -42°C in a water bath.

6. incubate 2min on ice.

7. add 350 µl pre-warmed (37 °C) SOC medium.

8. incubate 90min at 37°C and 225rpm.

9. centrifuge 3min at 8000xg.

10. remove the supernatant by decanting.

11. resolve the pellet in the residual SOC medium by gentle pipetting.

12. clean the bench with 80% ethanol, work next to a bunsen burner.

13. plate the cell solution onto LB plates containing ampicillin and X-gal.

14. incubate plates overnight at 37°C. 56 CHAPTER 3. MATERIALS AND METHODS

Overnight cultures After transformation, clones probably containing an insert (white, no β-galactosidase activity) and negative controls (blue, β-galactosidase activity) were picked and incubated in LB medium overnight.

Protocol: Overnight cultures 1. clean the bench with 80% ethanol, work next to a bunsen burner. Apply principle rules for sterile working.

2. pick positive colonies and a negative control colony with a sterile 200µl pipette tip.

3. eject the tip into a sterile glass or plastic tube ﬁlled with 3ml sterile LB medium.

4. close the tube with a metal or plastic cap to avoid excessive evaporation.

5. incubate overnight at 37°C, 225rpm.

Plasmid preparation Small scale plasmid preparations (”Miniprep”) were performed for plasmid isolation and purification. The PureYield™ Plasmid Miniprep System (Promega, Fitchburg, WI, USA) was used for plasmid preparations. This kit uses an endotoxin removal step in addition to standard silica membrane column purification. The kit was used according to the manufacturers instructions. Unfortunatly, buffer compositions are not given by the manufacturer. However, the principle of endotoxin removal is covered by U.S. Patent 6,194,562 and is described with possible buffer compositions there. After purification, plasmids were digested using Escherichia coli strain R nuclease I (EcoRI) (Fermentas Thermo Scientific): 1µl of the purified plasmid was digested with 0.5µl EcoRI (5U) in 10 µl EcoRI buffer (50 mM TRIS-HCl (pH 7.5), 10 mM MgCl2, 100mM NaCl, 0.02% Tri- ton X-100 and 0.1mg/ml bovine serum albumin (BSA)) at 37°C for 2h. The digested plasmids (5 µl) were subsequently separated as described in subsection 3.2.9. Purified plasmids containing inserts of the expected size were then used for sequencing.

Protocol: Plasmid preparation 1. chill the Neutralization Solution to 4-8°C before you begin.

2. use nuclease-free 1.5 ml microcentrifuge tubes.

3. add approximatly 1.5ml of the overnight culture to a 1.5ml microcentrifuge cup.

4. centrifuge 3min at 14000xg - 18000xg.

5. remove the supernatant by decanting. 3.2. MOLECULAR CLONING 57

6. decant the remaining overnight culture into the 1.5ml microcentrifuge cup.

7. centrifuge 3min at 14000xg - 18000xg.

8. remove the supernatant by decanting.

9. add nuclease-free H2O to approximatly 600µl.

10. resuspend the pellet.

11. add 100 µl Cell Lysis Buffer.

12. mix by inverting the microcentrifuge tube six times. If the solutions color changed from opaque to clear blue, continue with the next step right away (Do not wait longer than 2min, otherwise denaturing of plasmid DNA may occur).

13. add 350µl Neutralization Solution (chilled to 4-8°C).

14. mix thoroughly by inverting the microcentrifuge tube approximately four times. If the solution turns yellow and an precipitate forms, continue with the next step.

15. centrifuge 3min at 14000xg - 18000xg.

16. transfer the supernatant carefully to a PureYield™ Minicolumn assembly by pipetting.

17. discard the ﬂow-through.

18. add 200µl Endotoxin Removal Wash to the PureYield™ Minicolumn.

19. centrifuge 1min at 14000xg - 18000xg.

20. add 400µl Column Wash Solution to the PureYield™ Minicolumn.

21. centrifuge 1min at 14000xg - 18000xg.

22. transfer the PureYield™ Minicolumn to a new nuclease-free 1.5 ml microcentrifuge tube.

23. add 30 µl nuclease-free double-distilled water (ddH2O, pH7.0) to the PureYield™ Mini- column.

24. incubate 1min at room temperature.

25. centrifuge 1min at 14000xg - 18000xg.

26. store at -20°C. 58 CHAPTER 3. MATERIALS AND METHODS

3.2.12 Sequencing and sequence analysis

Puriﬁed plasmids were sent to analysis dissolved in ddH2O. Sanger sequencing was done by Euroﬁns MWG Operon (Ebersberg, Germany) or JenaGen GmbH (Jena, Germany). Sequences were assembled using Vector NTI (Life technologies, Carlsbad, CA, USA), alignments were done using Clustal Omega.

3.2.13 Motif prediction CDD (Conserved Domain Database, http://www.ncbi.nlm.nih.gov/cdd/)[174] and SMART (Sim- ple Modular Architecture Research Tool, http://smart.embl-heidelberg.de/) [298], [160] were used to predict putative functional domains. To predict nuclear localization signals, NLStradamus [207] was used (settings: 4state and 2state HMM static, posterior cutoff 0.6). In alignments, domains are shown as detected for R. maderae proteins unless noted otherwise.

3.2.14 Phylogenetic analysis For phylogenetic analysis, MEGA5 [324] was used. Sequences were aligned with CLUSTAL W, which is included in MEGA5. The neighbor-joining method [274] was used to infer evolutionary history. The consensus tree was calculated from 1000 bootstrap replicates [72] and evolutionary distances were computed using the Poisson correction method [370]. Gaps and missing sequences were eliminated for the analysis, resulting in 360 positions in PER and 421 in CRY2 being used. The accession numbers of the insect core feedback loop genes used for comparison are indicated in Table A.2. 3.2. MOLECULAR CLONING 59

3.2.15 Solutions

Oligotex suspension 50 x Advantage™ 2 Polymerase Mix Component Final Component Stock Final Oligotex particles 10 % w/v Titanium n.s. n.s. Titanium n.s. TaqDNA pol. TaqDNA pol. Proofreadingpol. n.s. n.s. TRIS-HCl 10 mM TaqStart antibody 1.1µg/µl 0.022µg/µl NaCl 500 mM Glycerol 50%v/v 1%v/v EDTA 1 mM TRIS-HCL 15 mM 0.3 mM SDS 0.1 % w/v (pH 8.0) Sodium azide 0.1 % w/v KCL(pH8.0) 75mM 1.5mM - pH 7.5. EDTA 50 µM 1 µM - provided by the manufacturer. - provided by the manufacturer.

Wash Buffer OW2 10 x Advantage™ 2 PCR Buffer Component Final Component Stock Final TRIS-HCl 10 mM Tricine-KOH 400 mM 40 mM NaCl 150 mM (pH 8.7) EDTA 1 mM KOAc 150 mM 15 mM - pH 7.5. Mg(OAc)2 35µg/µl 3.5µg/µl - provided by the manufacturer. BSA 37.5µg/ml 3.75µg/ml Tween20 0.05%v/v 0.3mM Elution Buffer OEB Nonidet-P40 0.05%v/v 0.3mM Component Final TRIS-HCl 5 mM - provided by the manufacturer. - pH 7.5. 6 x Gel loading buffer (type III [275]) - provided by the manufacturer. Component Stock Final Bromophenol 0.25%w/v 0.042%w/v Buffer OBB Component Final blue TRIS-HCl 20 mM XylenecyanolFF 0.25%w/v 0.042%w/v NaCl 1 M Glycerol 30%v/v 5%w/v EDTA 2 mM - in nuclease-free H2O. SDS 0.2 % w/v - pH 7.5. - provided by the manufacturer. 60 CHAPTER 3. MATERIALS AND METHODS

50 x TRIS-acetate-EDTA (TAE) Buffer Washing Buffer Component Stock Final Component Final TRIS-acetate 2 M 40 mM NaCl 20 mM EDTA 50mM 1 mM TRIS-HCL 2 mM For 1 l: Ethanol 40 % v/v - Prepare an EDTA stock solution (0.5M, pH - pH7.5. Provided by the manufacturer 8.0). without ethanol. - Dissolve 242g TRIS base in 750ml, add 57.1 ml glacial acetic acid. 2x Rapid Ligation Buffer, T4 DNA ligase - Combine 100ml EDTA stock solution with Component Stock Final the TRIS-acetate solution. TRIS-HCl 60 mM 30 mM (pH 7.8) - add ddH2O to 1 l. MgCl2 20 mM 10 mM - Dilute 1:50 with ddH2O before use. DTT 20 mM 10 mM Membrane Binding Solution ATP 2 mM 1 mM Component Final Polyethylene gly- 10%v/v 5%v/v Guanidine isoth- 4.5 M col iocyanate (MW8000, ACS Potassium acetate 0.5 M Grade) (pH 5.0) - Provided by the manufacturer. Sensi- - provided by the manufacturer. tive to repeated freeze-thaw cycles due to its ATP content. Split to single-use aliquots and Membrane Wash Solution store at -20 °C. Component Final Potassium acetate 10 mM SOC Medium (pH 5.0) Component Final Ethanol 80 % w/v Tryptone 20 g/l EDTA (pH 8.0) 16.7 µM Yeast extract 5 g/l - provided by the manufacturer without NaCl 0.5 g/l ethanol. KCl 2.5 mM MgCl2 10 mM Binding Buffer Glucose - 20mM Component Final - in ddH2O,pH7.0(NaOH). Guanidine isoth- 3 M - For 1l: Dissolve tryptone, yeast and NaCl iocyanate in 950 ml ddH2O. Add 10ml KCl in ddH2O TRIS-HCL 10 mM (250mM). Adjust to pH7.0 with NaOH and Ethanol 5 % v/v autoclave. Add 5ml MgCl2 in ddH2O (2 M, - pH 6.6. Provided by the manufacturer. sterile) and 20ml glucose (1M, sterile). 3.2. MOLECULAR CLONING 61

Luria-Bertani (LB) Medium - Agar plates Luria-Bertani (LB) Medium Component Final Component Final Tryptone - 10g/l Tryptone 10 g/l Yeast Extract - 5 g/l Yeast Extract 5 g/l NaCl - 10 g/l NaCl 10 g/l Agar - 15 g/l - in ddH2O,pH7.0(NaOH). Ampicillin - - For 1l: Dissolve tryptone, yeast and - in ddH2O,pH7.0(NaOH). NaCl in 900ml ddH2O. Adjust to pH7.0 with - For 1l: Dissolve tryptone, yeast and NaCl in NaOH. Adjust the volume to 1l and autoclave. 900 ml ddH2O. Adjust to pH7.0 with NaOH. Add the agar, adjust the volume to 1l and autoclave. 62 CHAPTER 3. MATERIALS AND METHODS

3.3 Quantitative real time polymerase chain reaction

3.3.1 Tissue sampling Whole brains and Malpighian tubules were dissected as described before in ”Molecular cloning: tissue sampling”. Residual fat body and tracheae were removed carefully. Accessory medullae and associated neurons (AMae) were dissected by cutting a triangle shaped piece of tissue from the brain (Fig. 3.3). Tissues were sampled every 4h starting at ZT0, from 30mins before to 30minsafter the time pointto be sampled. In the case of time points in the dark-phase and constant darkness, animals were dissected under red light (λ >610 nm) from a single high-power LED. For LD12:12 experiments, ten animals were dissected per time point and pool. In the case of LD6:18 and LD18:6 experiments, three animals were dissected per time point and pool. Tissue samples were snap-frozen in liquid nitrogen right Figure 3.3: Illustration of the piece of tis- after dissection and stored at -80°C until further use. sue that was excised to extract the AMe and associated neurons. Abbreviations: AMe, accessory medulla; La, lamina; 3.3.2 RNA extraction Lo, lobula; Me, medulla. Red: PDF- immunoreactive somata and projections. The RNeasy® Mini Kit (Qiagen N.V., Velden, Nether- Scalebar: 100 µM. lands) was used to extract total RNA in combination with on-column DNA digestion (RNase-Free DNase Set, Qiagen N.V., Velden, Netherlands). DNA digestion was appropriate due to the high sensitivity and susceptibility to genomic DNA contamination of qPCR using hydrolysis probes. Also, qPCR primers could not be designed to span exon borders due to the lack of genomic sequence data. The exact buffer compositions of the kit are being kept conﬁdential by Qiagen. However, some details are given: Buffer RLT contains high amounts of guanidine isothiocycanate, buffer RW1 contains a guanidine salt and ethanol, buffer RPE is described as a ”mild washing buffer” containing ethanol (ofﬁcial Qiagen website, FAQs). RNA handling followed principle RNA handling guidelines.

Protocol: RNA extraction Do not process more than six samples at once.

1. add 4 volumes of ethanol (96-100%, molecular biology grade) to buffer RPE before use.

2. add 10 µl β-mercaptoethanol per 1 ml to buffer RLT before use. 3.3. QUANTITATIVE REAL TIME POLYMERASE CHAIN REACTION 63

[152] 3. dissolve 1500KunitzU DNase 1 in 550 µl nuclease-free H2O by injecting the water into the DNase vial using a RNase-free needle and syringe.

4. mix the vial containing the DNAse by inverting. Do not vortex!

5. transfer tissue samples in 1.5 ml nuclease-free microcentrifuge tubes from -80 °C to ice.

6. add 100µl buffer RLT to AMe tissue, 150µl buffer RLT to brain and Malpighian tubules tissue.

7. homogenize tissue on ice using a motor driven plastic pestle.

8. add 250µl buffer RLT to AMe tissue, 450µl buffer RLT to brain and Malpighian tubules.

9. pass 5 times through a 20gauge needle ﬁtted to an RNase-free syringe.

10. if processing ten brains or Malpighian tubules from ten animals:

• pipette 300µl (brains) or 100µl (Malpighian tubules) of the homogenate into a new nuclease-free 1.5 ml microcentrifuge tube. • add 300µl (brains) or 500µl (Malpighian tubules) buffer RLT to the homogenate. • store the remaining homogenate at -80°C.

11. centrifuge 3min at 14000xg - 18000xg.

12. pipette the supernatant into a new 1.5ml nuclease-free microcentrifuge tube.

13. add one volume 70% ethanol.

14. mix by pipetting.

15. transfer up to 700µl including any precipitate to an RNeasy spin column / collection tube assembly.

16. centrifuge 1min at 8000xg.

17. discard the ﬂow-through.

18. if the volume of step 15 exceeded 700µl, transfer the remaining volume to the RNeasy spin column and repeat steps 16-17.

19. add 350µl buffer RW1 to the RNeasy spin column.

20. centrifuge 1min at 8000xg.

21. combine the following on ice (”DNAse1 incubation mix”):

• 10µl/sample DNase1 stock solution. 64 CHAPTER 3. MATERIALS AND METHODS

• 70µl/sample buffer RDD.

22. mix the DNAse1 incubation mix by gently inverting the tube.

23. pipette 80µl of the DNAse1 incubation mix directly onto the silica membrane of the RNeasy spin column, without touching it with the pipette tip.

24. incubate 15min at room temperature.

25. add 350µl buffer RW1 to the RNeasy spin column.

26. centrifuge 1min at 8000xg.

27. discard the ﬂow-through.

28. add 500µl buffer RPE to the RNeasy spin column.

29. centrifuge 1min at 8000xg.

30. discard the ﬂow-through.

31. repeat steps 22-24.

32. transfer the RNeasy spin column to a new collection tube.

33. centrifuge 1min at 14000xg - 18000xg.

34. transfer the RNeasy spin column to a new 1.5ml nuclease-free microcentrifuge tube.

35. add 30 µl nuclease-free H2O directly onto the silica membrane of the RNeasy spin column, without touching it with the pipette tip.

36. centrifuge 1min at 8000xg.

37. immediatly place the microcentrifuge tube containing the eluted RNA on ice.

38. determine RNA quality and concentration by measuring a full absorption spectrum.

39. store RNA at -80°C or continue with cDNA synthesis right away.

3.3.3 cDNA synthesis Random hexamer primers (Hexanucleotide mix, Roche, Rotkreuz, Switzerland) were used to prime ﬁrst-strand synthesis. 3.3. QUANTITATIVE REAL TIME POLYMERASE CHAIN REACTION 65

Protocol: cDNA synthesis 1. prepare the master mix (multiply by number of reactions plus one) in a 1.5ml nuclease-free microcentrifuge tube on ice:

• 2 µl 10x buffer RT • 2µl dNTP mix (5mM each) • 1µl random hexamer primer • 1 µl Sensiscript reverse transcriptase

• 9 µl nuclease-free H2O 2. pipette 15µl of the master mix per reaction into separate 0.2ml nuclease-free PCR tube.

3. add 5µl template RNA.

4. mix thoroughly by pipetting and ﬂicking the tube.

5. centrifuge brieﬂy to collect the contents on the bottom of the tube.

6. incubate in a thermal cycler:

1. 90 min - 37°C 2. hold - 4°C

7. store cDNA at 4°C or -20°C.

3.3.4 qPCR qPCR was performed using hydrolysis probes. The principle of hydrolysis probes uses the 5‘- 3‘-exonuclease activity of the Taq polymerase to hydrolyze a dual labeled probe during primer extension [107]. In detail, the probe consists of 13-16 nucleotides with a reporter dye bound to the 5’ end and a quencher bound to the 3’ end. In this study, fluorescein amidite (FAM) was used as reporter dye, and a minor groove binding non-fluorescent quencher (MGB-NFQ) was used as quencher. In contrast to a fluorescent quencher, which re-emits the absorbed energy as light, a non-fluorescent or ”dark” quencher dissipates the energy as heat. During PCR, the fluorescence of the reporter dye is measured at the end of each PCR cycle. If the probe is in- tact, Förster resonance energy transfer (FRET) occurs and the fluorescence of the FAM reporter dye is being quenched. In case the probe binds to the template between the two primers and primer elongation occurs, the probe is hydrolyzed by the 5‘-3‘-exonuclease of the Taq polymerase resulting in the reporter dye being released from the probe. Subsequently, a fluorescence signal occurs after illumination with the excitation wavelength of the dye, which increases during each cycle in direct correlation to the amplificate. R. maderae actin was used as reference to calculate relative changes in mRNA abundance of the target gene. The R. maderae actin sequence that was cloned was too small to be submitted to GeneBank. It is shown in Fig. A.1. 66 CHAPTER 3. MATERIALS AND METHODS

TaqMan® probes (Life technologies Carlsbad, CA, USA) were used (Table 3.3). To compensate for not PCR based fluorescence signals (pipetting errors, instrument sensor fluctuations etc.), a reference dye (Carboxy-X-rhodamine = ROX™) is included in the master mix to measure the background fluorescence in each sample and PCR cycle. Applied Biosystems 7300 and 7500 Fast Real-Time PCR Systems (Life technologies Carlsbad, CA, USA) were used in combination with MicroAmp™ Optical 96-Well Reaction Plates (Life technologies, Carlsbad, CA, USA) and Optical Adhesive Covers (Life technologies, Carlsbad, CA, USA).

Protocol: qPCR 1. use 96-well plates and adhesive covers suited for the qPCR cycler of choice only.

2. dilute cDNA 1:3.5 with nuclease-free H2O (20 µl cDNA + 70 µl nuclease-free H2O).

3. prepare the master mix (total volume = reaction volume · ((samples · 3) + 2)) for each gene in a 1.5 ml nuclease-free microcentrifuge tube on ice:

• 10 µl reactionvol. TaqMan® Universal PCR Master Mix, No AmpErase® UNG. • 1µl reactionvol. probe/primer mix.

• 4 µl reactionvol. nuclease-free H2O.

4. pipette the master mix into the well plates. Pipette the complete plate at once, use the same pipette for all wells. Pipette the volume on the side of the well and avoid introducing bubbles.

5. pipette the cDNA template into the wells as before the master mix.

6. seal the wellplate with an adhesive cover.

7. spin down residual ﬂuid on the well walls using a special plate centrifuge, or very gently tap the plate on the lab bench top to collect the reaction volume on the bottom of the well.

8. load the qPCR cycler with the 96-well plate.

9. use the following program (Applied Biosystems qPCR cyclers)

1. 10min - 95 °C 2. 15s - 95 °C 3. 1min - 60 °C 4. goto 2, repeat 40 times 3.3. QUANTITATIVE REAL TIME POLYMERASE CHAIN REACTION 67

Gene Forward Primer Reverse Primer Probe act gcgtctcgcatacagtaccaattta cggccagccagatcca ccccatgccatcctac per tggctccagtaacagaaaatgct gctgtgaccactcccactt ccactgccgcttccgt tim1 ctcatttgagcagagaagatagagaaca cagcagaataaacagacacttcaaatcc ctgtggctccaatttg cry2 tcgaagaagacccagaaccatttg gttctttgcacatcgccatgat ctggtcccgcacccgt

Table 3.3: qPCR probes and primers

3.3.5 Data analysis Data analysis was conducted by the qPCR cycler (Applied Biosystems 7300 and 7500 Fast Real- Time PCR Systems) up to the Ct calculations. Subsequent ratio calculations and plotting was done with Matlab (R2012b, Mathworks, Natick, MA, USA). In detail, first the reporter dye (FAM) fluorescence signal of the hydrolyzed probe and the background fluorescence signal (ROX™) is measured at the end of each cycle. To normalize the data, the reporter dye signal is then divided by the background fluorescence:

(1)

Rn = FAM/ROX

Rn: normalized reporter; FAM: FAM (or another reporter dye) ﬂuorescence signal; ROX: ROX™ ﬂuorescence signal.

Thereupon, the baseline (usually the arithmetic mean of the ﬁrst 10-15 cycles) is subtracted from the normalized ﬂuorescence signal:

(2)

∆Rn = Rn − baseline

∆Rn: normalized Rn; Rn: normalized reporter; baseline: arithmetic mean of Rn of the ﬁrst 10-15 cycles.

The baseline threshold (i.e. up to which cycle Rn is considered background) for the baseline is automatically set by the instrument. Subsequently, a threshold is being calculated (Ten times standard deviation above the mean baseline ﬂuorescence), which lies within the exponential phase of the PCR. The cycle, where ∆Rn crosses this threshold is called the crossing point (Cp) or threshold cycle (Ct). With the Ct value of the gene of interest and the reference gene, the relative amount of transcript of the gene of interest can then be calculated. There are different methods of calculating the relative amount of transcript. The easiest way is to use the ∆∆Ct method. In this method, the Ct of the gene of interest is subtracted from the Ct of the reference gene (=∆Ct). These values are calculated for each sample (including the control) and then the fold change is calculated with the following formula: 68 CHAPTER 3. MATERIALS AND METHODS

(3)

∆ ∆ f oldchange = 2− Ct (control)− Ct (sample)

In this method, optimal PCR reactions are assumed with the template doubling each cycle during the exponential phase of the PCR. Since this is not the case for most PCR reactions, a method that allows for PCR efficiency correction is more precise. If data for PCR efficiency is available or can be obtained, an efficiency corrected formula [235] should be used to calculate the relative change in expression levels. The efficiency was calculated by using four dilutions of cDNA (100, 10-1, 10-2, 10-3). The remaining cDNA of the previous experiments (LD12:12, AMe and brain without AMe) was combined across all timepoints. For each gene the Ct for AMe and brain without AMe samples was determined (except for rmTim1, where only brain without AMe mea- surements were taken in the 100 and 10-1 cDNA sample). The arithmetic mean of each dilution was calculated, and the slope of the linear regression line was determined for each gene. Then, −1/slope the PCR efficiency was calculated with the formula: E = 10 . If the Ct values and the efficiency are known, the relative expression ratio R can be calculated with the following formula [235]:

(4)

∆ Ct(goi)(control−sample) Egoi R = ∆ Ct(re f )(control−sample) Ere f

R: relative expression ratio; E: PCR efﬁciency; goi: gene of interest; ref: reference gene

Since a ”control” can not be denominated in circadian time series (because there is no ”treated” and ”untreated”, for example) the arithmetic mean across all time points was used as ”control”. To compare the daily mean of expression levels between different genes and within one gene between different light regimes, the percentage of actin expression was calculated with the following formula:

(5)

∑ZT20 Ct(re f )(ZT) Ct(goi)(ZT) ZT0 (Ere f )/(Egoi ) rel.expr = 100 · n

rel.expr: relative expression in % reference gene expression; E: PCR efﬁciency; goi: gene of interest; ref: reference gene; ZT: Zeitgeber time 3.3. QUANTITATIVE REAL TIME POLYMERASE CHAIN REACTION 69

Statistics One-way ANOVA was used to analyze differences between timepoints in the qPCR timeseries. If significant (α=0.05) differences were determined, a subsequent least significant difference (LSD) post-hoc test (α=0.05) was performed to determine which timepoints were significantly different from each other. One-way ANOVA was also used to determine significant differences between the daily mean of expression levels. 70 CHAPTER 3. MATERIALS AND METHODS

3.4 Antibody generation

Protein Host Antigen rmPER rabbit NH2-MEETATHNTKISDS-COOH (1) NH2-KSSTETPPSYNQLN-COOH (2) NH2-CRRETSATNTSQGSY-CONH2 (3) rmPER mouse NH2-CPQPPTTKRNKDKEH-CONH2 rmTIM1 rabbit NH2-CEQLETTALQQGNDLK-CONH2 rmCRY2 rabbit NH2-CEEDPEPFGRVRDQ-CONH2

Table 3.4: Peptides used for immunization.

Antipeptide antibodies directed against peptides derived from the rmPER, rmTIM1 and rmCRY2 amino acid sequences were custom generated by Coring Systems Diagnostix GmbH (Gernsheim a. Rhein, Germany). Two antibodies were raised against rmPER, one in rabbit and one in mouse. One antibody each was raised against rmTIM1 and rmCRY2 in rabbits. Immunization and keep- ing of the animals, blood collection and antibody purification was conducted by Coring Systems Diagnostix GmbH. For the generation of the anti rmPER (rabbit), anti rmTIM1 (rabbit) and anti rmCRY2 (rabbit) antibodies, two rabbits each, and for the anti rmPER (mouse) antibody four mice of the BALB/c strain were immunized. The peptides used for immunization were conjugated to a hemocyanine from Megathura crenulata (keyhole limpet hemocyanine, KLH), which was used as a carrier to amplify immunogenicity. A mixture of three peptides was used to gain an antibody specific for rmPER, one peptide each was used for rmTIM1 and rmCRY2 (Table 3.4). Rabbits were initially injected with 1ml of a 1:1 mixture of antigen solution and Freund’s complete adjuvant. Preimmune serum was taken from each animal. The animals immune reactions were boosted with 1ml of incomplete Freund’s adjuvant containing the antigen mixture 21, 42 and 62 days (anti rmCRY2 (rabbit) additionally on days 91 and 112) after the initial immunization. Blood was taken from each animal 53 and 74 (anti rmPER (rabbit) additionally 102, anti rmCRY2 (rabbit) additionally 102 and 123) days after the first immunization. The immunoreactivity was determined using enzyme linked im- munosorbent assays (ELISA, Table 3.5). The animals were sacrificed 88 days (anti rmTIM1), 116 (anti rmPER) and 138 (anti rmCRY2) days after the first immunization. The immuneserum of the rabbit with the highest immunoreactivity was chosen for immuno-affinity purification. The mice that were immunized with the rmPER peptide were injected with incompleted Fre- und’s adjuvant/antigen mixture on days 7, 14, 28 and 42 after initial immunization. Blood was collected before the first immunization and 53 days thereafter (Table 3.5), and the animals were sacrificed for the final bleed on day 67. Final bleed immunosera of all mice were pooled and used for immuno-affinity purification. Blood serum was immuno-affinity purified by the company using an immunogen-affinity resin with the peptides used for immunization. The peptide (in case of the anti rmPER (rabbit) all three peptides) was coupled to 1 ml activated sepharose beads and antisera were incubated with 3.4. ANTIBODY GENERATION 71 the beads for 1h at 37°C. After several washing steps, the purified antibodies were eluted in TRIS (0.1M) / glycine (pH7.8) buffer containing 0.02% sodium azide and antibody containing fractions were pooled. Final purified antibody concentrations were: anti rmPER (rabbit): 487µg/ml; anti rmCRY2 (rabbit): 87.3µg/ml; anti rmTIM1 (rabbit): 112.6µg/ml; anti rmPER (mouse): 4.6µg/ml. Sera and purified antibodies were aliquoted and stored at -80°C.

Target Host Day Titer Target Host Day Titer rmPER rabbitI 53 <500 (1) rmCRY2 rabbitI 53 2000 <500 (2) rabbitII 53 1000 <500 (3) rabbitI 74 1000 rabbitII 53 1000(1) rabbitII 74 4000 <500 (2) rabbitI 102 500 8000 (3) rabbitII 102 4000 rabbitI 74 500(1) rabbitI 123 500 <500 (2) rabbitII 123 4000 2000 (3) rabbitII 138 4000 rabbitII 74 1000(1) <500(2) 16000 (3) rabbit I 102 <1000 (1) <500(2) 8000 (3) rabbitII 102 <2000 (1) 500 (2) 64000 (3) rabbitII 116 2000(1) 500 (2) 16000 (3) rmPER mouseI 53 800 rmTIM1 rabbitI 53 <500 mouseII 53 800 rabbit II 53 <500 mouseIII 53 800 rabbit I 74 >64000 mouseIV 53 3200 rabbitII 74 16000 pool 67 3200 rabbit I 74 >64000 rabbitII 74 16000 rabbitI 88 16000

Table 3.5: Immunoreactivity of the host animals sera. Sera were diluted 1:500 to 1:64000 (rabbits, eight dilutions) or 1:200 to 1:25600 (mice, eight dilutions). Day means days after initial immunization, titer is the inverse of the highest dilution of serum with an optical density ≥1 at 450 nm in ELISAs. Experiments were conducted by Coring Systems Diagnostix GmbH. Sera used for immuno-afﬁnity puriﬁcation are indicated in bold. 72 CHAPTER 3. MATERIALS AND METHODS

3.5 Western blotting

3.5.1 Tissue sampling and homogenization As in qPCR experiments, six timepoints were chosen for dissection (unless stated otherwise in the respective ﬁgures), which were identical with the time points used in the qPCR experiments: ZT0, ZT4, ZT8, ZT12, ZT16 and ZT20. At each timepoint, ten R. maderae males were dissected and extracted tissue was pooled. Supraesophageal ganglia and Malpighian tubules tissues were sampled as described in subsection 3.3.1. Muscle tissue was taken from the thoracic leg’s coxa and terminal abdominal ganglia were sampled without adherent nerves. The tissue was homogenized in RIPA buffer and residual tissue debris was removed by centrifugation. Ho- mogenates were stored at -80 °C.

Protocol: Tissues sampling and homogenization 1. use 1.5ml microcentrifuge tubes.

2. dissect R. maderae tissue at the given time point (under red light λ >630nm if the time of dissection is in the dark phase), shock freeze tissue in liquid nitrogen.

3. store collected tissue at -80°C.

4. add 20µl RIPA buffer/supraesophageal ganglion, up to 200µl per cup.

5. use a motor driven pestle to homogenize the tissue, keep the tubes on ice.

6. pass the homogenate 5 times through a 23 gauge syringe.

7. centrifuge 10min, 17000xg, 4°C.

8. pipette the supernatant carefully into a new cup.

9. centrifuge 10min at 17000xg, 4°C.

10. pipette the supernatant into a new tube, store 5µl diluted 1:10 with RIPA buffer in a separate cup for the protein assay.

11. store the homogenate at -80°C until use.

3.5.2 Protein assay Pierce 660 nm The Pierce 660 nm protein assay (Thermo Scientific, Rockford, IL, USA) was used to determine the total protein concentration of the homogenates. The protocol was adapted to be used with a Nanodrop® ND 1000 (Thermo Scientific, Rockford, IL, USA) spectrophotometer. Since SDS is used in the homogenization buffer, addition of Ionic Detergent Compatibility Reagent (IDCR, Thermo Scientific, Rockford, IL, USA) to the assay solution according to the product’s protocol 3.5. WESTERN BLOTTING 73 was necessary. Before measuring the samples, a standard curve with the BSA standards that are included in the assay for each new 660 nm solution+IDCR, with at least three replicates per concentration was performed.

Protocol: protein assay Pierce 660 nm, standard curve 1. add 1g IDCR to 20ml 660nm reagent.

2. mix by vortexing.

3. run a standard curve, using the protocol mentioned below. Use the following BSA concentrations: 125µg/ml, 250µg/ml, 500µg/ml, 750µg/ml, 1000µg/ml, 1500µg/ml, 2000µg/ml.

4. store the 660nm solution (+IDCR) at room temperature.

Protocol: protein assay Pierce 660 nm 1. mix the 660nm solution (+IDCR) by vortexing before use.

2. combine 45µl 660nm solution + 3µl 1:10 diluted sample (660nm solution + 3µl RIPA as blank).

3. mix by vortexing.

4. incubate for 5min at room temperature in the dark.

5. mix by vortexing. Do not centrifuge.

6. measure 2µl.

7. repeat the assay at least two times with three replicates per sample.

3.5.3 Polyacrylamide gel electrophoresis

Mini-PROTEAN® II or Mini-PROTEAN® Tetra Cell (Bio-Rad) systems were used for sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) with a TRIS-glycine buffer system and gels of 0.75mm thickness. The concentration of polyacrylamide ranged from 5% to 10%, a 3% stacking gel was cast over the separating gel. Unless noted otherwise, 10µg total protein concentration per lane were used. Since loading equal volumes of homogenate when the same number of supraesophageal ganglia were sampled was more precise than loading the same total protein concentrations as determined by the Pierce 660nm assay, a ﬁxed volume of homogenate was used in later blots. Four different protein markers were used: Fermentas (Now part of Thermo Scientiﬁc) PageRuler Plus Prestained, Spectra Multicolor Broad Range Protein Ladder, PageRuler Unstained Protein Ladder and Invitrogen BenchMark™ Protein Ladder. 74 CHAPTER 3. MATERIALS AND METHODS

Protocol: Polyacrylamide gel electrophoresis

Separating Stacking 5% 7.5% 10% 12.5% 15% 3% Acrylamidestock 1,83ml 2,5ml 3,33ml 4,17ml 4,99ml 0.5ml 8x(4x)gelbuffer 1,25ml 1,25ml 1,25ml 1,25ml 1,25ml 1,25ml SDS(10%w/v) 100µl 100µl 100µl 100µl 100µl 50µl ddH2O 6,77ml 6.1ml 5.27ml 4.43ml 3.61ml 3.16ml APS 50µl 50µl 50µl 50µl 50µl 30µl TEMED 5µl 5µl 5µl 5µl 5µl 10µl

Table 3.6: Polyacrylamide gel component concentrations

1. clean the glass plates of the gel cassette (spacer: 0.75mm thickness) carefully with ddH2O, followed by 80% ethanol and again ddH2O. Dry with lint-free wipes. 2. assemble the minigel casting stand and the casting frame with the gel cassettes.

3. combine the following at room temperature in a beaker (volumes according to Table 3.6, for two minigels):

• acrylamide stock • 8x separating gel buffer • SDS (10%w/v)

• ddH2O 4. mix carefully.

5. add ammonium persulfate (APS) and Tetramethylethylendiamin (TEMED) (volumes according to Table 3.6).

6. mix carefully.

7. ﬁll the gel cassette with separating gel solution up to 2cm below the top of the inner glass plate (approx. 3.5ml). Make sure not to introduce any bubbles.

8. overlay the separating gel with ∼100 µl isopropanol.

9. incubate at room temperature for 30-45min. When polymerization is complete, two clearly separate phases will be visible: the polymerized gel on the bottom with a small phase of unpolymerized isopropanole/separating gel solution mixture on the top.

10. decant the upper phase with the isopropanole/unpolymerized separating gel mixture and remove residual ﬂuid using wipes. 3.5. WESTERN BLOTTING 75

11. ﬂush the surface of the separating gel carefully with ddH2O. 12. remove residual ﬂuid using wipes.

13. combine the following at room temperature in a beaker (Volumes according to Table 3.6, for two minigels:

• acrylamide stock • 4x stacking gel buffer • SDS (10%w/v)

• ddH2O 14. mix carefully.

15. add APS and TEMED (volumes according to Table 3.6).

16. ﬁll the gel cassette with the stacking gel solution and immediatly introduce the comb (10well comb, 0.75mm).

17. incubate at room temperature for 30-45min.

18. dilute homogenates 1:1 with sample buffer.

19. heat the sample buffer/homogenate mixture 5min at 95°C.

20. cool sample buffer/homogenate mixture on ice.

21. combine the gel cassettes with the electrode assembly and ﬁll with running buffer until ∼1mm above the upper edge of the inner glass plate.

22. pull the combs carefully.

23. spin down the sample.

24. vortex the sample brieﬂy.

25. load the gel with the appropriate volumes, use at least one size marker.

26. move electrode assembly to the mini tank carefully, ﬁll the mini tank with sample buffer up to the mark on the tank and close it.

27. apply 100V constant voltage and increase to 140V when the bromphenoleblueband reaches the separating gel.

28. switch off power as soon as the bromphenoleblue band leaves the separating gel.

29. disassemble the electrode assembly. 76 CHAPTER 3. MATERIALS AND METHODS

30. open the gel cassette carefully.

31. remove the stacking gel.

32. mark the upper left (i.e. the side of the ﬁrst lane) edge of the separating gel by removing a little piece of gel.

33. remove the separating gel from the gel cassette carefully and submerge in blotting buffer.

3.5.4 Blotting Two different types of blotting procedures were applied: semidry and wet blotting. In semidry blotting, layers of thick ﬁlterpaper are being used as buffer reservoir, in wet blotting the PAGE gel containing stack is completely submerged in buffer. Semidry blotting is faster and requires less material, however protein transfer proved to be lower and less reliable than with wet blotting. Nitrocellulose (0.45µm, Whatman/GE Healthcare) membrane was used for semidry blotting, polyvinylidene ﬂuoride membrane (PVDF, Roti®-PVDF, Carl-Roth GmbH + Co. KG, Karlsruhe, Germany) was used for wet blotting. Initially, SDS and methanol were used in the blotting buffer (semidry blotting). However, it turned out that neither SDS nor methanol in the buffer are necessary for satisfactory transfer of the analyzed proteins. Also, the methanol usually used to equilibrate the PVDF membrane was substituted with ethanol with equal results. Thus, the toxic and hazardous methanol was completely omitted from the blotting process.

Protocol: Semidry blotting 1. trim the extra thick blot papers (Biorad) and nitrocellulose or PVDF Membran to the approximate size of the polyacrylamide gel.

2. if using PVDF Membrane, wet the membrane in 100% ethanol (10min during equilibration of the other components).

3. equilibrate components for approximatly 10min in transfer buffer:

• Two extra thick blot papers per blot. • One membrane per blot (if using Nitrocellulose). • acrylamidegels of the preceding PAGE.

4. assemble the semidry blotting stack in the following order:

- Anode (cover) - Extra thick blot paper Equilibrated gel Membrane Extra thick blot paper 3.5. WESTERN BLOTTING 77

- Cathode (bottom) -

5. roll over the semidry blotting stack carefully using a 50ml centrifuge tube or a glass rod to remove residual air bubbles.

6. close the apparatus.

7. run at constant current: 1.5mA per square centimeter of gel (this applies to the footprint of the gel sandwich(es)= 80mA/minigel) for 60min.

8. if using nitrocellulose membrane, the protein transfer can be visualized after the blotting using Ponceau stain: incubate the blot with Ponceau stain for 1min on a shaker. Then wash the blot with ddH2O carefully until the background is clear.

Protocol: wet blotting 1. trim the gel-blotting-paper (Whatman 3MM, 0.34mm thickness) and PVDF membran to the approximate size of the ﬁberglass pads.

2. wet PVDF membrane in 100% ethanol (20min during equilibration of the other components).

3. equilibrate components for approximatly 20min in TRIS-glycine:

• ﬁberglass pads • Two Whatman ﬁlter-papers per blot • One PVDF membrane per blot • acrylamidegels of the preceding PAGE.

4. assemble the wet blotting stack in the following order (Gray side of the blot system on the bottom):

- Cathode - Fiberglass pad Whatman ﬁlterpaper Equilibrated membrane Equilibrated gel Whatman ﬁlterpaper Fiberglass pad - Anode -

5. roll over the wet blotting stack carefully using a 50ml centrifuge tube or a glass rod to remove residual air bubbles. 78 CHAPTER 3. MATERIALS AND METHODS

6. add a stir bar to the tank.

7. add the cooling unit to the tank.

8. ﬁll the tank with blotting buffer.

9. close the system.

10. place the blotting tank on a magnetic stirrer on medium speed to circulate the buffer without introducing bubbles.

11. run at constant voltage: 100V, 70min. Amperage should not exceed 350mA.

3.5.5 Detection After blotting, membranes were incubated in blocking solution containing low-fat dry milk powder to block unspecific binding. Subsequently, membranes were incubated with primary antibody solution containing the antibody against the protein of interest. The primary antibodies against rmPER, rmTIM1 and rmCRY2 are described in section ”Antibody Generation”. These antibodies were used in a dilution of 1:1000 and 1:10000, the dilution is given for each blot in the results section. The anti-TYR-TUBULIN antibody was a monoclonal antibody raised in mice and was used in a dilution of 1:10000. The rabbit anti-apPER antibody, a kind gift of Anthony Gotter, was an affinity purified antibody directed against the Antheraea pernyi PER-S peptide [279] and used in a dilution of 1:1000. As secondary antibodies, affinity purified Horseraddish peroxidase conjugated immunoglobulin G (IgG) antibodies raised in goat, directed against IgG heavy and light chain of the target species (rabbit and mouse, Jackson ImmunoResearch, West Grove, Pennsylvania, USA) were used. Freeze-dried secondary antibodies were rehydrated with ddH2O, aliquoted and stored at -80°C. Concentration of the antibody in stock solution was 1.5mg/ml in phosphate buffered saline (PBS, 0.01M sodium phosphate, 0.25M NaCl, pH7.6) containing 15mg/ml BSA (IgG- free, protease-free). Both secondary antibodies were used in a dilution of 1:5000 unless noted otherwise. Roti®-Lumin spray (Carl Roth GmbH + Co. KG, Karlsruhe, Germany) or SuperSignal West Pico (Thermo Scientific, Rockford, IL, USA) were used as chemiluminescent substrate according to the manufacturer’s specifications. The signal was detected using Amersham Hyperfilm ECL (GE Healthcare Bio-Sciences, Piscataway, NJ, USA).

Protocol: Detection 1. trim the blot with a clean scissor.

2. incubate in blocking solution either 2h at room temperature or at 4°C overnight on a shaker.

3. decant blocking solution. 3.5. WESTERN BLOTTING 79

4. incubate in primary antibody solution containing the primary antibody overnight at 4°C on a shaker.

5. wash three times shortly, then 5, 10, 30min with wash solution on a shaker at room temperature.

6. decant wash solution.

7. prepare secondary antibody solution: Use a HRP coupled antibody, directed against the primary antibody’s host species with a dilution of 1:5000, or try 1:30000 for a lower background.

8. incubate for 1 h at room temperature on a shaker.

9. wash three times shortly, then 5, 10, 30min with wash solution on a shaker at room temperature.

10. wash brieﬂy in ddH2O.

11. transfer the blot to a x-ray cassette.

12. use Roti®-Lumin spray according to the manufacturer’s manual (Or any other ECL).

13. cover the blot(s) with plastic ﬁlm (avoid air bubbles).

14. incubate for 5min.

15. expose to x-ray ﬁlm.

3.5.6 Preadsorption of the rabbit anti-rmPER antibody

To determine whether preadsoption with peptide 3 (NH2-CRRETSATNTSQGSY-CONH2) also abolishes rmPER signals in western blots, 3µl of the puriﬁed rabbit anti-rmPER antibody were incubated overnight in 3ml primary antibody solution containing 100µM or 10µM peptide 3 at 4°C. As positive control, the antibody was incubated the same way in primary antibody solution not containing peptide 3. The next day the preincubated primary antibody solution and control were used on semidry blotted membranes as described in subsection 3.5.5.

3.5.7 Stripping To allow for the detection of TUBULIN in cases where the protein of interest has an approximate molecular weight of TUBULIN (rmCRY2), the blot was stripped and reprobed against TUBULIN. 80 CHAPTER 3. MATERIALS AND METHODS

Protocol: Stripping 1. pre-heat the stripping buffer to 55°C in a water bath.

2. after detection, rinse the PVDF membrane shortly in ddH2O.

3. decant ddH2O. 4. incubate the membrane with stripping buffer for 15min at 55°C in a shaking water bath.

5. wash the membrane two times with wash solution for 5min each.

6. repeat steps 4 and 5 once.

7. incubate the membrane in blocking solution for 1h at room temperature.

8. proceed with the primary antibody incubation (see subsection ”Detection”). 3.5. WESTERN BLOTTING 81

3.5.8 Solutions

RIPA Buffer Separating gel buffer (8x concentrated) Component Final Component Final TRIS-HCl 50 mM TRIS 3 M NaCl 150 mM Triton X-100 or 1 % v/v - in ddH2O,pH8.8(HCl). Tergitol - For 100ml: Dissolve 36.3g TRIS in 80ml Sodium Deoxy- 1 % w/v ddH2O, adjust to pH8.8 (HCl), water to cholate 100ml, ﬁlter and store at 4°C. SDS 0.1 % w/v

Add 1 Roche Complete-Ultra tablet (+EDTA) Stacking gel buffer (4x concentrated) per 10ml RIPA, aliquot and store at -20°C. Component Final TRIS 500 mM

- in ddH2O,pH6.8(HCl). - For 100ml: Dissolve 6g TRIS in 80ml ddH2O, adjust to pH6.8 (HCl), water to 100ml, ﬁlter and store at 4°C.

Running Buffer (10x concentrated) Component Final Sample Buffer TRIS 250 mM Component Stock Final Glycine 1.92 M TRIS 500 mM 100 mM SDS 1 % v/v Glycerol 5 % v/v - in ddH2O Bromphenoleblue 1.5 mM - store at 4°C, dilute 1:10 before use. ß-Mercapto- 5 % v/v ethanol SDS 0.25 % Transfer Buffer w/v Component Final TRIS 48 mM - in ddH2O,pH6.81(HCl). - For 10 ml: Prepare a 500mM TRIS solution, SDS 0.037 % w/v adjust to pH6.81 (HCl). Mix 2ml TRIS Methanol 20 % v/v (pH6.81) solution, 2ml glycerol, 10mg - in ddH2O,pH9.2(glycine). Bromphenoleblue, 0.5ml ß-mercaptoethanol - Use fresh buffer every experiment. and 4ml 10% SDS in a 15ml centrifuge tube, - solve TRIS and SDS, adjust pH to 9.2 with water (ddH2O) to 10ml. Mix well, split into glycine, add methanol, water to 1l. 500µl aliquots and store at -20°C. - store at 4°C. 82 CHAPTER 3. MATERIALS AND METHODS

TRIS-glycine Buffer Primary antibody solution Component Final Component Final TRIS 25 mM PBS (10x) 10 % v/v Glycine 192 mM Tween 20 0.1 % v/v low-fatdrymilkpowder 2.5%w/v - in ddH2O - do not adjust pH. Adjusting the pH will - in ddH2O change the ion concentration and thus alter - ﬁlter. conductivity. - split into 10ml aliquots - store at 4°C. - store at -20°C.

Secondary antibody solution Component Final PBS (10x) 10 % v/v PBS (10x concentrated) Component Final Tween 20 0.1 % v/v NaCl 1.37 M -. in ddH2O KCl 27 mM - store at room temperature. Na2HPO4*2H2O 100 mM KH2PO4 18 mM Wash solution - in ddH2O,pH7.4. Component Final - autoclave. PBS (10x) 10 % v/v - store at room temperature °C. Tween 20 0.05 % v/v Triton X-100 1 % v/v SDS 0.1 % v/v

-. in ddH2O - store at room temperature. Blocking solution Component Final PBS (10x) 10 % v/v Stripping buffer Tween 20 0.1 % v/v Component Final low-fatdrymilkpowder 5%w/v Glycine 25 mM SDS 1 % w/v - in ddH2O - ﬁlter. - in ddH2O,pH2.2withHCl. - split into 40ml aliquots - adjust the pH prior to adding SDS. - store at -20°C. - store at room temperature. 3.6. IMMUNOHISTOCHEMISTRY 83

3.6 Immunohistochemistry

For immunohistochemistry using the antibodies directed against rmPER, rmTIM1, rmCRY2 and dPDF, a ﬂuorescent two-step indirect method was employed. The supraesophageal ganglion was either sectioned or treated as whole-mount, complete nervous system preparations were always processed as whole-mounts. In the case of time points in the dark-phase, animals were dissected under red light (λ >610nm) from a single high-power LED.

3.6.1 Primary antibodies

For details to rmPER, rmTIM1 and rmCRY2 primary antibodies, see section ”Antibody Gener- ation”. The anti-Drosophila-PDF (dPDF) antibody, a monoclonal antibody raised in mice [51], was purchased from the Developmental Studies Hybridoma Bank (DSHB, University of Iowa, USA) [51]. Speciﬁcity to R. maderae PDF was shown by Schulze et al. [299].

3.6.2 Secondary antibodies

As secondary antibodies, affinity purified fluorophore conjugated immunoglobulin G (IgG) antibodies raised in goat, directed against IgG heavy and light chain of the target species (Jackson ImmunoResearch, West Grove, Pennsylvania, USA) were used. Freeze-dried secondary antibodies were rehydrated with ddH2O, aliquoted and stored at -80°C. Concentration of the antibody in stock solution was 1.5mg/ml in phosphate buffered saline (PBS, 0.01M sodium phosphate, 0.25M NaCl, pH7.6) containing 15mg/ml BSA (IgG-free, protease-free) and 0.05% sodium azide. For an overview of the secondary antibodies used see Table 3.7, for details on fluorophores see Table 3.8 and fig. 3.5. Fluorophore conjugated secondary antibodies were used diluted 1:300, except the Alexa647 coupled anti mouse IgG antibody derived in goats (GaM-Alexa647) in combination with the anti dPDF antibody, which was diluted 1:5000. The secondary antibody solution also contained 1 µg/ml 4’,6-diamidino-2-phenylindole (DAPI) for nuclear counterstaining in most cases.

Primary Antibody Dilution Secondary Antibody Fluorophore Dilution anti rmPER (rabbit) 1:1000 - 1:20.000 GaR IgG (H+L) Cy™3 1:300 anti rmPER (rabbit) 1:1000 - 1:20.000 GaR IgG (H+L) Cy™5 1:300 anti rmTIM1 (rabbit) 1:500 - 1:10.000 GaR IgG (H+L) Cy™3 1:300 anti rmCRY2 (rabbit) 1:500 - 1:10.000 GaR IgG (H+L) Cy™3 1:300 antidPDF(mouse) 1:1000 GaM IgG (H+L) Cy™3 1:300 antidPDF(mouse) 1:1000 GaMIgG(H+L) AlexaFluor® 647 1:5000

Table 3.7: Primary and secondary antibody combinations and dilutions. Abbreviations: Cy, cyanine, GaM, goat anti-mouse; GaR, goat anti-rabbit; H+L, heavy and light chain; IgG, immunoglobulin G 84 CHAPTER 3. MATERIALS AND METHODS

3.6.3 Sections Sectioning was done using a vibrating blade microtome (Leica VT 1000, Wetzlar, Germany). Supraesophageal ganglia were sliced in sections of either 40µm or 100µm thickness. For comparison of Zeitgeber times, only 100µm sections were used. The protocol for 40µm sections followed the 100µm section protocol given below.

Protocol: Sections Dissection

1. in order to compare Zeitgeber times, use two animals per preparation (ZT4 and ZT16, or ZT 8 and ZT 20) and perform every step in parallel.

2. stun the animal for a few seconds in icewater.

3. remove the head.

4. pin down the head with insect pins #1 and open from the front.

5. remove fat body tissue and tracheae until the supraesophageal ganglion is exposed.

6. add a drop of ﬁxative (4 % paraform aldehyde (PFA)).

7. cut the esophageal connectives, then the optic nerves.

8. carefully remove the supraesophageal ganglion from the head capsule.

Fixation

1. use the same trough for all time points, mark the brains from different times (for example, leave a piece of antennal nerve leftover).

2. ﬁx the supraesophageal ganglia using 4% PFA overnight at 4°C.

Embedding

1. rinse three times brieﬂy in TRIS-buffered saline (TBS) + 0,1% Triton X-100.

2. wash four times for 15min in TBS + 0,1% Triton X-100.

3. remove any residual ﬂuid from the brain and embed in albumin/gelatin.

4. incubate the embedded brains in 10% Formaldehyde in TBS for 1h at 4°C. Use distigu- ishable molds for different time points.

5. remove the embedded brains carefully from the mold. Use pins with different colors to indicate different time points. Incubate in 10% Formaldehyde in TBS overnight at 4°C. 3.6. IMMUNOHISTOCHEMISTRY 85

Sectioning 1. wash the embedded brains shortly in TBS.

2. trim the albumin/gelatin block. Indicate different time points by removing a corner of the block for example. Slides of different time points must be distinguishable.

3. make 100µm sections. Collect series of two time points (i.e. two supraesophageal ganglia) in the same glass trough. It is very important to carry out every incubation step in the same trough with the same solution to ensure comparability. - For incubation steps use 1ml volume / glass trough - Make sure the sections are really ”free ﬂoating” and not sticking to each other or to the surface of the glass trough

Preincubation 1. wash three times for 10min in TBS + 0.1% Triton X-100.

2. incubate for 2h at room temperature in the following solution:

• 5% normal goat serum (NGS) • in TBS + 0.5% Triton X-100

Primary antibody incubation 1. incubate overnight at 4°C in the following solution: • optimal diluted antibody/antibodies • 2 % NGS • in TBS + 0.5% Triton X-100

Secondary antibody incubation 1. after primary antibody incubation rinse three times brieﬂy with TBS + 0.1% Triton X-100.

2. wash four times for at least 15min in TBS + 0.1% Triton X-100. 3. from now on every step should be performed in the dark.

4. incubate overnight at 4°C in the following solution:

• secondary antibody (1:300): GaX coupled with ﬂuorescent dye • 1 % NGS • in TBS + 0.5% Triton X-100 • for counterstaining: DAPI, 10µg/ml (=1µl/ml of the 10mg/ml stock solution) 86 CHAPTER 3. MATERIALS AND METHODS

Mounting

1. rinse three times brieﬂy with TBS + 0.1% Triton X-100.

2. wash three times for at least 20min in TBS + 0.3% Triton X-100.

3. rinse brieﬂy in ddH2O. 4. mount the sections on albumin coated slides. Mount two brains (= two time points) on one slide, one on each side.

5. dry the slides until any residual ﬂuid is evaporated, or overnight.

6. dehydrate using 50%, 70%, 90%, 96%, and two times 100% ethanol for 3min each.

7. incubate twice for 5min in 100% xylol or Roticlear.

8. embed in Entellan. Use coverslides of 170µm thickness.

9. store the slides horizontal under a fume hood overnight to allow the embedding medium to dry.

10. for longterm storage, store the slides in the dark at 4°C.

3.6.4 Whole-mounts For whole-mounts, either the supraesophageal ganglion with the ventral nerve cord, or merely the supraesophageal ganglion was used. Except for dissection and the ﬁnal embedding, the same protocol was used for both types of whole-mounts. For quantiﬁcation of rmPER immunoreactivity, ten brains each were dissected at ZT4, ZT16, ZT8 and ZT20. Brains from ZT4/ZT16 and ZT8/ZT20 were treated the same way and incubated in the same dishes and solutions.

Protocol: Whole-mounts Dissection (Central nervous system)

1. stun the animal 30min at 4°C.

2. remove the legs and wings.

3. open the abdomen with a clean scissor on both sides posterior to anterior. Try to cut more dorsal than ventral.

4. pin down the animal (dorsal side up) in a dish ﬁlled with wax, submerge the animal in cooled cockroach Ringer solution (alternatively: submerge in 4 % PFA in TBS).

5. open the abdomen by pulling carefully on the dorsal part. 3.6. IMMUNOHISTOCHEMISTRY 87

6. remove guts, glands and fat body tissue until you see the nervous system.

7. beginning at the terminal abdominal ganglion work your way carefully dissecting up to the metathoracic ganglion.

8. remove the lower shell of the abdomen, leaving the abdominal part of the ventral nerve cord ﬂoating in Ringer.

9. work your way up until only the head remains with the CNS ﬂoating freely in Ringer solution.

10. pin down the head from the back and carefully open it, not damaging the subesophageal ganglion (SOG).

11. remove muscle and fat body tissue until the brain is fully exposed.

12. cut the optic nerves.

13. carefully remove the brain from the head shell, watch out for the esophagus and the attached corpora cardiaca (CC) and corpora allata (CA). Be very careful not to damage the esophageal connectives.

14. remove any residual photoreceptor cells. Those will stain the nervous system and increase background ﬂuorescence.

15. pin down the dissected nervous system on wax and remove a part of the fat body tissue and tracheae moving alongside the CNS.

Dissection (Supraesophageal ganglion)

1. stun the animal approximately 30s in ice water.

2. decapitate the animal with a clean scissor. Keep the body on ice until the end of dissection and then freeze the remains at −20 °C.

3. pin down the head with the frons up using insect pins #1.

4. open the head capsule at the frons.

5. grab the now loosened frons at one of the antennae and carefully pull it to one side, cutting any muscles or nerves still being attached to the cuticle.

6. remove any fat body and tracheae on top of the supraesophageal ganglion.

7. remove the CC and CA under the supraesophageal ganglion carefully with a forceps from the dorsal side.

8. cut the esophageal connectives and the optic nerves on both sides. 88 CHAPTER 3. MATERIALS AND METHODS

9. carefully remove the supraesophageal ganglion from the head capsule and submerge it in fixative. When submerging in fixative, make sure the optic lobes do not turn down and stick to the central brain. To avoid this, either support the optic lobes with the two tips of a forceps, or carefully grab the supraesophageal ganglion on residual optic nerves of an optic lope. 10. remove any residual photoreceptor cells, tracheae and fat body. Fixation 1. fix the nervous system still pinned down using 4% PFA in TBS overnight at 4 °C. 2. after fixation continue carefully removing residual non neuronal tissue like fat body tissue or tracheae. For fixation of supraesophageal ganglions only, use 40x40mm glass troughs with glass cover and 1 ml fixative. Incubate either overnight at 4 °C or 2 h at room temperature on a shaker.

Enzymatic digestion 1. use 40x40mm glass troughs with glass cover and 1ml solution volume. 2. after fixation rinse three times briefly in TBS. 3. wash four times for at least 30 min in TBS. (Keep in mind: The fixative will also destroy the digesting enzyme, so wash very thoroughly, but carefully).

4. incubate 1 h at 37°C in 1 mg/ml Hyaluronidase (alternative: with 1mg/ml collagenase/dispase 15min in TBS at 37°C) in TBS. 5. rinse twice in TBS + 0.3% Triton X-100.

6. wash four times for 30min in TBS + 0.3% Triton X-100. Preincubation

incubate for 1-2 days at 4°C in the following solution, use 40x40mm glass troughs with glass cover and 1ml solution volume: • 5 % NGS • 0.02% sodium azide • in TBS + 0.3% Triton X-100 Primary antibody incubation

incubate for 3-5 days at 4°C in the following solution, use 40x40mm glass troughs with glass cover and 1ml solution volume: 3.6. IMMUNOHISTOCHEMISTRY 89

• optimal diluted primary antibody (antibodies) • 2 % NGS • 0.02% sodium azide • in TBS + 0.3% Triton X-100 Secondary antibody incubation 1. use 40x40mm glass troughs with glass cover and 1ml solution volume. 2. rinse twice with TBS + 0.3% Triton X-100. 3. wash four times for at least 20 min in TBS + 0.3% Triton X-100. 4. from now on every step should be done in the dark. 5. incubate for 3-5 days at 4°C in the following solution: • secondary antibody (1:300): GaX coupled with Cy3 or Cy5 (or both) • 1 % NGS • 0.02% sodium azide • in TBS + 0.3% Triton X-100 • for counterstaining: DAPI, 1µg/ml Mounting 1. rinse twice in TBS + 0.3% Triton X-100. 2. wash four times for at least 20min in TBS + 0.3% Triton X-100. Alternatively, to remove residual antibodies as completely as possible wash overnight. 3. rinse brieﬂy in TBS. 4. dehydrate using 50%, 70%, 90%, 96%, and two times 100% ethanol for 15min each. 5. incubate for 15min in 100% methylsalicylate. 6. incubate for at least 1h in 100% methylsalicylate. 7. embed in Permount on selfmade special coverslides (Fig. 3.4). Use 170 µm thick coverslides. 8. store the slides horizontal under the fume hood for some days to allow the embedding medium to dry on the surface. If air bubbles occur near the CNS, store the slide 45° vertically to allow the bubbles to move to one side. The CNS will stay in place. 9. for longterm storage, store the slides in the dark at 4°C. 90 CHAPTER 3. MATERIALS AND METHODS

6 50 17

-3- 24 26

76 Figure 3.4: Slides for complete nervous system whole-mounts. Numbers next to scalebars represent length in mm. 1: glass from standard slides, 6x26x1 mm; 2: coverslide, 50x24x0.17 mm; 3: coverslide, 60x24x0.17 mm; 4: stack of two fragments of coverslides, 50x5x0.17 mm each; 5: frosted glass from standard slide, 17x26x1 mm. The parts are glued together with Entellan.

3.6.5 Liquid phase preadsorption Peptides for liquid phase preadsorption were manufactured by PANAtecs GmbH, T¨ubingen. The anti rmPER (rabbit) antibody was preadsorbed for 5h with each of the three peptides used for immunization and a mixture thereof. The anti rmPER (rabbit) and anti dPDF (mouse) antibodies were used in a dilution of 1:1000. Secondary antibodies (GaR-Cy2 and GaM-Cy5) were diluted 1:300.

Protocol: Liquid phase preadsorption 1. dissect and ﬁx supraesophageal ganglia as described in subsection 3.6.3. 2. wash four times for 15 min with salt saline, TRIS-buffered (SST) + 0.1 % Triton X-100. 3. embed supraesophageal ganglia in albumin/gelatin as described in subsection 3.6.3. 4. prepare the following solutions: • For each peptide: four dilutions (100, 10, 1, 0.1 µM) of peptide, in SST + 0.5% Triton X-100, 5% NGS, primary antibody diluted as used in immunohistochemical stainings. • If several peptides were used for immunization, also prepare a series of dilutions as above with equimolar amounts of each peptide. • Control: primary antibody diluted as used in immunohistochemical stainings in SST + 0.5% Triton X-100, 5% NGS. 5. incubate peptide and control solutions 5h on a shaker at room temperature. 6. Make 40µm sections, distribute consecutive sections to n wells containing sodium phosphate buffer (NaPi), where n equals one well per peptide plus two (peptide mix and control). 3.6. IMMUNOHISTOCHEMISTRY 91

7. incubate sections on a shaker for 2h in SST + 0.5% Triton X-100 containing 5% NGS at room temperature.

8. incubate sections in peptide/control solutions on a shaker overnight at room temperature.

9. wash three times for 10min with SST + 0.1% Triton X-100.

10. incubate in SST + 0.5% Triton X-100 containing 1% NGS and secondary antibody (antibodies) for 1 h at room temperature.

11. wash three times for 10min with SST + 0.1% Triton X-100.

12. mount sections as described in subsection 3.6.3.

3.6.6 Imaging

Figure 3.5: Lambda scans of some ﬂuorophores used for whole-mounts. Emission spectra of secondary antibody coupled ﬂuorophores in whole-mounts are shown. Colored vertical lines represent excitation lasers, numbers next to lines indicate laser wavelength.

Antibody Fluorophore Abs. (λ) Ex. (λ) PMT (λ) Goat anti Rabbit Cy™3 bis-NHS ester 550 543 560-580 Goat antiMouse Cy™3 bis-NHS ester 550 543 560-580 Goat anti Rabbit Cy™5 bis-NHS ester 650 670 660-680 Goat anti Mouse Alexa®Fluor647carboxylicacid 651 633 659-679 - DAPI 350 405 434-484

Table 3.8: Characteristics of ﬂuorophores conjugated to secondary antibodies. Abbreviations: Abs., Absorption peak; Ex., Excitation peak; PMT Photomultiplier detection range

A Leica TCS SP5 Confocal Laserscanning Microscope (CLSM) was used for imaging. Scans were performed with a Leica HCX PL apochromate 20 /0.7 multi-immersion objective. Samples were always scanned in sequential mode to avoid bleedthrough, from the fluorophore with the highest to the fluorophore with the lowest excitation wavelength. Fluorophores were excited with a 405nm UV Diode laser, 543HeNe or 633nm HeNe laser. For emission spectra see Fig. 3.5, 92 CHAPTER 3. MATERIALS AND METHODS for fluorophore characteristics and photomultiplyer (PMT) settings see Table 3.8. X/Y resolution was 1024px, resulting in a square pixel size of 757nm, unless noted otherwise. Z-resolution was 1.98µm for whole-mounts, and system optimized resolution (629nm) for 100µm sections, unless noted otherwise.

3.6.7 Image processing Confocal stacks were imported from the proprietary Leica ﬁle format to ImageJ (v. 1.46t, Na- tional Institutes of Health, USA) using the Bio-Formats plugin from LOCI plugins (v. 4.4.1, Laboratory for Optical and Computational Instrumentation, Madison, WI, USA). For image stitching, if necessary, the Pairwise stitching plugin [244] for ImageJ was used. Colocalized pixel maps were computed using the Colocalization threshold plugin for ImageJ. Maximum projections were computed using the maximum intensity projection type.

3.6.8 Quantification Optic lobes were scanned with the same settings from each side, which resulted in four confocal stacks per undamaged brain. Confocal stacks were imported from Leica image files to ImageJ using the Bio-Formats plugin. RGB channels were split and saved as 8bit grayscale Tagged Image File Format (TIFF) stack. The grayscale TIFF stack of the fluorescence signal of interest was then imported into Matlab and the arithmetic mean pixel intensity for all pixels above a certain threshold was calculated resulting in only those signals (=pixels) being used in the calcu- lation that originate from the specific antibody signal. The Man-Whitney-U-test was then used to compare values between two Zeitgeber times. 3.6. IMMUNOHISTOCHEMISTRY 93

3.6.9 Solutions

TRIS Buffered Saline (TBS) Albumin-gelatin Component Final Component Final NaCl 137 mM Egg Albumin 30 % w/v KCl 2.7 mM Gelatin 4.8 % w/v TRIS 25 mM For approx. 100ml: - in ddH2O,pH7.4(HCl). - add 100ml ddH2O to 37.5g egg albumin - autoclave. and stir shortly. - store at 4°C. - add a stirring bar and stir overnight on a magnetic stirrer at 4 °C. Sodium phosphate buffer (NaPi) - let the egg albumin solution rest for 15min Component Final without stirring at room temperature. Na2HPO4*2H2O 46.5 mM - add 25ml ddH2O to 6g gelatin and stir NaH2PO4*2H2O 19 mM shortly. - in ddH2O,pH7.4(HCl). - incubate gelatin solution in a shaking water - store at 4°C. bath at 50°C until the gelatin completely dissolved and the solution is clear, uniform Salt saline, TRIS-buffered (SST) and free of air bubbles. Component Final - incubate the albumin solution at 50°C in a NaCl 300 mM shaking water bath for approx. 2min. TRIS 100 mM - with both solutions still in the water bath, - in ddH2O,pH7.4(HCl). pipette (5ml pipette) the gelatin solution in - store at 4°C. the albumin solution without introducing air bubbles. Slowly stir with the pipette tip while Cockroach Ringer pipetting. Component Final - continue stirring slowly with the pipette tip NaCl 156 mM for a few minutes. KCl 4 mM - split the albumin-gelatin into 5ml aliquots CaCl 6 mM without introducing air bubbles. Hepes 10 mM - solidify the albumin-gelatin solution at 4°C. Glucose 5 mM - store at -20°C. - in ddH2O,pH7.1(NaOH). - 380mOsm (Mannitol). - store at 4°C. 94 CHAPTER 3. MATERIALS AND METHODS

3.7 Running wheel assays

Male R. maderae were kept in running wheels for at least three weeks in a light proof wooden box. The box was illuminated with 100lx or 1000lx from a cold white 5mm SuperFlux LED array (World Trading Net GmbH & Co. KG, Bleicherode, Germany) at the animals level. Ani- mals were transferred from the respective colony to running wheels and entrained in LD cycles for at least one week. After this period of time, the light was turned off and the animals were kept in constant darkness for additional two weeks. Water and food (standard rodent chow, ssniff V2144, ssniff Spezialdiäten GmbH, Soest) were supplied ad libitum. Activity was recorded with a custom built data logger using an Arduino ATmega 2560 micro- controller board (Smart Projects Srl, Strambino, Italy) with a Mega mSD-Shield (v 1.2, Wat- terott electronic GmbH, Leinefelde, Germany) containing a microSD-Socket and a DS1302 Real Time Clock (Maxim Integrated, San Jose, USA). Signals from a reed switch were recorded with 115200 baud per channel. If the switch closed, a timestamp was logged in ASCII file format on a microSD card with the animal number that caused the signal. Timestamps were converted to a continuous timeline containing 1440 points (minutes) per day with the activity per minute summed up using MATLAB. Chi2 periodograms were calculated with the ImageJ plugin ActogramJ[295] with α=0.05 and a trial period of 16 to 32 hours. Rhythmic power was calculated by dividing the peak amplitude by the significance threshold at the period length where the peak was detected using MATLAB. If the rhythmic power was lower than 1.5 or if the peak width was smaller than two, animals were considered arhythmic and excluded from the analysis. Lomb-scargle periodograms were also calculated with similar results as the Chi2 periodogram. Actograms and periodograms were plotted using MATLAB. 4 Results

4.1 Molecular cloning

Initially, degenerate primers against per, tim1, cry1, cry2, clock, cycle, pdp1 and vri were made. Attempts to amplify a cDNA fragment of a R. maderae ortholog of cry1, clock, cycle, pdp1 and vri failed. The main reason for these difﬁculties was the limited availabilty of orthologous sequences in the databases at the time. Thus, it was decided to focus on per, tim1 and cry2. These genes are part of the ”core feedback loop” in many insects. To analyze their structure and expression in R. maderae, the cDNA of these genes was cloned, sequenced and their structure analyzed. With degenerate primers and RACE-PCR, full length cDNA sequences of rmPer, rmCry2 and a partial sequence of rmTim1 were acquired. In addition to these three genes, sequencing of clones subsequent to RACE-PCR revealed several R. maderae genes which were not targeted by the primers.

4.1.1 Period Cloning and sequencing of the Rhyparobia maderae period ortholog’s cDNA (rmPer) resulted in a sequence of 4261 basepairs (bp). The nucleotide sequence was submitted to Genbank and was assigned the accession number JX235363. It contains an initial ATG at position 133-135, a putative full length open reading frame (ORF), a stop codon (TAA) at position 3961-3963 and part of the 3’ and 5’ untranslated region (UTR). The protein encoded by the ORF has a predicted size of 1276 amino acids and an estimated molecular weight of 138.7 kDa (accession number AGA01525). A domain enhanced lookup time accelerated basic local alignment tool [21] (DELTA-BLAST, v. BLASTP 2.2.28+) database search revealed that the amino acid sequence of rmPER has69%(B. germanica, B. bisignata) to 79 % (P. americana) identity along the entire known amino acid sequences of these cockroach species. The E-value, which is a value for the probablity of coincidental sequence similarity, was 0 for these species. The maximum identi- ties for other well know insect species in circadian research ranged from 30 % (D. plexippus, E-value 8e-131) to 50% (G. bimaculatus, E-value: 0), with a maximum identity of 45% for D. melanogaster.

Protein domain structure and phylogenetic analysis Comparing the coding regions of D. melanogaster, D. pseudoobscura and D. virilis period, Colot et al. found six blocks of conserved amino acid sequences [42]. These blocks, named c1 to c6 (c for ”conserved”), are conserved across insect species and R. maderae as well (Fig. 4.1). A seventh conserved sequence block, defined in B. germanica [166] could also be identified in R. maderae (Fig. 4.1). This conserved region is also present in mammals, but is absent in Drosophila species. Using SMART [160, 298] and CDD [174], PAS-A (PER-ARNT-SIM-A), PAS-B (PER- ARNT-SIM-B) and PAC (PAS associated C-terminal) domains could be identified in rmPER. The

95 96 CHAPTER 4. RESULTS

PAS-A and PAS-B domains were redeﬁned as ”PAS fold” by Hefti et al [96]. The PAC domain lies within the cytoplasmatic localization domain (CLD) [272]. A Period C domain, including the c7 domain, was also predicted. Nuclear localization signal (NLS) prediction [207] resulted in two putative sites: One at the end of the c1 domain (residues 66-KRNKDKEHKKKKLK-79) is well deﬁned in D. melanogaster [272, 337]. A second NLS (residues 1239-KKKR-1242) was predicted near the C-terminal end. This motif is similar in some insects with c7 domain (Fig. 4.1, A.2-A.4). In addition to the two NLS predicted by NLStradamus, a bipartite NLS previously described in D. melanogaster at the end of the c3 domain [33] is also conserved in rmPER (Fig. 4.1, A.2-A.4). In cockroaches, there is no threonine-glycine-repeat (TG-repeat) as described for D. melanogaster, but serine-glycine is repeated four times at a similar position. This repeat is also known from other insects including B. germanica [166]. In phylogenetic analysis, rm- PER grouped with PER proteins of other cockroach species (B. germanica, B. bisignata and P. americana) and hemimetabolous insects (G. bimaculatus, P. humanus, Fig 4.2). 4.1. MOLECULAR CLONING 97

C1(60) C2(48) C3(47) C4(44) C5(32) C6(28) D. melanogaster 1218 aa NLS1(86) PAS-A(57) CLD(54) TG NLS2(71) PAS-B(47) CCID(32)

C4(36) C6(30) C1(80) C2(50) C3(49) C5(38) C7(63) A. mellifera 1124 aa NLS1(86) PAS-B(55) NLS2(75) PAS-A(57) CLD(65)

C4(21) C6(17) C1(26) C2(36) C3(32) C5(19) C7(45) A. pisum 1018 aa NLS1(26) PAS-B(39) NLS2(38) PAS-A(43) CLD(50)

C4(19) C6(32) C1(60) C2(58) C3(60) C5(16) C7(55) R. pedestris 1005 aa NLS1(71) PAS-B(66) NLS2(82) PAS-A(51) CLD(68)

C4(69) C6(63) C1(91) C2(80) C3(80) C5(77) C7(92) B. germanica 1225 aa NLS1(93) PAS-B(74) NLS2(100) PAS-A(83) CLD(76) SG

C1 C2 C3 C4 C6C5 C7 R. maderae 1276 aa NLS1 PAS-A PAS-B SG NLS2 CLD

Figure 4.1: Representation of rmPER and various insect PER proteins. C1-C6 domains [42] and the c7 domains [166] are indicated by black boxes. PAS-A (red) and PAS-B (orange) domains are shown as identiﬁed by SMART [160, 298] in rmPER. NLS (blue), CLD (green) and CCID (light gray) are indicated as identiﬁed in D. melanogaster [33, 272, 337]. Numbers in brackets represent percentage amino acid identity with rmPER. Protein size is shown at the end of each drawing. Abbreviations: PAS-A, PER-ARNT-SIM domain A; PAS-B, PER-ARNT-SIM domain B; C1-C7, conserved sequence blocks 1-7; CCID, CLK:CYC inhibition domain; CLD, cytoplasmic localization domain; NLS, nuclear localization signal. 98 CHAPTER 4. RESULTS

Figure 4.2: Phylogenetic relationship of rmPER, various insect and mouse PER proteins. The neighbor- joining algorithm as implemented in MEGA5 [324] with Poisson corrected distances was used. Numbers shown next to the branches represent the percentage of replicate trees in the bootstrap test (1000 replicates), shown only if higher than 50 %. The tree is drawn to scale, and the branch lengths indicate the evolutionary distance in the units of amino acid substitutions per site. 4.1. MOLECULAR CLONING 99

4.1.2 Timeless 1 Several attempts to obtain the full length coding sequence of the R. maderae timeless 1 ortholog using RACE failed. However, a 591bp fragment of rmTim1 was obtained using degenerate primers and RT-PCR (Fig. A.7). This fragment of the ORF was sufficient for quantitative PCR and to generate an antibody. The Figure 4.3: Drawing of D. melanogaster TIM1 iso- sequence was submitted to Genbank and as- form P, indicating the position of the rmTIM1 fragment. The black bar indicates the position of the signed the accession number JX266619. It en- fragment, numbers below the black bar indicate the codes fora 196 aminoacid longpart of the rm- number of conserved residues and total residues. Ab- TIM1 ORF (accession number: AGA01580). breviations: CLD: Cytoplasmatic localization do- Aligned with isoform P (1421aa) of dTIM1, main; NLS: Nuclear localization signal; PER-1: PER the obtained rmTIM1 fragment extends from binding domain 1; PER-2: PER binding domain 2 amino acids 645 to 839. It starts 35 aa 3’ of the first PER dimerization domain identified in dTIM1 [272], which includes the NLS, and spans 93 aa into the second PER dimerization domain (Fig. 4.3). A DELTA-BLAST database search showed that the amino acid sequence of the rmTIM1 fragment has 93% identity (E-value: 2e-125) with the only other known TIM1 protein in cockroaches, P. americana TIM1, 76% with G. bimaculatus and 67 % with D. melanogaster TIM1.

4.1.3 Cryptochrome 2 Cloning and sequencing of the R. maderae cry2 ortholog’s (rmCry2) cDNA resulted in a sequence of 2221 bp, with an initial ATG at position 117-119 and a stop codon (TGA) at position 1875-1877 (Fig. A.5-A.6). The sequence was submitted to Genbank and assigned the accession number JX266618. The predicted protein size was 586 aa, with an estimated molecular weight of 67.3 kDa (accession number: AGA01579). This was the ﬁrst cry gene reported in cockroaches. A BLAST search [5, 6] showed maximum identity values of 80% and 84% for two hemimetabolous insect cry2 genes, Riptortus pedestris and Pyrrhocornis apterus, 84 % for D. plexippus cry2, 87 % for A. mellifera cry2 and 87 % for Anopheles gambiae cry2. The E-values for all hits were zero.

Protein Domain structure and Phylogenetic Analysis A CDD search [174] predicted a DNA-photolyase (aa 10-175) and a ﬂavin adenine dinucleotide (FAD) binding domain (aa 217-492). Domains necessary for CLOCK:BMAL interaction (RD- 2a, RD-1 and RD-2b) in mice are conserved in rmCRY2, as well as a CLD domain [102] and coiled-coil (c-c) region [37] (Fig. 4.4, A.5-A.6). Phylogenetic analysis places rmCRY2 within the insect CRY2 proteins next to another hemimetabolous insect, R. pedestris (Fig. 4.5). 100 CHAPTER 4. RESULTS

Figure 4.4: Schematic representation of rmCRY2, various insect CRY2 proteins and D. melanogaster CRY1. DNA photolyase conserved domain and FAD binding site are indicated by black boxes and were identified by CDD [174]. CLD (green), RD-2a (reddish brown), RD-1 (dark red) and RD-2b (bright red) domains as identified in mouse CRY1 [102]; [151] and a C-C region (cyan) [37] are also shown. Numbers in brackets represent percentage amino acid identity with rmCRY2. Protein size is shown at the end of each drawing. Abbreviations: C-C, coiled-coil domain; CLD, cytoplasmic localization domain; CRY, CRYPTOCHROME; DNA-PL, DNA photolyase domain; FAD-B, flavin adenine dinucleotide binding domain; RD, repressive domain. 4.1. MOLECULAR CLONING 101

Figure 4.5: Phylogenetic relationship of different insect (including rmCRY2) and mouse cryptochromes and insect photolyases. The neighbor-joining algorithm as implemented in MEGA5 [324] with Poisson corrected distances was used. Numbers shown next to the branches represent the percentage of replicate trees in the bootstrap test (1000 replicates), shown only if higher than 50 %. The tree is drawn to scale, and the branch lengths indicate the evolutionary distance in the units of amino acid substitutions per site. Danio rerio CRY-DASH (Danio rerio DASH) was used as an outgroup. 102 CHAPTER 4. RESULTS

4.2 Putative core feedback loop gene expression

4.2.1 Putative core feedback loop gene expression in equinox light-dark cycles

First, it was examined whether there are daily oscillations in expression levels of the putative R. maderae core feedback loop genes rmPer, rmTim1, and rmCry2 in neuronal and non-neuronal tissues (Malpighian tubules) in LD12:12 and the ﬁrst day in constant darkness (DD). Detailed statistics of all qPCR experminents are shown in the Table A.1.

Neuronal tissue

In LD12:12, brain (with the AMe excised) as well as AMe rmPer, rmTim1 and rmCry2 mRNA levels cycled with the highest values occurring in the first half of the night and lowest values in the first half of the day (Fig. 4.6, 4.7, 4.8). Significant differences in mRNA levels between time points for all three genes in brain and AMe tissues were obtained (one-way ANOVA with post- hoc LSD test p<0.05), except for rmPer from AMe tissue (Fig 4.6). This lack of significance is due to a single pool of tissue at ZT4. This pool was not excluded as an outlier since the same sample grouped with the other pools for rmTim1, and rmCry2.

rmPer A rmPer B 2.2 2.2 Brain (−AMe) p=0.01738 AMe p=0.3997 1.8 ab a ab bc c ab ab a ab bc c ab 1.8

1.4 1.4

1 1

0.6 0.6 relative expression ratio 0.2 0.2 0 8 16 0 8 16 0 0 4 8 12 16 20 ZT ZT

Figure 4.6: rmPer mRNA levels in brain and AMe tissues cycle over the course of the day in LD conditions. Relative expression ratios of rmPer for brain (without AMe, black) and AMe (blue) tissues in 12:12 LD cycle are shown. (A) Two consecutive days with the same data are shown to illustrate oscillation. The efficiancy corrected method according to Pfaffl [235] was used to calculate expression ratios. The fitted curve was calculated using cosinor analysis [206]). The bars at the bottom of the plots indicate light (white) and dark (black) phases. Expression levels of two Zeitgeber times with different characters above are significantly different (one-way ANOVA with post-hoc LSD test p<0.05). Error bars represent standard error. The p-value in the upper right corner indicates the result of the one-way ANOVA (α=0.05). (B) Relative expression ratios calculated according to the ∆∆ct method. Each pool is plotted separately. Abbreviations: AMe, accessory medulla; LD, light:dark; ZT, Zeitgeber time. 4.2. PUTATIVE CORE FEEDBACK LOOP GENE EXPRESSION 103

rmTim1 A rmTim1 B 2.2 2.2 Brain (−AMe) p=2.962e−07 AMe p=0.00051 b a a b c b b a a b c b 1.8 1.8 ab a a b c bc ab a a b c bc

1.4 1.4

1 1

0.6 0.6 relative expression ratio 0.2 0.2 0 8 16 0 8 16 0 0 4 8 12 16 20 ZT ZT

Figure 4.7: rmTim1 mRNA levels in brain and AMe tissues cycle over the course of the day in LD conditions. Relative expression ratios of rmTim1 for brain (without AMe, black) and AMe (blue) tissues in 12:12 LD cycle are shown. For details see legend to Fig. 4.6.

rmCry2 A rmCry2 B 2.2 2.2 Brain (−AMe) p=2.471e−07 AMe p=4.098e−05 a a a a b b a a a a b b 1.8 1.8 ab b a ab c c ab b a ab c c

1.4 1.4

1 1

0.6 0.6 relative expression ratio 0.2 0.2 0 8 16 0 8 16 0 0 4 8 12 16 20 ZT ZT

Figure 4.8: rmCry2 mRNA levels in brain and AMe tissues cycle over the course of the day in LD conditions. Relative expression ratios of rmCry2 for brain (without AMe, black) and AMe (blue) tissues in 12:12 LD cycle are shown. For details see legend to Fig. 4.6. 104 CHAPTER 4. RESULTS

Malpighian tubules

In Malpighian tissue, variation between pools was much higher than in other tissues and no signiﬁcant differences between time points in any of the three genes were observed (Fig. 4.9, 4.10, 4.11).

rmPer A rmPer B 2.2 2.2 Malpighi p=0.08021

1.8 1.8

1.4 1.4

1 1

0.6 0.6 relative expression ratio 0.2 0.2 0 8 16 0 8 16 0 0 4 8 12 16 20 ZT ZT

Figure 4.9: rmPer mRNA levels in Malpighian tubules do not cycle over the course of the day in LD conditions. Relative expression ratios of rmPer for Malpighian tubules tissues in LD 12:12 cycle are shown. (A) Two consecutive days with the same data are shown to illustrate oscillation. The efficiancy corrected method according to Pfaffl [235] was used to calculate expression ratios. The fitted curve was calculated using cosinor analysis [206]). The bars at the bottom of the plots indicate light (white) and dark (black) phases. Error bars represent standard error. The p-value in the upper right corner indicates the result of the one-way ANOVA (α=0.05). (B) Relative expression ratios calculated according to the ∆∆ct method. Each pool is plotted separately. Abbreviations: ZT, Zeitgeber time.

rmTim1 A rmTim1 B 2.2 2.2 Malpighi p=0.1213

1.8 1.8

1.4 1.4

1 1

0.6 0.6 relative expression ratio 0.2 0.2 0 8 16 0 8 16 0 0 4 8 12 16 20 ZT ZT

Figure 4.10: rmTim1 mRNA levels in Malpighian tubules do not cycle over the course of the day in LD conditions. Relative expression ratios of rmTim1 for Malpighian tubules tissues in LD 12:12 cycle. For details see legend to Fig. 4.9. 4.2. PUTATIVE CORE FEEDBACK LOOP GENE EXPRESSION 105

rmCry2 A rmCry2 B 2.2 2.2 Malpighi p=0.4756

1.8 1.8

1.4 1.4

1 1

0.6 0.6 relative expression ratio 0.2 0.2 0 8 16 0 8 16 0 0 4 8 12 16 20 ZT ZT

Figure 4.11: Relative expression ratios of rmCry2 mRNA levels in Malpighian tubules do not cycle over the course of the day in LD 12:12 conditions. For details see legend to Fig. 4.9.

4.2.2 Putative core feedback loop gene expression in constant darkness Since there was hardly any difference in expression levels between AMe tissues and the remaining brain (Fig. 4.6, 4.7, 4.8), it was decided to use complete brains to monitor the expression levels in DD. Cycling of rmTim1 (Fig. 4.13), and rmCry2 (Fig. 4.14) mRNA in brain tissue continued the first day in DD with significant differences between time points. Highest values of rmPer expression occured during the subjective night and lowest during the subjective day, but they were not significantly different from each other (Fig. 4.12).

rmPer A rmPer B 2.2 2.2 Brain p=0.2001

1.8 1.8

1.4 1.4

1 1

0.6 0.6 relative expression ratio 0.2 0.2 0 8 16 0 8 16 0 0 4 8 12 16 20 ZT ZT

Figure 4.12: rmPer mRNA levels in brain tissue do not significantly cycle during the first day in constant darkness. (A) Two consecutive days with the same data are shown and relative expression ratios were calculated using the efficiancy corrected method [235]. The curve was fitted using cosinor analysis [206]). The bars at the bottom of the plots indicate light (gray) and dark (black) phases of the LD cycle used for entrainment. Error bars represent standard error. The p-value in the upper right corner indicates the result of the one-way ANOVA(α=0.05). (B) Relative expression ratios calculated according to the ∆∆ct method. Each pool is plotted separately. Abbreviations: ZT, Zeitgeber time. 106 CHAPTER 4. RESULTS

rmTim1 A rmTim1 B 2.2 2.2 Brain p=0.00168

1.8 d abc ab a bc c d abc ab a bc c 1.8

1.4 1.4

1 1

0.6 0.6 relative expression ratio 0.2 0.2 0 8 16 0 8 16 0 0 4 8 12 16 20 ZT ZT

Figure 4.13: rmTim1 mRNA levels in brain tissue cycle during the first day in constant darkness. Relative expression ratios of rmTim1 for brain tissue during first day in constant darkness are shown. Expression levels of two Zeitgeber times with different characters above are significantly different (one-way ANOVA with post-hoc LSD test p<0.05). For details see legend to Fig. 4.12.

rmCry2 A rmCry2 B 2.2 2.2 Brain p=0.00144

1.8 ab a a bc c c ab a a bc c c 1.8

1.4 1.4

1 1

0.6 0.6 relative expression ratio 0.2 0.2 0 8 16 0 8 16 0 0 4 8 12 16 20 ZT ZT

Figure 4.14: rmCry2 mRNA levels in brain tissue cycle during the first day in constant darkness. Relative expression ratios of rmCry2 for brain tissue during the first day in constant darkness are shown. Expression levels of two Zeitgeber times with different characters above are significantly different (one-way ANOVA with post-hoc LSD test p<0.05). For details see legend to Fig. 4.12.

4.2.3 Putative core feedback loop gene expression in different photoperiods After examining the daily expression of putative core feedback loop genes in LD12:12 conditions, the expression levels in cockroaches raised in different photoperiods were compared to analyse the influence of day length on clock gene expression in R. maderae. In both, short day (LD6:18) as well as long day (LD18:6) conditions, maxima occurred in the first half of the night with a higher amplitude under long day conditions. Significant differences between time points were found (one-way ANOVA with post-hoc LSD test p<0.05) for rmPer (Fig. 4.15), rmTim1 (Fig. 4.16), and rmCry2 (Fig. 4.17) in short and long day conditions. Aligned at ZT0, the long 4.2. PUTATIVE CORE FEEDBACK LOOP GENE EXPRESSION 107 day oscillation is phase-delayed with respect to the short day oscillation by about one sampling interval (4 hours).

rmPer A rmPer B 2.2 2.2 6:18 p=0.00897 18:6 p=5.76e−05 bc ab c c c a bc ab c c c a 1.8 1.8 b a a b bc c b a a b bc c

1.4 1.4

1 1

0.6 0.6 relative expression ratio 0.2 0.2 0 8 16 0 8 16 0 0 4 8 12 16 20 ZT ZT

Figure 4.15: The phase of rmPer oscillation adjusts relative to the beginning of the scotophase in different photoperiods. Relative expression ratios of rmPer for brain tissue under short day (6:18 LD, black) and long day (18:6 LD, red) conditions are shown. (A) Two consecutive days with the same data are shown to illustrate oscillation. The efficiancy corrected method according to Pfaffl [235] was used to calculate expression ratios. The fitted curve was calculated using cosinor analysis [206]. The bars at the bottom of the plots indicate light (white) and dark (black) phases. Expression levels of two Zeitgeber times with different characters above are significantly different (one-way ANOVA with post-hoc LSD test p<0.05). Error bars represent standard error. P-values in the upper right corner indicate the result of the one-way ANOVA (α=0.05). (B) Relative expression ratios calculated according to the ∆∆ct method. Each pool is plotted separately. Abbreviations: ZT, Zeitgeber time.

rmTim1 A rmTim1 B 2.2 2.2 6:18 p=3.18e−05 18:6 p=6.301e−06 a a b c bc a a a b c bc a 1.8 1.8 c a d a b b c a d a b b

1.4 1.4

1 1

0.6 0.6 relative expression ratio 0.2 0.2 0 8 16 0 8 16 0 0 4 8 12 16 20 ZT ZT

Figure 4.16: The phase of rmTim1 oscillation adjusts relative to the beginning of the scotophase in different photoperiods. Relative expression ratios of rmTim1 for brain tissue under short day (6:18 LD, black) and long day (18:6 LD, red) conditions are shown. For details see legend to Fig. 4.15. 108 CHAPTER 4. RESULTS

rmCry2 A rmCry2 B 2.2 2.2 6:18 p=0.01569 18:6 p=0.00349 a a ab a b a a a ab a b a 1.8 1.8 ab bc a a bc c ab bc a a bc c

1.4 1.4

1 1

0.6 0.6 relative expression ratio 0.2 0.2 0 8 16 0 8 16 0 0 4 8 12 16 20 ZT ZT

Figure 4.17: The phase of rmCry2 oscillation adjusts relative to the beginning of the scotophase in different photoperiods. Relative expression ratios of rmCry2 for brain tissue under short day (6:18 LD, black) and long day (18:6 LD, red) conditions are shown. For details see legend to Fig. 4.15.

Constant darkness

Cycling continued the first day in constant darkness with significant differences between time points, except for rmPer, and rmCry2 in short day conditions (one-way ANOVA with post-hoc LSD test p<0.05, Fig 4.18, 4.19, 4.20). The amplitude in long day conditions was generally higher in the LD cycle than the first day in DD. Strikingly, the expression level of rmTim1 from animals raised under short day conditions suddenly dropped to low levels at ZT 12 the first day in DD.

rmPer A rmPer B 2.2 2.2 6:18 p=0.1066 18:6 p=0.00238 1.8 1.8 c b a a bc ab c b a a bc ab

1.4 1.4

1 1

0.6 0.6 relative expression ratio 0.2 0.2 0 8 16 0 8 16 0 0 4 8 12 16 20 ZT ZT

Figure 4.18: Oscillation of rmPer continues during the ﬁrst day of constant darkness after entrainment to LD 6:18 and LD 18:6. Relative expression ratios of rmPer for brain tissue under short day (6:18 LD, black) and long day (18:6 LD, red) conditions during the ﬁrst day in constant darkness are shown. For details see legend to Fig. 4.15. 4.2. PUTATIVE CORE FEEDBACK LOOP GENE EXPRESSION 109

rmTim1 A rmTim1 B 2.2 2.2 6:18 p=0.00244 18:6 p=8.046e−06 a a b c ab a a a b c ab a 1.8 1.8 c b a d a bc c b a d a bc

1.4 1.4

1 1

0.6 0.6 relative expression ratio 0.2 0.2 0 8 16 0 8 16 0 0 4 8 12 16 20 ZT ZT

Figure 4.19: Oscillation of rmTim1 continues during the ﬁrst day of constant darkness after entrainment to LD 6:18 and LD 18:6. Relative expression ratios of rmTim1 for brain tissue under short day (6:18 LD, black) and long day (18:6 LD, red) conditions are shown. For details see legend to Fig. 4.15.

rmCry2 A rmCry2 B 2.2 2.2 6:18 p=0.2386 18:6 p=0.03853 1.8 1.8 c c ab bc abc a c c ab bc abc a

1.4 1.4

1 1

0.6 0.6 relative expression ratio 0.2 0.2 0 8 16 0 8 16 0 0 4 8 12 16 20 ZT ZT

Figure 4.20: Oscillation of rmCry2 continues during the ﬁrst day of constant darkness after entrainment to LD 6:18 and LD 18:6. Relative expression ratios of rmCry2 for brain tissue under short day (6:18 LD, black) and long day (18:6 LD, red) conditions are shown. For details see legend to Fig. 4.15. 110 CHAPTER 4. RESULTS

Daily mean expression levels The daily mean of rmPer, rmTim1, and rmCry2 expression levels was significantly higher in long day than in short day conditions (one-way ANOVAp<0.05, Fig. 4.21, B). The daily mean of rm- Per, and rmCry2 expression was much higher than the rmTim1 expression level in all tissues (Fig. 4.21, A). When moved from short day to DD, the expression levels significantly increased for rmPer and rmTim1 and did not change in rmCry2 (one-way ANOVA p<0.05). Reversely, when moved to constant darkness from long day conditions, expression levels significantly dropped for all three genes (one-way ANOVA p<0.05, Fig. 4.21, B).

Figure 4.21: The daily mean of expression levels depends on the light regime. Daily mean of the expression levels of rmPer, rmTim1, and rmCry2 for different tissues (A) and different photoperiods (B) are shown. Error bars represent standard error. Asterisks indicate significant differences (one-way ANOVA p<0.05). Abbreviations: AMe, accessory medulla; DD, first day in constant darkness; LD, light:dark; n.s., not significant. 4.3. CIRCADIAN OSCILLATIONS IN PROTEIN ABUNDANCE 111

4.3 Circadian oscillations in protein abundance

To examine if the circadian changes observed in mRNA levels are also present on protein levels, western blots were performed using antibodies directed against rmPER, rmTIM1 and rmCRY2. As in quantitative PCR experiments, brains were dissected every four hours and ten brains per sampling point were pooled. On mRNA level, the amplitude and mean expression level (Fig. 4.15, 4.16, 4.17) of daily oscillation was higher in long day conditions. Due to the rather low illuminance (100lx) in the mass culture boxes, male R. maderae were taken from the mass cultures three days before dissection and kept in a LD12:12 rhythm with an illuminance of approximately 1000lx, to try to increase amplitudes of gene expression for the time series experiments. In the ﬁrst experiments performed, each lane of the SDS-PAGE was loaded with 10 µg ac- Control + 100µM + 10µM cording to the Pierce 660nm protein assay.

However, loading the SDS-PAGE with the kDa ZT4 ZT16 ZT4 ZT16 ZT4 ZT16 same volumina of homogenate proved to be more precise than loading the gel according to the total protein concentrations determined by 200 the Pierce 660 nm protein assay (see subsection 5.3.1). Thus, the gels for SDS-PAGEs of 150 some western blots shown were loaded with identical volumina. In these cases 2 µl corre- < PER 100 spond to approximately 10µg of total protein.

85 4.3.1 PERIOD

The antibody against rmPER (rabbit) rec- Figure 4.22: Preinkubation of the anti rmPER (rab- ognized three prominent bands in western bit) antibody with peptide 3 (RRETSATNTSQGSY) blots using brain tissue: one over 150 kDa, used for immunization abolished the putative rm- one above 120 kDa (at or over 120 kDa or PER specific signal. The antibody was preincu- just below 140kDa, depending on the protein bated at 4 °C in primary antibody solution overnight marker), and one at 70 to 85 kDa. Preincuba- with 100 µM and 10 µM peptide 3 respectively. The tion with Peptide 3 (C-RRETSATNTSQGSY) anti rmPER (rabbit) antibody was diluted 1:1000 and used for immunization, which abolished rm- 10 µg total protein were loaded per lane on a 5 % gel. PER like immunofluorescence completely, strongly reduced the signal of all bands (Fig. 4.22). However, the lowest molecular weight band at 70 to 85 kDa was still faintly visible as well as another very faint band above 200kDa. The band above 120 kDa can split into up to three, mostly two, distinct bands. The protein(s) represented by this band were the only proteins that always gave a signal in all western blots using the anti rmPER (rabbit) antibody, and is thus labeled ”PER” in figures. The rmPER (mouse) antibody recognized no proteins above background using the wet blotting technique, however using the semi-dry method it recognized a band over 120 kDa, hardly above background (Fig. A.8). An antibody directed against B. germanica PER also recognized a pro- 112 CHAPTER 4. RESULTS

tein above 120 kDa, but the background was too strong to clearly distiguish background from specific signal (not shown). Another antibody against the Antheraea pernyi PER-S peptide also recognized a band above 120 kDa (Fig. A.9). The A. pernyi s-peptide differs from the rm- PER s-peptide in only one amino acid residue (A. pernyi: KSSTETPLSYNQLN, R. maderae: KSSTETPPSYNQLN). The secondary antibody (GaR-HRP, diluted 1:5000) did not bind to any proteins on the membrane (Fig. 4.23, A, rightmost lane). Probing different tissues with the anti rmPER (rabbit) antibody resulted in a strong signal in brain tissue and a faint signal in tissue from the terminal abdominal ganglion. There was almost no signal in Malpighian tubules, and none in muscle tissue (Fig. 4.23, A). Using a different blocking solution (BSA, which also gave a high background, instead of milk powder) or a prolonged x-ray film exposure the rmPER signal was stronger in Malpighian tubules, but still weaker than in brain tissue. The rmPER (mouse) antibody only weakly recognized a protein above 120 kDa in brain tissue (Fig. A.8). In contrast to the results obtained from quantitative PCR using cDNA, no clear rhythm was observed in LD 12:12 when sampling at different Zeitgeber times. Differ- ent preparations suggested maxima at different time points, with no differences in the loading control (Fig. 4.23, B; Fig. A.10). However, even when loading two different gels with the same homogenates, slight differences between Zeitgeber times were observed, with no differences in the loading control (western blots shown in Fig. 4.23, B and Fig. A.11) were loaded with the same homogenate). Upon prolonged x-ray film exposure these slight differences between Zeit- geber times vanished.

A B Muscle Malpighi Brain TAG ZT kDa 6 18 6 18 6 18 6 18 - kDa 0 4 8 12 16 20

250 200 260 150 120 < PER 140 130 100 < PER 85 70 100 100 60 70 50 70 < TUB

Figure 4.23: (A) rmPER can be detected in nervous and Malpighian tubules tissues, but not in muscle tissue. A western blot of the different R. maderae tissues is shown. The anti rmPER (rabbit) antibody was diluted 1:1000. 10 µg total protein were used on a 5 % gel. On the last lane (-), only the secondary antibody was used. (B) rmPER does not cycle during the course of the day. A western blot time series of R. maderae brain tissues is shown. The anti rmPER (rabbit) and TYR-TUBULIN antibody were diluted 1:10000. 2 µl homogenate were used on a 7.5 % gel.

4.3.2 TIMELESS1 Employing the purified anti rmTIM1 (rabbit) antibody on western blots, two prominent bands were recognized (Fig. 4.24, B): one at 100-120 kDa and another one representing a higher 4.3. CIRCADIAN OSCILLATIONS IN PROTEIN ABUNDANCE 113 molecular weight protein at 150-200 kDa. The preimmune sera showed no band at 100-120 or 150-200 kDa in neither of the two rabbits used for immunization, whereas the serum of rabbit 12 used 74 days after immunization gave a strong signal at 150-200 kDa (Fig. 4.24, A). Using the purified anti rmTIM1 (rabbit) antibody, the lower molecular weight band also faded upon lower total protein concentrations, lower antibody concentrations or shorter x-ray film exposure. Similar to rmPER, no clear circadian rhythms were detected in western blots (Fig. 4.24, B).

B A Day 0 Day 74 Rabbit 12 Rabbit 14 Rabbit 12 Rabbit 14 ZT kDa 5 17 5 17 5 17 5 17 kDa 0 4 8 12 16 20

260 200 200 250 200 150 150 < TIM 150 < TIM 140 130 120 120 120 100 100 100 100 100 85 85 85 70 70 70 70 70 60 60 60 50 < TUB

Figure 4.24: (A) The signal above 150 kDa when using the anti rmTIM1 (rabbit) antibody in western blots represents a speciﬁc reaction to the immunization. Western blots of R. maderae brains are shown. Anti-rmTIM1 preimmune serum and immune serum sampled 74 days after immunization were used. Sera were diluted 1:1000, 5 µl homogenate were used on a 5 % gel. Numbers above blots indicate ZTs when sampling was performed. (B) rmTIM1 does not cycle during the day in LD 12:12 conditions in brain tissue homogenates. Time series of R. maderae brain tissues. The anti rmTIM1 (rabbit) antibody was diluted 1:1000, the TYR-TUBULIN antibody 1:10000. 2 µl homogenate were used on a 7.5 % gel.

4.3.3 CRYPTOCHROME 2 The immunopurified anti rmCRY2 (rabbit) antibody only recognized several faint bands between 40 and 50 kDa (Fig. 4.26, B) hardly above background. These bands were slightly stronger at ZT12, ZT16 and ZT20 than at ZT0, ZT4 and ZT8 in the blot shown (Fig. 4.26, B), with the background correlating with these differences. The loading control showed hardly any difference between Zeitgeber times. Additional blots using the purified anti rmCRY2 (rabbit) antibodies did not recognize any bands at all. The blood serum obtained from rabbit 3 123 days after immunization recognized a strong band between 70 and 85 kDa as well as several other bands hardly above background, amongst others some bands between 40 and 50 kDa. The band at 70 to 85 kDa did not show any consistent differences between time points (Fig. 4.26, A). However, preimmunesera of rabbit 1, and weakly also rabbit 2 showed a band between 70 and 85 kDa also indicating that the antibodies recognizing this protein were not produced in a specific reaction to the immunization with the antigen (Fig. 4.25). This is also supported by the fact that the immunopurified antibody did not recognize a protein of that size. 114 CHAPTER 4. RESULTS

Day 0 Day 123 Rabbit 1 Rabbit 2 Rabbit 1 Rabbit 2 kDa 5 17 5 17 5 17 5 17 kDa

200 150 150 120 120 100 100 85 85 70 70 < 60 60 *

Figure 4.25: Immune sera of rabbits immunized with an rmCRY2 peptide did probably not contain rm- CRY2 speciﬁc antibodies. Anti-rmCRY2 preimmune serum and immune serum sampled 123 days after immunization were used. Sera were diluted 1:1000, 5 µl homogenate were used on a 5 % gel. Num- bers above blots indicate Zeitgeber times sampled (ZT 5 and ZT 17). The asterisk indicates a probably unspeciﬁc signal.

ZT B A ZT kDa 0 4 1 2 16 20 kDa kDa 0 4 1 216 20 kDa 260 250 200 260 250 140 130 150 200 120 140 130 150 100 100 100 120 85 100 70 < 100 100 70 70 70 85 55 60 * 70 70 60 50 55 40 50 35 40 40 40 35 25 30 35

55 60 55 60 < TUB < TUB 50 50

Figure 4.26: The rmCRY2 immune serum does most probably not recognize rmCRY2 in western blots of R. maderae brain tissue. (A) A western blot time series of R. maderae brain tissues sampled from animals kept in LD 12:12 is shown. The immune serum sampled 123 days after immunization and the anti TYR-TUBULIN antibody were diluted 1:10000. (B) The anti rmCRY2 (rabbit) antibody does most probably not recognize rmCRY2 in western blots of R. maderae brain tissue. The same homogenates as in (A) were used. The puriﬁed antibody was diluted 1:1000, the TYR-TUBULIN antibody 1:10000. 2 µl homogenate were used on a 10 % gel. The asterisk indictes a probably unspeciﬁc signal. 4.4. ANTI RMPER-IMMUNOREACTIVITY IN THE NERVOUS SYSTEM 115

4.4 Anti rmPER-immunoreactivity in the nervous system

To assay the spatial expression pattern of the circadian core feedback loop genes in the nervous system of R. maderae, antibodies directed against rmPER, rmTIM1 and rmCRY2 were used in immunohistochemical experiments. The puriﬁed antibodies directed against rmTIM1 and rm- CRY2 did not yield any signal above background at concentrations ranging from 5x10-2 to 10-4, neither in 100µm sections nor in whole-mounts (Fig. A.12, B, C). The anti-rmPER antibody raised in mice also did not show any immunoreactivity in concentrations ranging from 10-2 to 10-4, neither in sections nor in whole-mounts (Fig. A.12, A). The anti rmPER antibody raised in rabbits recognized an epitope in the nuclei.

4.4.1 Anti rmPER-immunoreactivity in the central nervous system

Deploying two-step fluorescence immunohistochemistry on R. maderae nervous tissue using the anti rmPER (rabbit) antibody resulted in a mass staining of cells in the central nervous system (Figs. 4.30, 4.33). Neurons as well as probably almost all glial cells were rmPER-ir. Lower concentrations (5x10-4, 1x10-4, 5x10-5) of the antibody did not lower the number of stained cells, but resulted in a reduced overall staining. In the optic lobes, the anterior (aPDFMe) (Fig. 4.27) and posterior (pPDFMe) PDF-ir neurons of the medulla (Fig. 4.28) as well as the PDF-ir neurons in the dorsal (dPDFLa) and ventral (vPDFLa) laminae (Fig. 4.27) were all rmPER-ir in the nucleus. To examine if the rmPER-ir was nuclear, cytoplasmatic or both, DAPI (4’,6-diamidino- 2-phenylindole) was used to counterstain the tissue. DAPI is a nuclear marker, that binds to AT-rich regions of the DNA minor groove. Anti rmPER-ir was exclusively nuclear, not a single cell exhibiting cytoplasmatic staining was recognized. The rmPER-ir was generally stronger in the nuclei of large cells, conversely to the DAPI counterstain (Fig. 4.28). Thus, DAPI and PER seemed not to colocalize in very large or very small cells. A single experiment using heat induced epitope retrieval (HIER, Citric acid buffer, 30 min at 90 °C after fixation) resulted in poor tissue preservation and reduced staining, with the same predominantly nuclear staining pattern as in other experiments (not shown). To test for specificity of the purified rmPER antibody to the antigens used for immunization, the antibody was preadsorbed with each of the three peptides used for immmunization, as well as with a mixture of the three peptides. Preadsorption with peptide 1 (MEETATHNTKISDS-C) and 2 (KSSTETPPSYNQLN-C) in concentrations ranging from 100 (Fig. 4.29 A, B) to 0.1µM had no effect on immunoreactivityin the central nervous system. Only preadsorption with peptide 3 (C-RRETSATNTSQGSY) in concentrations of 100µM (Fig. 4.29, C) and 10 µM abolished rmPER immunoreactivity completely. Negative controls (preincubation without peptide) all showed rmPER-ir. A BLAST database search of peptide 3 scored mainly PER proteins of various species, with only cockroach (R. maderae, P. americana, B. germanica and B. bisignata) and cricket (G. bimaculatus, Modicogryllus siamensis, Laupala paranigra and Laupala cerasina) PER proteins having E-values below 10. Merely R. maderae PER shares 100% sequence identity with this peptide. 116 CHAPTER 4. RESULTS

A AeM B C

Figure 4.27: (A) All anterior PDF-ir neurons (red, white arrows: large aPDFMe; black arrows: medium aPDFMe; arrowheads: small aPDFMe) are also rmPER-ir (green) in the nucleus. (B, C) dPDFLa (B, red, arrowheads) and vPDFLa (C, red, arrowheads) are also rmPER-ir (green) in the nucleus. Shown are maximum projections (Z-direction: 14 µm (A), 31 µm (B), 22 µm (C)) of CLSM images containing dPDF- ir. Abbreviations: dPDFLa, dorsal PDF-ir lamina neurons; aPDFMe, anterior PDF-ir medulla neurons; AMe, accessory medulla; vPDFLa, ventral PDF-ir lamina neurons. Scalebar: 20 µm.

A B

* *

Figure 4.28: All posterior PDF-ir (red) medulla neurons colocalize rmPER-ir (green) in the nucleus. Maximum projections (Z-direction: 14 µm) of rmPER-ir (green) and dPDF-ir (red, A) in the pPDFMe and colocalization (white, B) of rmPER-ir (green) with DAPI (blue). Arrowheads: pPDFMe; asterisk: large neuron with typical weak DAPI staining. Scalebar: 20 µm. 4.4. ANTI RMPER-IMMUNOREACTIVITY IN THE NERVOUS SYSTEM 117

A B C

Figure 4.29: Only peptide 3 abolished all rmPER-ir in preadsorption assays. The anti rmPER (rabbit) antibody was preadsorbed with the peptides used for immunization: Peptide 1, MEETATHNTKISDS-C (A); Peptide 2, KSSTETPPSYNQLN-C(B); Peptide 3 RRETSATNTSQGSY-C (C) Shown are maximum projections (Z-direction: 15 µm) of CLSM image stacks scanned by Andreas Arendt. The antibody was diluted 1:1000 and the concentration of the respective peptide was 100 µM. Green: rmPER-ir, red: dPDF- ir. Scalebar: 50 µm.

Figure 4.30: rmPER-ir is widespread in the nuclei of cells in the central brain. A maximum projection (Z-direction: 39 µm) of rmPER-ir (green) and dPDF-ir (red) in the central brain is shown. Abbreviations: Al, antennal lobe; Ca, calyces; Lo, lobula. Scalebar: 100 µm. 118 CHAPTER 4. RESULTS

Anti rmPER-immunoreactivity at different Zeitgeber times

To examine whether the staining in the dPDF- ir cells stays nuclear at all ZTs and to test if a difference in the overall staining intensity can be observed, whole-mounts as well as 100µm n.s. n.s. sections were made using the anti rmPER (rabbit) antibody. Due to the fact that it was unknownif there is a shift between mRNA and protein concentration maxima as observed in D. melanogaster, two different pairs of ZTs were processed, ZT4 and ZT16 or ZT8 and ZT 20. Each ZT pair was treated in parallel, and images were acquired and processed with the same settings. Overall, ten brains per ZT were examined in whole-mounts and five each in 100µm sections with the same result: In op- Figure 4.31: rmPER-ir does not cycle in a circadian tic lobes (Fig. 4.33) and in particular in the manner in the cells of the optic lobes. Comparison of aPDFMe neurons (Fig. 4.32), the rmPER spe- the mean pixel intensities of rmPER-ir fluorescence cific staining remained nuclear at all Zeitgeber in the optic lobes. times examined, in whole-mount dissections as well as in 100µm sections. The mean pixel intensity of all rmPER-ir specific signals of each confocal stack (whole-mounts) was calculated and the values for each ZT were tested for normal distribution using the Lilliefors-test [162]. Since data was not normally distributed, the Mann- Whitney-U-test (α = 0.05) was used to test for significant differences between ZTs. There was no significant difference in overall staining intensity, neither between ZT4 and ZT 16 (p=0.6546) nor between ZT8 and ZT20 (p=0.6509, Fig. 4.31).

e A AMe B AM C AMe D AM

ZT 4 ZT 16 ZT ZT

Figure 4.32: rmPER-ir (green) stays nuclear at all ZTs monitored (ZT 4, A; ZT 16, B; ZT 8, C and ZT 20, D)in the anterior PDF-ir cells (red) of the medulla. Maximum projections of CLSM images containing dPDF-ir are shown. (100 µm sections, Z-direction: 25 µm). Abbreviations: AMe, accessory medulla. Scalebar: 20 µm. 4.4. ANTI RMPER-IMMUNOREACTIVITY IN THE NERVOUS SYSTEM 119

A B

ZT 4 ZT 16

C D

ZT 8 ZT 20

Figure 4.33: rmPER-ir (green) is widespread in the optic lobes at all ZT monitored (ZT 4, A; ZT 16, B; ZT 8, C; ZT 20, D). Maximum projections of CLSM images containing dPDF-ir (red, Z-direction: ZT 4, 166 µm; ZT 16, 198 µm; ZT 8, 188 µm; ZT 20, 209 µm) are shown. Scalebar: 50 µm. 120 CHAPTER 4. RESULTS

4.4.2 Anti rmPER-immunoreactivity in the ventral nerve cord Since rmPER-ir was widespread in the central nervous system, the question arose whether also neurons and ant glia in the remaining nervous system, the ventral nerve cord are rmPER-ir. To answer this question, the complete central nervous system with the ventral nerve cord post was isolated and whole-mount immunohistochemistry was performed to label rmPER. Like in the supraesophageal ganglion, almost every cell in the ventral nerve cord from subesophageal ganglion (Fig. 4.37), thoracic (Fig. 4.38) and abdominal (Fig. 4.36) ganglia to terminal abdominal ganglion (Fig. 4.35) was rmPER- ir in the nucleus in homogenousdensity. Again, as in the central nervous system, the nuclei of large cells were stronger stained than those of smaller cells. Conversely to the DAPI stain, this resulted in rmPER-ir nuclei with Figure 4.34: Putative glial cells in no visible DAPI staining in very large neurons. How- the ventral nerve cord connectives are ever, despite not obviously colocalized with DAPI, this rmPER-ir in the nucleus. Colocalization staining was nuclear. Not only cells in the ganglia were (white) of rmPER-ir (green) with DAPI stained, but also nuclei of putative glial cells on the con- (blue) is shown with a maximum projec- nectives (Fig. 4.34), indicating that the rmPER antibody tion (Z-direction: 183 µm) of a CLSM im- not only recognizes an epitope in neurons, but also in age stack. Scalebar: 50 µm. Abbrevia- glial cells. tions: ant, anterior; post, posterior.

A B C D

Figure 4.35: rmPER-ir is widespread in the nuclei of cells of the terminal abdominal ganglion. Colocal- ization (white) of rmPER-ir (green) and DAPI (blue) is shown with maximum projections (Z-direction: 40 µm) of CLSM image stacks from dorsal (A) to ventral (D). Scalebar: 100 µm. 4.4. ANTI RMPER-IMMUNOREACTIVITY IN THE NERVOUS SYSTEM 121

A B C D

Figure 4.36: rmPER-ir is widespread in the nuclei of cells of the abdominal ganglion. Colocalization (white) of rmPER-ir (green) and DAPI (blue) in the ﬁrst abdominal ganglion is shown with maximum projections (Z-direction: 40 µm) of CLSM image substacks from dorsal (A) to ventral (D). Scalebar: 100 µm.

A B

C D

Figure 4.37: rmPER-ir is widespread in the nuclei of cells of the subesophageal ganglion. Colocalization (white) of rmPER-ir (green) and DAPI (blue) is shown with maximum projections (Z-direction: 40 µm) of CLSM image substacks from dorsal (A) to ventral (D). Scalebar: 100 µm. 122 CHAPTER 4. RESULTS

A B

C D

Figure 4.38: rmPER-ir is widespread in the nuclei of cells of the thoracic ganglia. Colocalization (white) of rmPER-ir (green) and DAPI (blue) in the prothoracic ganglion is shown with maximum projections (Z-direction: 40 µm) of CLSM image substacks from dorsal (A) to ventral (D). Scalebar: 100 µm. 4.5. RUNNING WHEEL ACTIVITY IN DIFFERENT PHOTOPERIODS 123

4.5 Running wheel activity in different photoperiods

To compare locomotor activity rhythms in different photoperiods, the activity of male R. maderae was monitored using running wheel assays. Animals were transferred from the respective colony to running wheels and entrained in LD cycles for at least one week. After this period of time, the light was turned off and the animals were kept in constant darkness for additional two weeks. Activity was recorded using a custom built data logger as described in Materials and Methods, actograms of all recorded animals are shown in Figs. A.13-A.20. The period length and rhythmic power was determined using chi2 periodogram analysis. Since data was not normally distributed (Lilliefors-test p<0.05), the Mann-Whitney-U-Test was used to test for signiﬁcant differences. If the animals’ period length changed in constant darkness, the posterior period length was used for statistics. Two light intensities were tested: 100lx and 1000lx. For each set of experiments, new animals were used.

A B n.s. n.s. n.s. n.s. *** *** * ***

Figure 4.39: The free-running period was signiﬁcantly longer when male R. maderae were entrained to LD 6:18 than to LD 18:6 in bright light conditions, but there were no differences when entrained with lower light intensities. The period length of male R. maderae in short (LD 6:18) and long (LD 18:6) days and the free-running period after entrainment to these light regimes (A, 100 lx; B, 1000 lx) are shown. The Mann-Whitney-U-test (α=0.05) was used to compare two groups. Abbreviations: LD, light:dark cycle; DD, constant darkness; *, p<0.05; ***, p<0.001; n.s., not signiﬁcant.

4.5.1 Low light activity Under light intensities of approximately 100lx, animals showed a tendancy to shorter free- running period lengths in LD18:6 (Fig. 4.41 and 4.42). However, there were no signiﬁcant differences in period length, neither between different photoperiods, nor between LD and DD (Fig. 4.39, A). A light intensity of 100lx correlates approximately to the light intensity in the rearing boxes. 124 CHAPTER 4. RESULTS

Interestingly, some animals exhibited a change in free-running period after some time * in constant darkness (Fig. 4.43): Nine out of n.s. 23 animals kept previouslyin LD6:18 (100lx) * n.s. changed the period length from T>24 h to T<24h after some time in DD. Of animals kept in 18:6 LD, the period length of two out of 21 animals changed in that manner. Also, three animals showed a shorter period length than 24h in LD6:18, and four animals in LD18:6 showed a longer period length than 24h (Table 4.1). This variation and the changing free-running period in some animals were the reason for insignificant differences between groups. The free-running rhythmic Figure 4.40: Rhythmic Power of the free-running power was significantly lower in long day an- periods in different photoperiods and light intensi- imals than in short day animals (p = 0.0173, ties. Abbreviations: *, p<0.05; n.s., not significant. Fig. 4.40).

4.5.2 Bright light activity Under light intensities of approximately 1000lx, the free-running period of animals kept in LD6:18 was significantly longer (p = 7.19·10−6) than the free-running period of animals kept in LD18:6 (Fig 4.39, B). Comparing the free-running period to the period in LD, the free-running period of animals entrained to LD6:18 was significantly longer (p = 1.99·10−4) than the period in LD in all but one animal. Conversely, cockroaches entrained to LD18:6 had a significantly shorter (p = 6.24·10−6) free-running period (Fig 4.39, B; Table 4.1). Additionally, in contrast to lower light intensities, only one animal in LD6:18, and none in LD18:6 exhibited a change in the free-running period within two weeks after lights off (Table 4.1). There was no significant difference in rhythmic power between short and long day animals, however in long day animals the rhythmic power was significantly higher (p = 0.0383) under 1000lx compared to 100lx (Fig 4.40). 4.5. RUNNING WHEEL ACTIVITY IN DIFFERENT PHOTOPERIODS 125

A B 23.7

24.23

Figure 4.41: (A) Example of a double-plot actogram of a male R. maderae’s running wheel activity in LD 6:18 (100 lx) and subsequent DD. The red bars indicate that no data was recorded at that time. (B) Chi2 periodogram analysis of 7 days before the beginning of DD and 14 days thereafter (C). The red line indicates the threshold (0.05), the number next to the peak the period length.

A B

24.0

22.68

Figure 4.42: (A) Example of a double-plot actogram of a male R. maderae’s running wheel activity in LD 18:6 (100 lx) and subsequent DD. The red bars indicate that no data was recorded at that time. (B) Chi2 periodogram analysis of 7 days before the beginning of DD and 14 days thereafter (C). The red line indicates the threshold (0.05), the number next to the peak the period length. 126 CHAPTER 4. RESULTS

DD6:18(100lx) DD18:6(100lx) DD6:18(1000lx) DD18:6(1000lx) Total 24 24 16 16 τ≥24 h 10(41.7%) 4(16.7%) 14(87.5%) 0 τ<24 h 3(12.5%) 11(45.8%) 1(6.25%) 16(100%) τ Change 9(37.5%) 1(4.1%) 1(6.25%) 0 Arhythmic 1(4.1%) 5(20.8%) 0 0 † 1(4.1%) 3(12.5%) 0 0 Free-running τ 1435 1424 1448 1412 (median), min Power (median) 3.3 2.5 3.1 3.8

Table 4.1: Effects of different photoperiods on the free-running period of male R. maderae during the ﬁrst two weeks in constant darkness. ”τ change” refers to animals whose free-running period changes from τ>24 h to τ<24 h. 100 lx/1000 lx refers to the light intensity used for entrainment. Abbreviations: τ, tau (Period length).

A B

24.06

24.1

23.37

Figure 4.43: (A) Example of a double-plot actogram of a male R. maderae’s running wheel activity in LD 6:18 (100 lx) and subsequent DD, exhibiting a change in period length. The red bars indicate that no data was recorded at that time. (B) Chi2 periodogram analysis of seven days before the beginning of DD, 1-9 days (C) and 10-17 (D) days thereafter. The animals free running period changed approximately ten days after the beginning of DD. The red line indicates the threshold (0.05), the number next to the peak the period length. 5 Discussion

To begin analyzing the molecular circadian system of the cockroach R. maderae, the structure and expression profile of the core feedback loop genes per, tim1 and cry2 were analyzed. Using degenerate primers and RACE, the full length sequences of rmPer and rmCry2, and a fragment of rmTim1 could be cloned. Phylogenetic analysis grouped rmPER and rmCRY2 with other hemimetabolous insects, and their structure suggests a similar function. Using quantitative PCR, it could be shown that the transcripts of rmPer, rmTim1 and rmCry2 cycle in phase in different light regimes and that peak levels adjust relative to the onset of the scotophase in different photoperiods, with maxima in the night. In constant darkness, cycling of rmTim1 and rmCry2 continues, but no significant differences between ZTs could be observed for rmPer, probably due to low amplitude cycling. Expression levels of rmPer and rmCry2 were much higher than rmTim1 expression levels, suggesting a more specific role for rmTim1. Based on the acquired ORF sequences, peptide antibodies against the putative aminoacid sequences of rmPER, rmTIM1 and rmCRY2 were generated. rmPER and rmTIM1 antibodies detected a most probably specific epitope in supraesophageal ganglion tissues, and results did not suggest cycling in protein abundance. The anti rmCRY2 (rabbit) antibody probably did not detect any specific epitopes in western blots. To gain information about the cellular expression of the core feedback loop genes, fluorescent immunocytochemistry was employed. Only the rmPER antibody derived in rabbits worked and suggested a widespread expression of rmPER in the CNS with no oscillation, similar to the western blots. The staining was exclusively nuclear at all times and almost every cell in the central nervous system was PER-ir. Running wheel assays were used to examine the locomotor activity of the short and long day photoperiods used in the quantitative PCR experiments. There was no significant difference in the free-running periods of animals kept at 100lx illumination. However at 1000lx animals kept in short days exhibited a highly significant longer free-running period than animals kept in long days.

5.1 Core feedback loop genes in R. maderae

5.1.1 Period Using degenerate primers and RACE, the full length coding sequence of rmPer and part of the 5’ and 3’ UTR could be obtained. Phylogenetic analysis of the putative rmPER protein grouped it with other hemimetabolous insects (G. bimaculatus, P. humanus corporis) and within cockroach PER proteins (P. americana, B. germanica, B. bisignata, Fig. 4.2). A. siphon PER does not group with the other hemimetabolous insects and forms its own branch due to the highly derived nature of the A. siphon circadian clock [46]. Functional domains that have been shown to be necessary for PER speciﬁc function in D. melanogaster are conserved in rmPER (Fig. 4.1). Conservation of the PAS domain region suggests rm- TIM1 interaction as in D. melanogaster. The regionthatalignswith theCCID in D. melanogaster [33] shares only 32% identity with D. melanogaster PER. However, the c3 and c4 conserved se-

127 128 CHAPTER 5. DISCUSSION

quence blocks within the CCID, that contain DBT (c3, dPDBT [136]) and CLK (c4, CBD [317]) interaction domains share a higher identity with D. melanogaster PER. This suggests that rm- PER interacts with TIM1, DBT and CLK like in D. melanogaster, although functional studies are necessary to confirm this. Both NLS identified in D. melanogaster are also conserved in rmPER, suggesting a similar nuclear translocation. In addition, a third NLS (1238-KKKRR-1243) was identified at the C- terminal end by NLS prediction using NLStradamus [207]. This sequence resembles a classic monopartite NLS [176]. These NLS mediate binding to IMPORTIN α, which in turn binds IM- PORTIN β that mediates nuclear import through the nuclear pores. Whether it is a functional NLS remains to be determined. However, this NLS might suggest additional functionality and protein interaction not present in D. melanogaster. RmPER lacks a threonine-glycine repeat between the C-Domain and the dPDBT, similar to all non-Drosophilid species examined so far. Instead, it contains a short repeat of serine-glycine residues like other insect PER proteins [166]. It is unknown whether this short stretch of aminoacids and the adjacent low complexity region has a similar role in temperature compensation and/or courtship song cycle as suggested for D. melanogaster [264, 292]. The C-terminal c7 domain, which was first defined in B. germanica, is also conserved in rmPER, similar to all non- Dipteran species examined so far. The C-terminal region in the mammalian R. norvegicus PER2, was shown to bind CRY1 (a type 2 CRY) [331]. However, although mammals and various insects share the C-terminal c7 region, the aminoacid stretch necessary for CRY1 binding in R. norvegicus is hardly conserved in R. maderae. Nevertheless, the C-terminal region may mediate CRY2 binding in insects too, although so far no functional studies have been deducted. However, Ne- matoceran insects like Culex, Anopheles and Aedes lack a c7 domain like D. melanogaster, but express CRY2. If the c7 domain indeed mediates CRY2-PER interaction in insects, CRY2-PER interaction may be different in Nematoceran insects.

5.1.2 Cryptochrome The aminoacid sequence and domain structure of rmCRY2 indicate that the cloned cDNA is indeed a type 2 CRY. Functional domains that were ﬁrst deﬁned in mammalian CRY2 [102, 37] are conserved in rmCRY2 (Fig. 4.4). A small region that was shown to mediate PER2 binding in mammals is also conserved in rmCRY2 (aminoacids 475-501, 81% identity), including two aminoacids (Arg-487 and Lys-489) that are necessary for PER2 interaction [215]. Interaction of PER:TIM1 with insect CRY2 was shown in D. plexippus [366], and repressor function of insect CRY2s were experimentally proven for 6 insect CRY2s [119, 359]. However, CRY2 proteins can essentially be activated by light, as has been shown in mammals [103], and it is not entirely clear if CRY2 function is light-dependent, independent or both in insects. Several attempts to clone a fragment of cry1 in R. maderae using degenerate primers failed. Al- though this does not prove that this gene is absent from the genome of R. maderae, there are several indications that the R. maderae clock does not receive direct light input via CRY1. Thus this gene might be absent from the genome, or at least it does not function as a blue-light photoreceptor. First of all, the complex eyes appear to be the only source of light entrainment in cockroaches [261, 208]. Second, the action spectrum for entrainment gives no hint to an involve- 5.2. CORE FEEDBACK LOOP GENE EXPRESSION RHYTHMS 129

ment of a blue-light sensitive CRY like D. melanogaster CRY1 [193]. Third, cuticle deposition rhythms in R. maderae are independent from the locomotor pacemaker and unentrainable by light [348]. These observations are in contrast to ﬁndings in D. melanogaster, where these rhythms have been shown to be light-entrainable via CRY1 [122]. Cuticle deposition rhythms are controlled by the clock genes in D. melanogaster like the locomotor pacemaker [122]. If this is also the case in R. maderae, this hints again to a lack of direct light input via CRY1, at least in this peripheral oscillator.

5.1.3 Timeless 1 Only a short fragment of rmTim1 could be cloned (Fig. A.7). Further attempts to acquire the full-length ORF of rmTim1 using RACE were unsuccessful and 3’ and 5’ RACE products did not contain any rmTim1 coding sequences. Similar problems were reported by Cortes et al. [46], who only obtained sequences containing stop codons in both directions of the initial tim1 fragment. The difficulties with the rmTim1 RACE may be explained with the low expression level of rmTim1 compared to the other two circadian genes cloned. The obtained ORF sequence however was sufficient for qPCR and the generation of a specific antibody.

In summary, the putative rmPER aminoacid sequence is conserved in R. maderae and contains all domains necessary for core clock function. Similar to many other insects but in contrast to D. melanogaster, it also contains a conserved C7 region with a putative additional NLS, which suggests additional regulational mechanisms. Cloning and analysis of the putative rmCRY2 sequence revealed that it is a type 2 CRY, is conserved in R. maderae and contains all domains necessary for transcriptional repressor function as in other insects and vertebrates. The rmCry2 sequence cloned is the ﬁrst record of a type 2 cry in cockroaches.

5.2 Core feedback loop gene expression rhythms

5.2.1 Daily expression rhythms in LD cycles In LD12:12 conditions, expression of rmPer, rmTim1, and rmCry2 is rhythmic and transcript levels cycle in the same phase in the brain of the Madeira cockroach. Maxima occur in the ﬁrst half of the dark phase around ZT16-ZT20, and minima around ZT4-ZT8. This is similar to D. melanogaster [93, 302], D. plexippus [366] and A. mellifera [269] and indicates, together with the conserved domain structure, that the expression rhythms are driven by the same regulatory elements as in these insects. It also conﬁrms that rmCry2 is indeed a type 2 cry, since cry1 is like clk regulated by vri and pdp1 in other insects like D. melanogaster [50, 84] and thus oscillates in phase with clk and not with per/tim1. The amplitude of rmPer, rmCry2 (<2-fold) and rmTim1 (∼2.5-fold) expression is rather low compared to other insects. This weak circadian modulation was also reported in other hemimetabolous insects like P. apterus [105, 319], R. pedestris [117] and A. pisum [46]. In contrast, 130 CHAPTER 5. DISCUSSION

holometabolous insects often exhibit amplitudes more that 3-fold [46, 50, 117, 254, 308] and up to 7-fold in A. mellifera cry2 [269]. The weak circadian modulation in hemimetabolous insects was suggested to be caused by peripheral tissue rhythms out of phase with the neuronal pacemakers when large compartments of the heads were used [105, 117]. However if out-of- phase oscillators are responsible for the observed low amplitude rhythms, differently phased pacemakers must reside in the neuronal tissue itself, since precisely explanted neuronal tissues lacking peripheral tissues like photoreceptor cells or the neurohemal complex were used in this study.

5.2.2 Daily expression rhythms in DD Cycling of rmTim1 and rmCry2 continued the first day in DD, which indicates that these genes are very likely circadian clock genes in R. maderae. However, cycling in constant darkness is no absolute proof that a gene is part of the core set of circadian genes necessary for circadian rhythmicity, since oscillations can also be observed in genes directly regulated by the circadian machinery. Significant differences for rmPer cycling on the first day in constant darkness could not be observed. The amplitude of rmPer cycling was already very low in LD, and one possible explanation for the insignificant cyling in constant darkness is that the amplitude further dampened due to uncoupling oscillators in the brain.

5.2.3 Core feedback loop gene expression in the AMe The neurons associated with the AMe have been shown to contain the pacemaker for circadian locomotor rhythmicity in cockroaches [252]. Explanted tissues containing the AMe cycled in the same phase as the rest of the supraesophageal ganglion for all three genes examined. Not all cells seem to be phase-locked, since variation between pools were higher in AMe containing tissues, especially in rmPer. This may be caused by different circadian pacemaker cells in the AMe which cycle out of phase and was most probably not an artefact, since the same pool grouped well with the rest of the data in rmTim1 and rmCry2 and was measured on the same qPCR well plate. Although the majority of cells in the excised piece of tissue containing the AMe are not PDF-ir, many of those cells are probably associated with the AMe. The possibility cannot be ruled out that the rhythmsobserved in the AMe tissues are not primarily the rhythms of the PDF-ir neurons. However, the supraesophageal tissues with the AMe region excised did deﬁnitely not contain any PDFMe-ir neurons, so that the rhythms observed in the brain tissues with the AMe excised are rhythms of neuronal pacemakers located outside the pacemaker controlling locomotor rhythms in R. maderae. Since these tissues did not contain photoreceptor cells, other peripheral tissues or the neurohemal organ, other pacemakers must reside in the supraesophageal ganglion outside the AMe. In addition to the PDF-ir neurons associated with the AMe, PDF-ir neurons also reside in two clusters of cells (dPDFLa and vPDFLa) in the lamina [109, 234]. In D. melanogaster it has been shown that all PDF-ir neurons rhythmically express clock genes, and it is likely that the PDFLa neurons also express all genes necessary for circadian pacemakers. However, it has never been directly shown in R. maderae that PDF-ir neurons indeed are circadian pace- 5.2. CORE FEEDBACK LOOP GENE EXPRESSION RHYTHMS 131

makers controlling locomotor activity. Explanted tissues which restored locomotor rhythms also contained other neurons [252], and although recovery of circadian activity correlated with the regrowth of PDF-ir ﬁbres after the optic stalks were cut [312], other non-identiﬁed neurons may have also recovered. In addition to neurons glial cells have also been shown to express circadian genes in D. melanogaster [69, 316, 361], and glial cells of the Madeira cockroach brain may express circadian genes as well. This is supported by the widespread PER-ir in supraesophageal ganglion, where the majority of cells express rmPer.

5.2.4 Core feedback loop gene expression in Malpighian tubules

To examine the expression of core feedback loop genes outside the neuronal pacemakers, malpighian tubules were probed the same way as the supraesophageal ganglion. No signiﬁcant differences in expression levels of rmPer, rmTim1 and rmCry2 between timepoints were found. Compared to actin, their expression level was even higher than in neuronal tissue. However, since it is unknown how expression levels of actin differ between tissues in R. maderae, these values cannot be compared between tissues. The lack of circadian oscillation in Malpighian tubules is similar to G. bimaculatus, where rhythmic expression of neither per nor tim1 could be observed [336]. In contrast, per and tim1 expression is rhythmic in Malpighian tubules of D. melanogaster [81]. This hints to yet another tissue speciﬁc non-circadian role of the clock genes in some insects. Indeed it has been shown in P. apterus, that clock genes can attain other roles in some tissues independent of the central circadian clock [11]. Another explanation could be that expression of these genes cycles in a circadian manner also in Malpighian tubules, but differently phased oscillators reside in the sampled tissue, resulting in a seemingly constitutive expression.

5.2.5 The total expression level compared to actin

When observing the daily mean of expression levels compared to the reference gene actin, the expression level of rmTim1 was much lower (<2 % of actin expression) than the expression level of rmPer (<10 % of actin expression) and rmCry2 (∼4 % of actin expression). Nevertheless it exhibited the highest amplitude cycling. This indicates that rmTim1 may be expressed in less cells and may be more tightly regulated by light than rmPer or rmCry2. RmPer and rmCry2 expression may be more widespread than rmTim1 expression. Indeed immunstainings showed that rmPER appears to be expressed in the majority of CNS cells. The role of TIM1 is diverse among insect species. This was demonstrated with knock-down experiments in G. bimaculatus and Thermobia domestica. Whereas rmTim1 is an essential component for circadian rhythmicity in T. domestica [129], it seems to be dispensable in G. bimaculatus [53]. The fragment of rmTim1 that was cloned shares the highest identity with P. americana TIM1 (93 %), followed by G. bimaculatus (75 %) and T. domestica TIM1 (74 %). However, with the experiments performed and only part of rmTim1 cloned, it cannot be decided whether rmTim1 is an essential component of the circadian clock or merely has a modulatory role. 132 CHAPTER 5. DISCUSSION

5.2.6 Photoperiod dependent plasticity

Experiments using shorter or longer photophases than 12h revealed that the phase of core feedback loop gene expression adjusted relative to the onset of the scotophase. This mechanism is similar to D. melanogaster [173, 246] and two other flies, S. bullata and P. terraenovae [86, 195], where the phase of per and tim1 mRNA cycling also adjusts relative to the onset of the scotophase. This adjustment can be explained with the circadian clock’s light entrainment: In D. melanogaster, light induces CRY1 mediated degradation of TIM1, which delays TIM1 accumulation and subsequently produces a phase delay. Per and tim1 mRNA peaks hardly run out of phase, which can be easily explained by both genes’ expression being regulated by the same elements. The adjustment of the peak relative to the onset of the scotophase is interesting in terms of photoperiodic models, since this is a requirement for the external coincidence model. If a photoinducible phase in the later scotophase is indeed responsible for a photoperiodic response, the phase adjustment relative to the beginning of the scotophase is a coercive requirement. In S. bullata, tim1, but not per, mRNA levels dampened in longer photoperiods, whereas per and tim1 mRNA peak levels were unaffected in D. melanogaster. In hemimetabolous insects expression levels were also lower in long days like in P. apterus [60] or A. pisum [46] or remained unchanged as in R. pedestris [117]. In contrast, the peak amplitude of rmPer, rmTim1 and rm- Cry2 as well as the daily mean of expression levels was higher in long days compared to short days. However this apparent conflict may be none: In P. apterus for example insects ”destined- to-diapause” were monitored. The strong changes in per mRNA levels may reflect changes in physiological conditions because of the animals going into diapause rather than being a direct consequence of the changing photoperiod [60]. In R. maderae, the higher mRNA levels in long day conditions may be explained with the slower protein accumulation and overall lower protein levels. Since the proteins of the core feedback loop genes per, tim1 and cry2 are known to inhibit their own expression in other insects, lower mean protein levels may result in an increased expression rate. In addition to higher mRNA levels in long days, daily mean mRNA levels also significantly increased when transferred to constant darkness from LD 6:18 (short day) in rmPer and rmTim1. However, they significantly decreased upon transfer from LD 18:6 (long day) to constant darkness. mRNA levels in light-dark cycles and otherwise constant conditions (like temperature) are influenced by light and by the endogenous feedback system itself. If moved to constant darkness from different photoperiods, the mean daily mRNA level should move towards its endogenous state. If light indeed drives rmPer, rmTim1 and rmCry2 expression, the drop in mRNA levels when the animals were transfered to constant darkness from long day can be explained. How- ever, the significant increase in rmPer and rmTim1 expression levels after transfer of animals from short day conditions to constant darkness can not be explained by lower light input.

In summary, rmPer, rmTim1 and rmCry2 cycle in phase in light-dark cycles with maxima in the ﬁrst half of the scotophase similar to most insects examined so far. The phase adjusts relative to the onset of the scotophase, providing the basis for photoperiodic plasticity. Cycling continues in constant conditions except for rmPer due to the low amplitude cycling which probably damp- ens quickly in constant darkness due to uncoupling of pacemakers. Explanted AMe tissue as well 5.3. CIRCADIAN RHYTHMS ON PROTEIN LEVEL 133 as the remaining supraesophageal ganglion cycle in phase. This suggests additional pacemakers outside the known locomotor activity controlling AMe pacemaker. Neither rmPer, nor rmTim1 or rmCry2 cycle in Malpighian tubules in contrast to D. melanogaster, suggesting a different function or regulation in this tissue.

5.3 Circadian rhythms on protein level

To determine, whether clock gene mRNA and protein abundance cycle with a speciﬁc phase difference as in D. melanogaster, time series were performed with rmPER and rmTIM1 speciﬁc antibodies.

5.3.1 PERIOD

Specificity of the PER antibodies Of the bands observed in western blots, the signal above 120kDa probably represents full length rmPER. Depending on the marker, the antigen represented by this band has a molecular weight of slightly more than 120 kDa or approximately 130-140kDa, which corresponds well with the calculated molecular weight of 138kDa. An antigen of this size was detected by the anti rmPER- S peptide (mouse), the anti A. pernyi PER-S peptide and the anti rmPER (rabbit) antibody. The secondary antibody control without primary antibody did not give any signal at all, showing that none of the epitopes are recognized unspecifically directly by the secondary antibody. Preincu- bation of the anti rmPER (rabbit) antibody with peptide 3, which was mainly responsible for the immunoreaction of the immunized animals, abolished the signal almost completely, but other bands were also reduced.This shows that the antibody recognizes this specific peptide. Further evidence that the protein detected by the antibody is indeed rmPER would be given with knockout mutants or RNAi experiments. Since there is no genome or transcriptome available, not even of another cockroach species and since R. maderae is a ovoviviparous species, knock-out mutants are not generated easily. However, RNAi was already successfully employed on other hemimetabolous such as G. bimaculatus [190, 334, 335] and R. pedestris [116, 118, 119] and may prove to be a valuable tool in the future. Due to no genome or transcriptome available it can also not be tested if the epitope that was recognized by the antibody is also present on another R. maderae protein with the approximate size of rmPER. However, it is not very likely. A BLAST search using peptide 3 against all non-redundant GenBank coding sequences resulted exclusively in insect PER proteins with E-values below ten. In P. americana, the anti A. pernyi PER-S antibody only detected an antigen of approximately 110 kDa, which may correspond to the putative full-length rmPER signal [371]. An antibody directed against the D. melanogaster full-length PER detected two antigens with a molecular weight of approximatly 70 and 80 kDa [371]. In R. maderae, all antibodies used in western blots also detected one or two bands of this size in brain tissue homogenates. Although this may be a crossreacting protein, it is conspicuous that this antigen is being detected using antibodies directed against three different peptides (A. pernyi and D. melanogaster/cockroach PER-S and peptide 3) and full length dmPER in cock- 134 CHAPTER 5. DISCUSSION

roaches (Fig. 4.22, 4.23, A.8, A.9, and Fig. 2 in [371]). Possibly, this lower molecular weight epitope recognized by various anti PER antibodies is a shorter isoform or PER cleavage product, or a second PER is encoded in the genome of cockroaches (males and females), in addition to the ∼120kDa PER protein. Such a second per gene was for example discovered in females of A. pernyi [88].

Daily rhythms in PER abundance Although four western blots suggested protein maxima in the scotophase, six did not. How- ever, as explained in the results section, even when the same homogenates were used in different western blot experiments results were different. Overall results were not consistent and slight differences between Zeitgeber times were probably due to different loading or other variations of the method. Something similar was perviously observed in a diploma thesis that used an antibody directed against D. melanogaster PER with R. maderae brain extracts [368]: Signals at 190, 120 and 70 kDa were observed, which probably correlate with the signals observed in this work with the anti rmPER (rabbit) antibody above 150, above 120 and at approximately 70 kDa respectively. In time series only two out of nine western blots exhibited differences in staining intensities in the two higher molecular weight signals. Even when differences were observed, they were very faint (cf. Fig. 20 and 21 in [368]). Like for the qPCR experiments, explanted brains were used for the western blot experiments. Ten isolated brains were homogenized for each time point, and only undamaged complete brains were used. The Pierce 660nm kit and a Nanodrop ND-1000 were used to determine overall protein concentrations in the homogenates. Although the method itself worked as proven by standard curves, it turned out to be less precise than loading equal volumina of homogenate. When loading the amounts as suggested by the 660nm kit, sometimes differences between Zeitgeber times were suggested that were exactly inverse to the determined total protein concentration, but correlated with the loaded homogenate volumina. When using the same volumina, there was no obvious oscillation. There was no difference in reference protein signal strength (TYR-TUB) between time points when using the same total protein amounts as determined by the kit, or the same homogenate volumina. With the amount of total protein loaded needed to detect the proteins of interest, TYR-TUB concentration in the homogenates is probably much higher than the concentration of the circadian proteins. This results in the membrane being saturated with the protein and the x-ray ﬁlm being maximally exposed. Thus smaller differences in total protein concentration are not visible. This is a general problem with reference proteins and can be observed in many published western blots. If there is a daily oscillation in protein concentration in rmPER, it may be undetectable with western blots due to the very low amplitude, that already shows on mRNA level. However, comparisons between mRNA and protein oscillations have to be estimated very carefully, since various posttranscriptional and posttranslational events can uncouple mRNA from protein concentration rhythms. In D. melanogaster, PER is consecutively phosphorylated and PER bands thus shift in a circadian manner [63]. Although several clustered bands were sometimes observed that may represent a different phosphostatus of the protein, there was no obvious circadian shift in R. maderae. PER concentration rhythms in total head or brain extracts were found in D. melanogaster [63], 5.3. CIRCADIAN RHYTHMS ON PROTEIN LEVEL 135

but already other ﬂies do not exhibit PER rhythms: per mRNA cycles in Ceratitis capitata and M. domestica, but PER is constitutively detected in western blots of these species [40, 182]. Also in the marine mollusk Bulla gouldingiana PER abundance does not appears to cycle in a daily manner in the eye, gut and ganglia [44]. There are two obvious explanations for the lack of rhythmic PER concentrations: Either, PER abundance within each cell does indeed not cycle, or differently phased oscillators result in undetectable ryhthms when collectively observed. If there is indeed no circadian oscillation of PER protein, PER may attain a different role in the circadian system of some animals. Maybe, CRY2 is the main rhythmic repressor and PER merely serves as some kind of constitutive present co-factor. More likely however, PER protein concentrations cycle in putative pacemaker cells, and either some pacemakers are differently phased, or PER expression is not rhythmic in some cells. Considering, that the core feedback loop transcriptional activators and repressors regulate the expression of much more genes than only the clock genes (more than 1000 in D. melanogaster), cell-speciﬁc regulation may be necessary in many cells. In addition, since several individuals were pooled, the population may not be perfectly synchronized. Both would result in either very low expression rhythms as observed in mRNA time-series, or even no rhythms as observed in western blots.

5.3.2 TIMELESS1

Specificity of the rmTIM1 antibody Although several epitopes were recognized in high concentrations of the rmTIM1 antibody in western blots, it recognized a single epitope of approximately 150-200kDa in higher dilutions (Fig. 4.24, B). The antibodies raised against the peptide were a specific reaction to the immunization, as shown by the western blots using the preimmunsera (Fig. 4.24, A). The full length coding sequence of rmTim1 could not be cloned, but the molecular weight of the recognized epitope corresponds well with the molecular weight of the longer isoforms of D. melanogaster TIM1 (∼156kDa) and of A. pisum TIM1 (∼168 kDa). However, the circadian clocks of these two species are rather derived, and the TIM1 protein of R. maderae should more likely be com- parable to more basic insects like G. bimaculatus and T. domestica. Indeed, a DELTA-BLAST database search revealed that the aminoacid sequence of the cloned rmTIM1 fragment has the highest similarity with these two species after another cockroach, P. americana. The TIM1 protein of both, G. bimaculatus and T. domestica though has a much lower calculated molecular weight (G. bimaculatus: 134kDa [53]; T. domestica: 136kDa [129]) than D. melanogaster or A. pisum TIM1. If the antibody indeed recognized rmTIM1, the protein may either be larger than these species’ TIM1, or it just appears to be larger on western blots due to some kind of posttranslational modification.

Daily rhythms in TIM1 abundance Interestingly, similar to PER, no consistent rhythms could be detected. Either no rhythms were observed, or there were very faint differences between Zeitgeber times that suggested different 136 CHAPTER 5. DISCUSSION

maxima. In these cases, the background was also higher, indicating that these slight differences are due to unequal loading or other variability of the method (Fig. 4.24 B). In D. melanogaster, TIM1 is almost undetectable during the light phase, and phase resetting is thought to be mainly achieved by light input to the core feedback loop via CRY1 and TIM1. Also low sensitivity to light during the day is explained by the very low or undetectable TIM1 levels. However, in R. maderae, TIM1 can clearly be detected during the photophase, with no or little difference compared to the scotophase. This indicates, that there is no direct light input via CRY1 or other photosensitive proteins to TIM1 at least in the majority of cells, since TIM1 levels are not lower during the photophase. Either, as explained for PER, differently phased cells, or cells constitutively expressing TIM1 mask a putative oscillation of the protein, or light input and phase reset are not TIM1 dependently mediated in R. maderae.

In summary, neither oscillation of rmPER, nor of rmTIM1 could be observed in R. maderae brain tissue on protein level. Although this was also observed in PER abundance of other insects, it is unlikely that both genes do not cycle on protein level in circadian pacemaker cells. Most probably either masking occurred or the antibodies did only detect some forms of the respective protein.

5.4 Distribution of PER-ir cells in the central nervous system

Using the anti rmPER (rabbit) antibody, widespread but exclusively nuclear staining was observed. The staining was nuclear at all Zeitgeber times observed, and similar to the western blots, no differences between the monitored Zeitgeber times could be detected. Widespread nuclear, but also cytoplasmatic staining in some cells was observed with the polyclonal full length anti dPER antibody in P. americana [371]. In contrast, the anti A. pernyi PER-S antibody exclusively recognized cytoplasmatic staining in P. americana [371]. This correlates with the results obtained in R. maderae, since the PER-S peptide did hardly evoke any immunoreaction in the rabbits used for immunization, and no cytoplasmatic staining was observed. The aminoacid sequence of the D. melanogaster PER-S peptide is identical with the R. maderae PER-S peptide, and the A. pernyi anti PER-S antibody recognizes the same cells and compartments as the D. melanogaster anti PER-S antibody (Saumann unpublished, mentioned in [371]). Thus, both PER-S peptide antibodies most probably recognize the same epitope, which is not that surprising, since the D.melanogaster/R. maderae PER-S peptide differs from the A. pernyi PER-S peptide by only one aminoacid. That the polyclonal anti rmPER (rabbit) serum did hardly contain any antibodies directed against the PER-S peptide was also conﬁrmed by preincubation experiments: The PER-S peptide did not abolish PER-ir (Fig. 4.29), and preincubation with peptide 3 alone strongly weakened the signal of all bands in western blots. In P. americana, the anti A. pernyi PER-S antibody recognized a single band at approximately 110kDa, whereas an antibody directed against the complete D. melanogaster PER protein exclusively recognized two bands at ∼70 and ∼80 kDa [371]. Bands of a similar molecular weight were also present in western blots using the rmPER antibody. These results suggest that the nuclear PER-ir observed represents the protein(s) detected at ∼80 kDa by western blots in cockroaches. Since full-length rmPER has 5.4. DISTRIBUTION OF PER-IR CELLS IN THE CENTRAL NERVOUS SYSTEM 137

a calculated molecular weight of 138 kDa and the antibody also recognizes an epitope at this molecular weight, one may suspect that the nuclear PER-ir represents cross-reactivity in cockroaches. However, the anti full-length dmPER antibody also stained the cytoplasm in some P. americana cells, indicating that the protein represented by the band at approximatly 80 kDa also resides in the cytoplasm of some cells. In addition, the rmPER antibody also recognized a higher molecular protein above 120 kDa, which most probably represents full-length PER.

5.4.1 PER-ir in different Zeitgeber times In P. americana, the cytoplasmatic staining did not show any circadian changes. However, the nuclear staining cycled with minima during the night. Oddly, the nuclear staining abolished during constant darkness indicating that PER levels are driven by light [371]. In contrast, there was no difference in nuclear staining between different Zeitgeber times in R. maderae (Fig. 4.32, 4.33). In B. germanica and its sister species B. bisignata, an antibody derived against a peptide of the conserved C5 region stained exclusively the cytoplasm of distinct cell groups in the optic lobe, proto-, deuto- and tritocerebrum, and SOG. Unfortunatly only one Zeitgeber time was probed. Thus, it is unknown whether PER-ir cycles in these cells [342]. However, the location of the groups is very similar to the groups showing cytoplasmatic staining observed in P.americana, which did not show any differences in different Zeitgeber times [371]. One explanation for the different staining patterns with various antibodies in the four cockroach species may be due to the protocol used. As noted before by Sandrelli et al. [276], reports on exclusively cytoplasmatic staining when using enzyme based immunohistochemistry instead of fluorescent methods may have to be assessed carefully. In M. domestica, enzyme based immunohistochemistry revealed only cytoplasmatic staining, but fluorescent immunohistochemistry in combination with confocal microscopy in the same species revealed nuclear staining in PER and TIM1 colocalizing neurons and ZT dependent translocation [40]. Another probable explanation is possible antigen masking. The tertiary structure, various posttranslational modifications and complex formation may mask the antigens used for immunization. If antibodies derived against single peptides were used, either the nucleus or the cytoplasm was stained. In contrast, the antibody derived against the full length D. melanogaster proteins stained cell nuclei as well as the cytoplasm. However, this cannot be the only explanation for the observed staining patterns. Even with the antibody that stained both, the cytoplasm and the nuclei in P. americana, no daytime dependent nuclear import was observed. The PER-ir either stayed nuclear or cytoplasmatic. As described before, the nuclear epitope may be linked to the lower molecular weight epitope observed at approximately 80kDa in western blots, and the cytoplasmatic to the higher molecular weight epitope which may represent full-length PER. If the lower molecular weight epitope and nuclear staining respectively indeed represents a crossreacting protein, the question arises why only the nuclear staining cycles in P. americana, and neither cycling nor nuclear import was observed within the cellgroups showing cytoplasmatic staining. Taken together, anti PER antibodies either stain an epitope in the nucleus of almost all neuronal and glial cells, distinct cell groups in the cytoplasm which appear to be identical in different cockroach species, or both. Cycling was only observed in nuclear staining in LD in P. ameri- 138 CHAPTER 5. DISCUSSION

cana, and no nuclear translocation was ever observed. In summary, several explanations (or a mixture thereof) are possible for the exclusively nuclear PER-ir in R. maderae and the lack of circadian oscillation:

1. The nuclear or cytoplasmatic staining represents a crossreacting protein in some species and only some forms of PER or PER complexes are recognized by the antibody.

2. There may be two different forms of PER genes encoded in the genome of cockroaches which cycle in antiphase.

3. PER undergoes some form of tissue speciﬁc expression or modiﬁcation, and is indeed almost constitutively expressed in cockroaches. This is also indicated by western blots as well as quantitative PCR, where expression rhythms were very low or absent. CRY2 may be the major repressor in R. maderae, and maybe PER has the role of a co-factor in the nucleaus of the majority of cells. Indeed there are some indications for such a role in D. plexippus DpN1 cells. CRY2 alone potently inhibits E-box driven expression in this system, whereas PER does not. Also PER stays >90% nuclear in this cellular context (unpublished, mentioned in [366]). However, in other cells, D. plexippus PER can repress E-box driven expression [366] indicating that the function may depend on the cellular background.

In summary, rmPER appears to occur in all cells of the central nervous system exclusively in nuclei. Similar results were observed in P. americana, however in contrast to this species rmPER did not cycle. This is different to almost all other insects examined so far and suggests a different role and regulation of rmPER in the majority of cells in the central nervous system.

5.5 Locomotor activity in different photoperiods

Running wheel assays showed, that cockroaches kept in low light conditions (∼100lx) did not exhibit any significant differences in free-running period length between the two photoperiods examined. In contrast, animals transferred from ∼100lx to bright light conditions (∼1000 lx) for one week did express significantly longer free-running rhythms when kept in LD6:18, and shorter period lengths when kept in LD18:6. One explanation, why the animals exhibit significant effects only in higher light intensities could be the range of entrainment. Light pulse length and light intensity have similar effects, and the range of entrainment becomes greater with higher light intensities [286]. Compared to other insects, R. maderae requires rather high intensities of light to evoke phase shifts [221, 347]. Maybe, 100lx is not enough to produce significant effects. Light intensity also influences the coupling strength [343]. If there are two different oscillators, one with a short and one with a long τ, one may dominate in short day and one in long day, and higher light intensities can enhance coupling. Thus, the respective ”dominant” oscillator can couple the weaker better in higher light intensities, which leads to a stronger effect in overt rhythms. 5.5. LOCOMOTOR ACTIVITY IN DIFFERENT PHOTOPERIODS 139

In R. maderae, there seem to be two different physiological mechanisms that influence the free- running period length: First, the total period length of the light regime during the larval development influences the adult free-running period: longer T during larval development results in longer free-running τ in adult animals [221]. Second, entrainment to different period lengths in adults results in transient ”after-effects” in subsequent free-running periods. It seems to be especially the latter mechanism that produced the observed effects. Since both groups grew up in 100lx, period length shortening or lengthening in animals kept in higher light intensities is an effect of adult plasticity and no permanent developmental effect. Also in animals entrained with low light intensities, the free-running period length of some animals changed after some time in constant conditions, which indicates that the observed photoperiodic effects were only transient. This change in period length in DD was not observed in higher light intensities, which indicates that the duration of the photoperiodic response depends on the entrainment light intensity. Since no change was observed in higher light intensities, it can not be concluded that the effect is transient. However, only two weeks of constant conditions after entrainment were monitored, and after-effects can last up to four weeks [225]. Thus, the photoperiodic entrainment effect may have worn off after more than two weeks in constant conditions. The original paper describ- ing after effects of entrainment used two different photoperiods: LD8:14 and LD8:18, and the free-running period length was longer in LD8:18. Since the photophase was kept constant, this indicates that the night length determines the free-running period. This corresponds well with the effects observed in R. maderae, since the period length was also significantly longer when the scotophase of the lightregime used for entrainment was prolonged. The mechanism underlying the photoperiodic effect may be explained as follows: In long days (or in other words: more light), more PDF accumulates. This is reflected by the number of PDF-ir neurons [340, 131]. Increased PDF signaling then leads to an increase in phase delays, resetting the phase to peak in the dark phase despite long days (like in the qPCR experiments). In addition, PDF may serve as some kind of ”photoperiodic counter” as it takes 2-6 weeks for PDF to accumulate to amounts that produce visible effects in immunostainings. Above a certain threshold, PDF then triggers photoperiodic output. The amount of PDF also influences locomotor behaviour, since more PDF leads to increased phase delays, which is also reflected in a longer free-running period. However, the effect of the photoperiod on τ in running wheel behaviour was either none (100lx), or the exact opposite of what would be expected, if phase delays by PDF are indeed responsible for longer free-running periods. Maybe, PDF determines the magnitude of the response, because effects were only observed in higher light intensities and more cells are stained in longer photophases also, and the night length determines the direction of the response.

In summary, the photoperiodic plasticity also manifested on behavioural level: longer photophases decreased period length and higher light intensities were necessary for a signiﬁcant effect. The period length appears to depend on the scotophase, whereas light intensity appears to regulate the strength of the period length change. 140 CHAPTER 5. DISCUSSION

5.6 Conclusion and future perspectives

This thesis identified the first circadian clock genes of the Madeira cockroach Rhyparobia maderae and showed circadian cycling of the mRNA of the core feedback loop genes period, timeless 1, and cryptochrome 2. The domain structure and daily expression profile of these genes were analyzed. The findings show that the circadian clock of the Madeira cockroach belongs to the basic type of insect clocks, which is more closely related to vertebrate clocks than to the derived clocks of D. melanogaster and A. pisum. No statistically significant circadian oscillation in protein abundance of neither rmPER nor rmTIM1 were observed in western blots and immunohistochemical analysis of whole brain tissues. Instead, rmPER appears to stay nuclear at all times in R. maderae nervous tissue cells. Most likely, no rmPER-protein oscillations were observed due to low amplitude cycling and because oscillations in different circadian pacemaker cells were not phase-coupled. In addition, the anti rmPER (rabbit) antibody appeared to only recognize the nuclear form of the PER protein. In the future, it is important that the genome of R. maderae is being sequenced. Various insect genomes are already published, but there is still no cockroach genome available. Although the genome of B. germanica has been sequenced recently, it is not available to the public at this time. Nevertheless, since the Madeira cockroach is an important model organism of circadian research, it will be necessary to obtain the genome of R. maderae in the near future. With the full genome available, it will be much easier to identify the complete set of circadian genes, design specific primers and analyze their expression. In addition to the genome, transcriptome data will also help to unravel the circadian system of cockroaches further. RNA-Seq may be employed instead of qPCR to obtain the circadian expression profile of various genes at once. As a next step towards analyzing the function of the circadian genes in R. maderae, gene silencing employing RNAi should be used to analyze the function of the circadian genes. Although some functional data on hemimetabolous insects has already been published, for example by the lab of Kenji Tomioka and coworkers, there is still no single hemimetabolous insect species with at least the core feedback loop genes functionally analyzed. Since it became more and more obvious in the last decade that there are significant differences between insect species, it will be very interesting to understand how the molecular system of the circadian clock works in a more basic insect. mRNA sequence data of rmPer, rmTim1 and rmCry2 is now available, and gene silencing using RNAi in combination with running wheel assays may clarify the function of these genes. In summary, this study provides the first data on the molecular circadian system of the cockroach R. maderae. In the future, a sequenced genome or transcriptome will help to characterize the complete set of circadian genes in R. maderae, and functional studies using RNAi will be used to clarify the function of clock genes in the Madeira cockroach. In addition, classical methods like RACE and qPCR will be used to analyze individual genes. With more sequence data available, specific antibodies against additional circadian genes like CLK or CYC can be generated to further analyze the cellular expression pattern and subcellular localization. Bibliography

[1] ABRUZZI,K.C.,RODRIGUEZ,J.,MENET,J.S.,DESROCHERS,J.,ZADINA,A.,LUO, W., TKACHEV,S., AND ROSBASH, M. Drosophila CLOCK target gene characterization: implications for circadian tissue-speciﬁc gene expression. Genes Dev 25, 22 (Nov 2011), 2374–2386.

[2] AKTEN,B.,JAUCH,E.,GENOVA,G.K.,KIM, E. Y., EDERY,I.,RAABE, T., AND JACKSON, F. R. A role for CK2 in the Drosophila circadian oscillator. Nat Neurosci 6, 3 (Mar 2003), 251–257.

[3] ALLADA,R., AND CHUNG, B. Y. Circadian organization of behavior and physiology in Drosophila. Annu Rev Physiol 72 (Mar 2010), 605–624.

[4] ALLADA,R.,WHITE,N.E.,SO, W. V., HALL,J.C., AND ROSBASH, M. A mutant Drosophila homolog of mammalian Clock disrupts circadian rhythms and transcription of period and timeless. Cell 93, 5 (May 1998), 791–804.

[5] ALTSCHUL, S. F., MADDEN, T. L., SCHAFFER¨ ,A.A.,ZHANG,J.,ZHANG,Z.,MILLER, W., AND LIPMAN, D. J. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25, 17 (Sep 1997), 3389–3402.

[6] ALTSCHUL, S. F., WOOTTON,J.C.,GERTZ,E.M.,AGARWALA,R.,MORGULIS,A., SCHAFFER¨ ,A.A., AND YU, Y.-K. Protein database searches using compositionally adjusted substitution matrices. FEBS J 272, 20 (Oct 2005), 5101–5109.

[7] ASCHOFF, J. Circadian timing. Ann N Y Acad Sci 423 (1984), 442–468.

[8] BACHLEITNER, W., KEMPINGER,L.,WULBECK¨ ,C.,RIEGER,D., AND HELFRICH-FORSTER¨ , C. Moonlight shifts the endogenous clock of Drosophila melanogaster. Proc Natl Acad Sci U S A 104, 9 (Feb 2007), 3538–3543.

[9] BAE,K., AND EDERY, I. Regulating a circadian clock’s period, phase and amplitude by phosphorylation: insights from Drosophila. J Biochem 140, 5 (Nov 2006), 609–617.

[10] BAE,K.,LEE,C.,SIDOTE,D.,CHUANG, K. Y., AND EDERY, I. Circadian regulation of a Drosophila homolog of the mammalian Clock gene: PER and TIM function as positive regulators. Mol Cell Biol 18, 10 (Oct 1998), 6142–6151.

[11] BAJGAR,A.,JINDRA,M., AND DOLEZEL, D. Autonomous regulation of the insect gut by circadian genes acting downstream of juvenile hormone signaling. Proc Natl Acad Sci U S A 110, 11 (Mar 2013), 4416–4421.

[12] BAO,S.,RIHEL,J.,BJES,E.,FAN, J. Y., AND PRICE, J. L. The Drosophila double-timeS mutation delays the nuclear accumulation of period protein and affects the feedback regulation of period mRNA. J Neurosci 21, 18 (Sep 2001), 7117–7126.

[13] BARRETT,R.K., AND PAGE, T. L. Effects of light on circadian pacemaker development. I. The freerunning period. J Comp Physiol A 165, 1 (Apr 1989), 41–49.

141 142 BIBLIOGRAPHY

[14] BAYLIES,M.K.,WEINER,L.,VOSSHALL,L.B.,SAEZ,L.,A., AND YOUNG, M. W. Genetic, molecular, and cellular studies of the per locus and its products in Drosophila melanogaster. In: M.W Young (Ed.), Molecular Genetics of Biological Rhythms, Marcel Dekker, New York. (1993), 123–153.

[15] BEAVER,L.M.,GVAKHARIA,B.O.,VOLLINTINE, T. S., HEGE,D.M.,STANEWSKY,R., AND GIEBULTOWICZ, J. M. Loss of circadian clock function decreases reproductive ﬁtness in males of Drosophila melanogaster. Proc Natl Acad Sci U S A 99, 4 (Feb 2002), 2134–2139.

[16] BENITO,J.,HOXHA, V., LAMA,C.,LAZAREVA,A.A.,FERVEUR, J.-F., HARDIN, P. E., AND DAUWALDER, B. The circadian output gene takeout is regulated by Pdp1 epsilon. Proc Natl Acad Sci U S A 107, 6 (Feb 2010), 2544–2549.

[17] BENITO,J.,ZHENG,H.,NG, F. S., AND HARDIN, P. E. Transcriptional feedback loop regulation, function, and ontogeny in Drosophila. Cold Spring Harb Symp Quant Biol 72 (2007), 437–444.

[18] BENNA,C.,BONACCORSI,S.,WULBECK¨ ,C.,HELFRICH-FORSTER¨ ,C.,GATTI,M.,KYRIA- COU, C. P., COSTA,R., AND SANDRELLI, F. Drosophila timeless2 is required for chromosome stability and circadian photoreception. Curr Biol 20, 4 (Feb 2010), 346–352.

[19] BLAU,J., AND YOUNG, M. W. Cycling vrille expression is required for a functional Drosophila clock. Cell 99, 6 (Dec 1999), 661–671.

[20] BODENSTEIN,C.,HEILAND,I., AND SCHUSTER, S. Temperature compensation and entrainment in circadian rhythms. Phys Biol 9, 3 (Jun 2012), 036011.

[21] BORATYN,G.M.,SCHAFFER¨ ,A.A.,AGARWALA,R.,ALTSCHUL, S. F., LIPMAN,D.J., AND MADDEN, T. L. Domain enhanced lookup time accelerated BLAST. Biol Direct 7 (2012), 12.

[22] BORSON,N.D.,SALO, W. L., AND DREWES, L. R. A lock-docking oligo(dT) primer for 5’ and 3’ RACE PCR. PCR Methods Appl 2, 2 (Nov 1992), 144–148.

[23] BRADSHAW, W. E., EMERSON,K.J., AND HOLZAPFEL, C. M. Genetic correlations and the evolution of photoperiodic time measurement within a local population of the pitcher-plant mosquito, Wyeomyia smithii. Heredity (Edinb) 108, 5 (May 2012), 473–479.

[24] BRADSHAW, W. E., HOLZAPFEL,C.M., AND MATHIAS, D. Circadian rhythmicity and photoperiodism in the pitcher-plant mosquito: can the seasonal timer evolve independently of the circadian clock? Am Nat 167, 4 (Apr 2006), 601–605.

[25] BRAND,A.H., AND PERRIMON, N. Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118, 2 (Jun 1993), 401–415.

[26] BUSZA,A.,EMERY-LE,M.,ROSBASH,M., AND EMERY, P. Roles of the two Drosophila CRYP- TOCHROME structural domains in circadian photoreception. Science 304, 5676 (Jun 2004), 1503– 1506.

[27] BUNNING¨ . Die endogene Tagesrhythmik als Grundlage der Photoperiodischen Reaktion. Berichte der Deutschen Botanischen Gesellschaft 54 (1936), 590–607.

[28] BUNNING¨ ,E. The physiological clock: circadian rhythms and biological chronometry. 1973. BIBLIOGRAPHY 143

[29] BUNSOW¨ ,R.C. Uber¨ tages- und jahresrhythmische Anderungen¨ der photoperiodischen Lichtempﬁndlichkeit bei Kalanchoe blossfeldiana und ihre Beziehungen zur endogenen Tages- rhythmik. Zeitschrift fur¨ Botanik 41 (1953), 257–276.

[30] CERIANI, M. F., DARLINGTON, T. K., STAKNIS,D.,MAS´ , P., PETTI,A.A.,WEITZ,C.J., AND KAY, S. A. Light-dependent sequestration of TIMELESS by CRYPTOCHROME. Science 285, 5427 (Jul 1999), 553–556.

[31] CERIANI, M. F., HOGENESCH,J.B.,YANOVSKY,M.,PANDA,S.,STRAUME,M., AND KAY, S. A. Genome-wide expression analysis in Drosophila reveals genes controlling circadian behavior. J Neurosci 22, 21 (Nov 2002), 9305–9319.

[32] CHANG,D.C.,MCWATTERS,H.G.,WILLIAMS,J.A.,GOTTER,A.L.,LEVINE,J.D., AND REPPERT, S. M. Constructing a feedback loop with circadian clock molecules from the silkmoth, Antheraea perny. J Biol Chem 278, 40 (Oct 2003), 38149–38158.

[33] CHANG,D.C., AND REPPERT, S. M. A novel C-terminal domain of Drosophila PERIOD inhibits dCLOCK:CYCLE-mediated transcription. Curr Biol 13, 9 (Apr 2003), 758–762.

[34] CHATTERJEE,A., AND HARDIN, P. E. Time to taste: circadian clock function in the Drosophila gustatory system. Fly (Austin) 4, 4 (2010), 283–287.

[35] CHATTERJEE,A.,TANOUE,S.,HOUL,J.H., AND HARDIN, P. E. Regulation of gustatory physiology and appetitive behavior by the Drosophila circadian clock. Curr Biol 20, 4 (Feb 2010), 300–309.

[36] CHAVES,I.,POKORNY,R.,BYRDIN,M.,HOANG,N.,RITZ, T., BRETTEL,K.,ESSEN,L.-O., VAN DER HORST, G. T. J., BATSCHAUER,A., AND AHMAD, M. The cryptochromes: blue light photoreceptors in plants and animals. Annu Rev Plant Biol 62 (Jun 2011), 335–364.

[37] CHAVES,I.,YAGITA,K.,BARNHOORN,S.,OKAMURA,H., VAN DER HORST, G. T. J., AND TAMANINI, F. Functional evolution of the photolyase/cryptochrome protein family: importance of the C terminus of mammalian CRY1 for circadian core oscillator performance. Mol Cell Biol 26, 5 (Mar 2006), 1743–1753.

[38] CHIU,J.C.,VANSELOW, J. T., KRAMER,A., AND EDERY, I. The phospho-occupancy of an atypical SLIMB-binding site on PERIOD that is phosphorylated by DOUBLETIME controls the pace of the clock. Genes Dev 22, 13 (Jul 2008), 1758–1772.

[39] CLARIDGE-CHANG,A.,WIJNEN,H.,NAEF, F., BOOTHROYD,C.,RAJEWSKY,N., AND YOUNG, M. W. Circadian regulation of gene expression systems in the Drosophila head. Neuron 32, 4 (Nov 2001), 657–671.

[40] CODD, V., DOLEZEL,D.,STEHLIK,J.,PICCIN,A.,GARNER,K.J.,RACEY,S.N.,STRAAT- MAN,K.R.,LOUIS,E.J.,COSTA,R.,SAUMAN,I.,KYRIACOU, C. P., AND ROSATO,E. Circadian rhythm gene regulation in the houseﬂy Musca domestica. Genetics 177, 3 (Nov 2007), 1539–1551.

[41] COLLINS,B.,MAZZONI,E.O.,STANEWSKY,R., AND BLAU,J. Drosophila CRYP- TOCHROME is a circadian transcriptional repressor. Curr Biol 16, 5 (Mar 2006), 441–449. 144 BIBLIOGRAPHY

[42] COLOT, H. V., HALL,J.C., AND ROSBASH, M. Interspeciﬁc comparison of the period gene of Drosophila reveals large blocks of non-conserved coding DNA. EMBO J 7, 12 (Dec 1988), 3929–3937.

[43] COLWELL,C.S., AND PAGE, T. L. The electroretinogram of the cockroach Leucophaea maderae. Comp Biochem Physiol A Comp Physiol 92, 1 (1989), 117–123.

[44] CONSTANCE,C.M.,GREEN,C.B.,TEI,H., AND BLOCK,G.D. Bulla gouldiana period exhibits unique regulation at the mRNA and protein levels. J Biol Rhythms 17, 5 (Oct 2002), 413–427.

[45] CORNWELL, P. G. The Cockroach: a laboratory insect and an industrial pest. Hutchinson of London, 1968.

[46] CORTES´ , T., ORTIZ-RIVAS,B., AND MART´INEZ-TORRES, D. Identiﬁcation and characterization of circadian clock genes in the pea aphid Acyrthosiphon pisum. Insect Mol Biol 19 Suppl 2 (Mar 2010), 123–139.

[47] COSTA,R.,PEIXOTO,A.A.,BARBUJANI,G., AND KYRIACOU, C. P. A latitudinal cline in a Drosophila clock gene. Proc Biol Sci 250, 1327 (Oct 1992), 43–49.

[48] CURRIE,J.,GODA, T., AND WIJNEN, H. Selective entrainment of the Drosophila circadian clock to daily gradients in environmental temperature. BMC Biol 7 (2009), 49.

[49] CURTIN,K.D.,HUANG,Z.J., AND ROSBASH, M. Temporally regulated nuclear entry of the Drosophila period protein contributes to the circadian clock. Neuron 14, 2 (Feb 1995), 365–372.

[50] CYRAN,S.A.,BUCHSBAUM,A.M.,REDDY,K.L.,LIN,M.-C.,GLOSSOP,N.R.J.,HARDIN, P. E., YOUNG, M. W., STORTI, R. V., AND BLAU, J. Vrille, Pdp1, and dClock form a second feedback loop in the Drosophila circadian clock. Cell 112, 3 (Feb 2003), 329–341.

[51] CYRAN,S.A.,YIANNOULOS,G.,BUCHSBAUM,A.M.,SAEZ,L.,YOUNG, M. W., AND BLAU, J. The double-time protein kinase regulates the subcellular localization of the Drosophila clock protein period. J Neurosci 25, 22 (Jun 2005), 5430–5437.

[52] CZARNA,A.,BERNDT,A.,SINGH,H.R.,GRUDZIECKI,A.,LADURNER,A.G.,TIMIN- SZKY,G.,KRAMER,A., AND WOLF, E. Structures of Drosophila cryptochrome and mouse cryptochrome1 provide insight into circadian function. Cell 153, 6 (Jun 2013), 1394–1405.

[53] DANBARA, Y., SAKAMOTO, T., URYU,O., AND TOMIOKA, K. RNA interference of timeless gene does not disrupt circadian locomotor rhythms in the cricket Gryllus bimaculatus. J Insect Physiol 56, 12 (Dec 2010), 1738–1745.

[54] DARLINGTON, T. K., WAGER-SMITH,K.,CERIANI, M. F., STAKNIS,D.,GEKAKIS,N., STEEVES, T. D., WEITZ,C.J.,TAKAHASHI,J.S., AND KAY, S. A. Closing the circadian loop: CLOCK-induced transcription of its own inhibitors per and tim. Science 280, 5369 (Jun 1998), 1599–1603.

[55] DELA PAZ FERNANDEZ´ ,M.,CHU,J.,VILLELLA,A.,ATKINSON,N.,KAY,S.A., AND CERI- ANI, M. F. Impaired clock output by altered connectivity in the circadian network. Proc Natl Acad Sci U S A 104, 13 (Mar 2007), 5650–5655. BIBLIOGRAPHY 145

[56] DE MAIRAN, J. J. Observation Botanique. Hist.Acad.R.Sci. (1729), 35–36.

[57] DEBRUYNE, J. P., WEAVER,D.R., AND REPPERT, S. M. CLOCK and NPAS2 have overlapping roles in the suprachiasmatic circadian clock. Nat Neurosci 10, 5 (May 2007), 543–545.

[58] DENLINGER,D.,YOCUM,G., AND RINEHART, J. Hormonal Control of Diapause. In Insect Endocrinology, L. I. Gilbert, Ed. Academic Press, San Diego, 2012, pp. 430 – 463.

[59] DISSEL,S.,CODD, V., FEDIC,R.,GARNER,K.J.,COSTA,R.,KYRIACOU, C. P., AND ROSATO, E. A constitutively active cryptochrome in Drosophila melanogaster. Nat Neurosci 7, 8 (Aug 2004), 834–840.

[60] DOLEZEL,D.,SAUMAN,I.,KOST’AL´ , V., AND HODKOVA, M. Photoperiodic and food signals control expression pattern of the clock gene, period, in the linden bug, Pyrrhocoris apterus. J Biol Rhythms 22, 4 (Aug 2007), 335–342.

[61] DOLEZELOVA,E.,DOLEZEL,D., AND HALL, J. C. Rhythm defects caused by newly engineered null mutations in Drosophila’s cryptochrome gene. Genetics 177, 1 (Sep 2007), 329–345.

[62] DOWSE,H.B., AND PALMER, J. D. Entrainment of circadian activity rhythms in mice by elec- trostatic ﬁelds. Nature 222, 5193 (May 1969), 564–566.

[63] EDERY,I.,ZWIEBEL,L.J.,DEMBINSKA,M.E., AND ROSBASH, M. Temporal phosphorylation of the Drosophila period protein. Proc Natl Acad Sci U S A 91, 6 (Mar 1994), 2260–2264.

[64] EMERSON,K.J.,DAKE,S.J.,BRADSHAW, W. E., AND HOLZAPFEL, C. M. Evolution of photoperiodic time measurement is independent of the circadian clock in the pitcher-plant mosquito, Wyeomyia smithii. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 195, 4 (Apr 2009), 385–391.

[65] EMERY, I. F., NOVERAL,J.M.,JAMISON, C. F., AND SIWICKI, K. K. Rhythms of Drosophila period gene expression in culture. Proc Natl Acad Sci U S A 94, 8 (Apr 1997), 4092–4096.

[66] EMERY, P., SO, W. V., KANEKO,M.,HALL,J.C., AND ROSBASH, M. CRY, a Drosophila clock and light-regulated cryptochrome, is a major contributor to circadian rhythm resetting and photosensitivity. Cell 95, 5 (Nov 1998), 669–679.

[67] EMERY, P., STANEWSKY,R.,HELFRICH-FORSTER¨ ,C.,EMERY-LE,M.,HALL,J.C., AND ROSBASH, M. Drosophila CRY is a deep brain circadian photoreceptor. Neuron 26, 2 (May 2000), 493–504.

[68] ENGELMANN, W., HELLRUNG, W., AND JOHNSSON, A. Circadian locomotor activity of Musca ﬂies: recording method and effects of 10 Hz square-wave electric ﬁelds. Bioelectromagnetics 17, 2 (1996), 100–110. Magneto.

[69] EWER,J.,FRISCH,B.,HAMBLEN-COYLE,M.J.,ROSBASH,M., AND HALL, J. C. Expression of the period clock gene within different cell types in the brain of Drosophila adults and mosaic analysis of these cells’ inﬂuence on circadian behavioral rhythms. J Neurosci 12, 9 (Sep 1992), 3321–3349. 146 BIBLIOGRAPHY

[70] FABRICIUS,J. Species insectorum exhibentes eorum differentias speciﬁcas, synonyma auctorum, loca natalia, metamorphosin. No. Bd. 1 in Species insectorum exhibentes eorum differentias speci- ﬁcas, synonyma auctorum, loca natalia, metamorphosin. 1781.

[71] FANG, Y., SATHYANARAYANAN,S., AND SEHGAL, A. Post-translational regulation of the Drosophila circadian clock requires protein phosphatase 1 (PP1). Genes Dev 21, 12 (Jun 2007), 1506–1518.

[72] FELSENSTEIN, J. Conﬁdence limits on phylogenies: an approach using the bootstrap. Evolution 39 (1985), 783–791.

[73] FOGLE,K.J.,PARSON,K.G.,DAHM,N.A., AND HOLMES, T. C. CRYPTOCHROME is a blue-light sensor that regulates neuronal ﬁring rate. Science 331, 6023 (Mar 2011), 1409–1413.

[74] FOURTNER,C.R.&KAARS, C. J. Anatomy of the central nervous system and its usefulness as a model for neurobiology. In: Cockroaches as models for neurobiology: applications in biomedical research. Volume I. Eds.: Huber, A., Masler, E. P. & Rao, B. R. CRC press Florida, USA. (1990), 65–87.

[75] FRANC¸ OIS, P., DESPIERRE,N., AND SIGGIA, E. D. Adaptive temperature compensation in circadian oscillations. PLoS Comput Biol 8, 7 (2012), e1002585.

[76] FROY,O.,GOTTER,A.L.,CASSELMAN,A.L., AND REPPERT, S. M. Illuminating the circadian clock in monarch butterﬂy migration. Science 300, 5623 (May 2003), 1303–1305.

[77] GEGEAR,R.J.,CASSELMAN,A.,WADDELL,S., AND REPPERT, S. M. Cryptochrome mediates light-dependent magnetosensitivity in Drosophila. Nature 454, 7207 (Aug 2008), 1014–1018.

[78] GEGEAR,R.J.,FOLEY,L.E.,CASSELMAN,A., AND REPPERT, S. M. Animal cryptochromes mediate magnetoreception by an unconventional photochemical mechanism. Nature 463, 7282 (Feb 2010), 804–807.

[79] GEKAKIS,N.,SAEZ,L.,DELAHAYE-BROWN,A.M.,MYERS, M. P., SEHGAL,A.,YOUNG, M. W., AND WEITZ, C. J. Isolation of timeless by PER protein interaction: defective interaction between timeless protein and long-period mutant PERL. Science 270, 5237 (Nov 1995), 811–815.

[80] GIEBULTOWICZ,J.M., AND HEGE, D. M. Circadian clock in Malpighian tubules. Nature 386, 6626 (Apr 1997), 664.

[81] GIEBULTOWICZ,J.M.,STANEWSKY,R.,HALL,J.C., AND HEGE, D. M. Transplanted Drosophila excretory tubules maintain circadian clock cycling out of phase with the host. Curr Biol 10, 2 (Jan 2000), 107–110.

[82] GLASER, F. T., AND STANEWSKY, R. Temperature synchronization of the Drosophila circadian clock. Curr Biol 15, 15 (Aug 2005), 1352–1363.

[83] GLOSSOP,N.R.,LYONS,L.C., AND HARDIN, P. E. Interlocked feedback loops within the Drosophila circadian oscillator. Science 286, 5440 (Oct 1999), 766–768. BIBLIOGRAPHY 147

[84] GLOSSOP,N.R.J.,HOUL,J.H.,ZHENG,H.,NG, F. S., DUDEK,S.M., AND HARDIN, P. E. VRILLE feeds back to control circadian transcription of clock in the Drosophila circadian oscillator. Neuron 37, 2 (Jan 2003), 249–261.

[85] GOTO, S. G. Roles of circadian clock genes in insect photoperiodism. Entomological Science 16, 1 (2013), 1–16.

[86] GOTO,S.G., AND DENLINGER, D. L. Short-day and long-day expression patterns of genes involved in the ﬂesh ﬂy clock mechanism: period, timeless, cycle and cryptochrome. J Insect Physiol 48, 8 (Aug 2002), 803–816.

[87] GOTO,S.G.,HAN,B., AND DENLINGER, D. L. A nondiapausing variant of the ﬂesh ﬂy, Sar- cophaga bullata, that shows arrhythmic adult eclosion and elevated expression of two circadian clock genes, period and timeless. J Insect Physiol 52, 11-12 (2006), 1213–1218.

[88] GOTTER,A.L.,LEVINE,J.D., AND REPPERT, S. M. Sex-linked period genes in the silkmoth, Antheraea pernyi: implications for circadian clock regulation and the evolution of sex chromo- somes. Neuron 24, 4 (Dec 1999), 953–965.

[89] GRIMA,B.,CHELOT´ ,E.,XIA,R., AND ROUYER, F. Morning and evening peaks of activity rely on different clock neurons of the Drosophila brain. Nature 431, 7010 (Oct 2004), 869–873.

[90] GUBLER,U., AND HOFFMAN, B. J. A simple and very efﬁcient method for generating cDNA libraries. Gene 25, 2-3 (Nov 1983), 263–269.

[91] HAO,H.,ALLEN,D.L., AND HARDIN, P. E. A circadian enhancer mediates PER-dependent mRNA cycling in Drosophila melanogaster. Mol Cell Biol 17, 7 (Jul 1997), 3687–3693.

[92] HARDIN, P. E. Molecular genetic analysis of circadian timekeeping in Drosophila. Adv Genet 74 (2011), 141–173.

[93] HARDIN, P. E., HALL,J.C., AND ROSBASH, M. Feedback of the Drosophila period gene product on circadian cycling of its messenger RNA levels. Nature 343, 6258 (Feb 1990), 536–540.

[94] HARKER, J. E. Diurnal rhythms in Periplaneta americana L. Nature 173, 4406 (Apr 1954), 689–690.

[95] HARKER, J. E. Factors controlling the diurnal rhythm of activity of Periplaneta americana L. J. Exp. Biol. 33 (1956), 224–234.

[96] HEFTI,M.H.,FRANC¸ OIJS,K.-J., DE VRIES,S.C.,DIXON,R., AND VERVOORT, J. The PAS fold. A redeﬁnition of the PAS domain based upon structural prediction. Eur J Biochem 271, 6 (Mar 2004), 1198–1208.

[97] HELFRICH-FORSTER¨ ,C.,SHAFER, O. T., WULBECK¨ ,C.,GRIESHABER,E.,RIEGER,D., AND TAGHERT, P. Development and morphology of the clock-gene-expressing lateral neurons of Drosophila melanogaster. J Comp Neurol 500, 1 (Jan 2007), 47–70.

[98] HELFRICH-FORSTER¨ ,C.,WINTER,C.,HOFBAUER,A.,HALL,J.C., AND STANEWSKY,R. The circadian clock of fruit ﬂies is blind after elimination of all known photoreceptors. Neuron 30, 1 (Apr 2001), 249–261. 148 BIBLIOGRAPHY

[99] HELFRICH-FORSTER¨ ,C.,YOSHII, T., WULBECK¨ ,C.,GRIESHABER,E.,RIEGER,D.,BACH- LEITNER, W., CUSAMANO, P., AND ROUYER, F. The lateral and dorsal neurons of Drosophila melanogaster: new insights about their morphology and function. Cold Spring Harb Symp Quant Biol 72 (2007), 517–525.

[100] HEMSLEY,M.J.,MAZZOTTA,G.M.,MASON,M.,DISSEL,S.,TOPPO,S.,PAGANO,M.A., SANDRELLI, F., MEGGIO, F., ROSATO,E.,COSTA,R., AND TOSATTO, S. C. E. Linear motifs in the C-terminus of D. melanogaster cryptochrome. Biochem Biophys Res Commun 355, 2 (Apr 2007), 531–537.

[101] HENNIG,S.,STRAUSS,H.M.,VANSELOW,K.,YILDIZ,O.,SCHULZE,S.,ARENS,J., KRAMER,A., AND WOLF, E. Structural and functional analyses of PAS domain interactions of the clock proteins Drosophila PERIOD and mouse PERIOD2. PLoS Biol 7, 4 (Apr 2009), e94.

[102] HIRAYAMA,J.,NAKAMURA,H.,ISHIKAWA, T., KOBAYASHI, Y., AND TODO, T. Functional and structural analyses of cryptochrome. vertebrate cry regions responsible for interaction with the clock:bmal1 heterodimer and its nuclear localization.

[103] HOANG,N.,SCHLEICHER,E.,KACPRZAK,S.,BOULY, J.-P., PICOT,M.,WU, W., BERNDT, A.,WOLF,E.,BITTL,R., AND AHMAD, M. Human and Drosophila cryptochromes are light activated by ﬂavin photoreduction in living cells. PLoS Biol 6, 7 (Jul 2008), e160.

[104] HODGE,J.J., AND STANEWSKY, R. Function of the Shaw potassium channel within the Drosophil circadian clock. PLoS One 3, 5 (2008), e2274.

[105] HODKOVA,M.,SYROVA,Z.,DOLEZEL,D., AND SAUMAN,I. Period gene expression in relation to seasonality and circadian rhythms in the linden bug, Pyrrhocoris apterus (Heteroptera). European Journal of Entomology 100, 2 (2003), 267–273.

[106] HOGENESCH,J.B.,GU, Y. Z., JAIN,S., AND BRADFIELD, C. A. The basic-helix-loop-helix- PAS orphan MOP3 forms transcriptionally active complexes with circadian and hypoxia factors. Proc Natl Acad Sci U S A 95, 10 (May 1998), 5474–5479.

[107] HOLLAND, P. M., ABRAMSON,R.D.,WATSON,R., AND GELFAND, D. H. Detection of spe- ciﬁc polymerase chain reaction product by utilizing the 5’—-3’ exonuclease activity of Thermus aquaticus DNA polymerase. Proc Natl Acad Sci U S A 88, 16 (Aug 1991), 7276–7280.

[108] HOLTON, T. A., AND GRAHAM, M. W. A simple and efﬁcient method for direct cloning of PCR products using ddT-tailed vectors. Nucleic Acids Res 19, 5 (Mar 1991), 1156.

[109] HOMBERG,U.,WURDEN¨ ,S.,DIRCKSEN,H., AND RAO, K. R. Comparative anatomy of pigment-dispersing hormone-immunoreactive neurons in the brain of orthopteroid insects. Cell and Tissue Research 266 (1991), 343–357.

[110] HOSOKAWA,N.,HATAKEYAMA, T. S., KOJIMA, T., KIKUCHI, Y., ITO,H., AND IWASAKI,H. Circadian transcriptional regulation by the posttranslational oscillator without de novo clock gene expression in Synechococcus. Proc Natl Acad Sci U S A 108, 37 (Sep 2011), 15396–15401.

[111] HOUL,J.H.,YU, W., DUDEK,S.M., AND HARDIN, P. E. Drosophila CLOCK is constitutively expressed in circadian oscillator and non-oscillator cells. J Biol Rhythms 21, 2 (Apr 2006), 93–103. BIBLIOGRAPHY 149

[112] HU,K.,REICHERT,H., AND STARK, W. Electrophysiological characterization of Drosophila ocelli. Journal of comparative physiology 126, 1 (1978), 15–24.

[113] HUANG,Z.J.,CURTIN,K.D., AND ROSBASH, M. PER protein interactions and temperature compensation of a circadian clock in Drosophila. Science 267, 5201 (Feb 1995), 1169–1172.

[114] HUANG,Z.J.,EDERY,I., AND ROSBASH, M. PAS is a dimerization domain common to Drosophila period and several transcription factors. Nature 364, 6434 (Jul 1993), 259–262.

[115] HUNTER-ENSOR,M.,OUSLEY,A., AND SEHGAL, A. Regulation of the Drosophila protein timeless suggests a mechanism for resetting the circadian clock by light. Cell 84, 5 (Mar 1996), 677–685.

[116] IKENO, T., KATAGIRI,C.,NUMATA,H., AND GOTO, S. G. Causal involvement of mammalian- type cryptochrome in the circadian cuticle deposition rhythm in the bean bug Riptortus pedestris. Insect Mol Biol 20, 3 (Jun 2011), 409–415.

[117] IKENO, T., NUMATA,H., AND GOTO, S. G. Molecular characterization of the circadian clock genes in the bean bug, Riptortus pedestris, and their expression patterns under long- and short-day conditions. Gene 419, 1-2 (Aug 2008), 56–61.

[118] IKENO, T., NUMATA,H., AND GOTO, S. G. Circadian clock genes period and cycle regulate photoperiodic diapause in the bean bug Riptortus pedestris males. J Insect Physiol 57, 7 (Jul 2011), 935–938.

[119] IKENO, T., NUMATA,H., AND GOTO, S. G. Photoperiodic response requires mammalian-type cryptochrome in the bean bug Riptortus pedestris. Biochem Biophys Res Commun 410, 3 (Jul 2011), 394–397.

[120] IKENO, T., TANAKA,S.I.,NUMATA,H., AND GOTO, S. G. Photoperiodic diapause under the control of circadian clock genes in an insect. BMC Biol 8 (2010), 116.

[121] INGRAM,K.K.,KUTOWOI,A.,WURM, Y., SHOEMAKER,D.,MEIER,R., AND BLOCH,G. The molecular clockwork of the ﬁre ant Solenopsis invicta. PLoS One 7, 11 (2012), e45715.

[122] ITO,C.,GOTO,S.G.,SHIGA,S.,TOMIOKA,K., AND NUMATA, H. Peripheral circadian clock for the cuticle deposition rhythm in Drosophila melanogaster. Proc Natl Acad Sci U S A 105, 24 (Jun 2008), 8446–8451.

[123] IVANCHENKO,M.,STANEWSKY,R., AND GIEBULTOWICZ, J. M. Circadian photoreception in Drosophila: functions of cryptochrome in peripheral and central clocks. J Biol Rhythms 16, 3 (Jun 2001), 205–215.

[124] IWAI,S.,FUKUI, Y., FUJIWARA, Y., AND TAKEDA, M. Structure and expressions of two circadian clock genes, period and timeless in the commercial silkmoth, Bombyx mori. J Insect Physiol 52, 6 (Jun 2006), 625–637.

[125] JARAMILLO,A.M.,ZENG,H.,FEI,H.,ZHOU, Y., AND LEVITAN, I. B. Expression and function of variants of slob, slowpoke channel binding protein, in Drosophila. J Neurophysiol 95, 3 (Mar 2006), 1957–1965. 150 BIBLIOGRAPHY

[126] JARAMILLO,A.M.,ZHENG,X.,ZHOU, Y., AMADO,D.A.,SHELDON,A.,SEHGAL,A., AND LEVITAN, I. B. Pattern of distribution and cycling of SLOB, Slowpoke channel binding protein, in Drosophila. BMC Neurosci 5 (Jan 2004), 3.

[127] JEPSON,J.E.C.,SHAHIDULLAH,M.,LAMAZE,A.,PETERSON,D.,PAN,H., AND KOH, K. Dyschronic, a Drosophila homolog of a deaf-blindness gene, regulates circadian output and Slowpoke channels. PLoS Genet 8, 4 (2012), e1002671.

[128] KADENER,S.,STOLERU,D.,MCDONALD,M.,NAWATHEAN, P., AND ROSBASH,M. Clock- work orange is a transcriptional repressor and a new Drosophila circadian pacemaker component. Genes Dev 21, 13 (Jul 2007), 1675–1686.

[129] KAMAE, Y., AND TOMIOKA,K. Timeless is an essential component of the circadian clock in a primitive insect, the ﬁrebrat Thermobia domestica. J Biol Rhythms 27, 2 (Apr 2012), 126–134.

[130] KANEKO,M., AND HALL, J. C. Neuroanatomy of cells expressing clock genes in drosophila: transgenic manipulation of the period and timeless genes to mark the perikarya of circadian pacemaker neurons and their projections. J Comp Neurol 422, 1 (Jun 2000), 66–94.

[131] KAPPA, S. Photoperiode-abh¨angige Plastizit¨at des circadianen Schrittmacherzentrums der Sch- abe Rhyparobia maderae - Immunzytochemiche Untersuchungen Pigment-Dispersing Faktor und Orcokinin enthaltender Neurone. Diploma Thesis, Johannes-Gutenberg-Universitat¨ Mainz (2012), 118.

[132] KEEGAN, K. P., PRADHAN,S.,WANG, J.-P., AND ALLADA, R. Meta-analysis of Drosophila circadian microarray studies identiﬁes a novel set of rhythmically expressed genes. PLoS Comput Biol 3, 11 (Nov 2007), e208.

[133] KEVAN, D. K. M. Names involving the madeira and surinam cockroaches (Dictuoptera, Blattodea, Nauphoetidae). The entomologist’s Record 92 (1980), 77–82.

[134] KIM, E. Y., BAE,K.,NG, F. S., GLOSSOP,N.R.J.,HARDIN, P. E., AND EDERY, I. Drosophila CLOCK protein is under posttranscriptional control and inﬂuences light-induced activity. Neuron 34, 1 (Mar 2002), 69–81.

[135] KIM, E. Y., AND EDERY, I. Balance between DBT/CKIepsilon kinase and protein phosphatase activities regulate phosphorylation and stability of Drosophila CLOCK protein. Proc Natl Acad Sci U S A 103, 16 (Apr 2006), 6178–6183.

[136] KIM, E. Y., KO, H. W., YU, W., HARDIN, P. E., AND EDERY, I. A DOUBLETIME kinase binding domain on the Drosophila PERIOD protein is essential for its hyperphosphorylation, transcriptional repression, and circadian clock function. Mol Cell Biol 27, 13 (Jul 2007), 5014–5028.

[137] KING,H.A.,HOELZ,A.,CRANE,B.R., AND YOUNG, M. W. Structure of an enclosed dimer formed by the Drosophila period protein. J Mol Biol 413, 3 (Oct 2011), 561–572.

[138] KIVIMAE¨ ,S.,SAEZ,L., AND YOUNG, M. W. Activating PER repressor through a DBT-directed phosphorylation switch. PLoS Biol 6, 7 (Jul 2008), e183. BIBLIOGRAPHY 151

[139] KLOSS,B.,PRICE,J.L.,SAEZ,L.,BLAU,J.,ROTHENFLUH,A.,WESLEY,C.S., AND YOUNG, M. W. The Drosophila clock gene double-time encodes a protein closely related to human casein kinase I epsilon. Cell 94, 1 (Jul 1998), 97–107.

[140] KLOSS,B.,ROTHENFLUH,A.,YOUNG, M. W., AND SAEZ, L. Phosphorylation of PERIOD is inﬂuenced by cycling physical associations of DOUBLE-TIME, PERIOD, and TIMELESS in the Drosophila clock. Neuron 30, 3 (Jun 2001), 699–706.

[141] KNOWLES,A.,KOH,K.,WU, J.-T., CHIEN, C.-T., CHAMOVITZ,D.A., AND BLAU, J. The COP9 signalosome is required for light-dependent TIMELESS degradation and Drosophila clock resetting. J Neurosci 29, 4 (Jan 2009), 1152–1162.

[142] KO, H. W., JIANG,J., AND EDERY, I. Role for Slimb in the degradation of Drosophila Period protein phosphorylated by Doubletime. Nature 420, 6916 (Dec 2002), 673–678.

[143] KO, H. W., KIM, E. Y., CHIU,J.,VANSELOW, J. T., KRAMER,A., AND EDERY, I. A hierarchical phosphorylation cascade that regulates the timing of PERIOD nuclear entry reveals novel roles for proline-directed kinases and GSK-3beta/SGG in circadian clocks. J Neurosci 30, 38 (Sep 2010), 12664–12675.

[144] KOH,K.,ZHENG,X., AND SEHGAL, A. JETLAG resets the Drosophila circadian clock by promoting light-induced degradation of TIMELESS. Science 312, 5781 (Jun 2006), 1809–1812.

[145] KONOPKA,R.J., AND BENZER, S. Clock mutants of Drosophila melanogaster. Proc Natl Acad Sci U S A 68, 9 (Sep 1971), 2112–2116.

[146] KONOPKA,R.J.,PITTENDRIGH,C., AND ORR, D. Reciprocal behaviour associated with altered homeostasis and photosensitivity of Drosophila clock mutants. J Neurogenet 6, 1 (Sep 1989), 1–10.

[147] KOSTˇ AL´ , V. Insect photoperiodic calendar and circadian clock: independence, cooperation, or unity? J Insect Physiol 57, 5 (May 2011), 538–556.

[148] KRISHNAN,B.,DRYER,S.E., AND HARDIN, P. E. Circadian rhythms in olfactory responses of Drosophila melanogaster. Nature 400, 6742 (Jul 1999), 375–378.

[149] KRISHNAN,B.,LEVINE,J.D.,LYNCH,M.K.,DOWSE,H.B.,FUNES, P., HALL,J.C., HARDIN, P. E., AND DRYER, S. E. A new role for cryptochrome in a Drosophila circadian oscillator. Nature 411, 6835 (May 2001), 313–317.

[150] KULA-EVERSOLE,E.,NAGOSHI,E.,SHANG, Y., RODRIGUEZ,J.,ALLADA,R., AND ROS- BASH, M. Surprising gene expression patterns within and between PDF-containing circadian neurons in Drosophila. Proc Natl Acad Sci U S A 107, 30 (Jul 2010), 13497–13502.

[151] KUME,K.,ZYLKA,M.J.,SRIRAM,S.,SHEARMAN, L. P., WEAVER,D.R.,JIN,X.,MAY- WOOD,E.S.,HASTINGS,M.H., AND REPPERT, S. M. mCRY1 and mCRY2 are essential components of the negative limb of the circadian clock feedback loop. Cell 98, 2 (Jul 1999), 193– 205.

[152] KUNITZ, M. Crystalline desoxyribonuclease; isolation and general properties; spectrophotometric method for the measurement of desoxyribonuclease activity. J Gen Physiol 33, 4 (Mar 1950), 349–362. 152 BIBLIOGRAPHY

[153] LANDSKRON,J.,CHEN, K. F., WOLF,E., AND STANEWSKY, R. A role for the PE- RIOD:PERIOD homodimer in the Drosophila circadian clock. PLoS Biol 7, 4 (Apr 2009), e3.

[154] LANKINEN, P., AND FORSMAN, P. Independence of genetic geographical variation between photoperiodic diapause, circadian eclosion rhythm, and Thr-Gly repeat region of the period gene in Drosophila littoralis. J Biol Rhythms 21, 1 (Feb 2006), 3–12.

[155] LEAR,B.C.,LIN,J.-M.,KEATH,J.R.,MCGILL,J.J.,RAMAN,I.M., AND ALLADA, R. The ion channel narrow abdomen is critical for neural output of the Drosophila circadian pacemaker. Neuron 48, 6 (Dec 2005), 965–976.

[156] LEE,C.,BAE,K., AND EDERY, I. [the Drosophila clock protein undergoes daily rhythms in abundance, phosphorylation, and interactions with the per-tim complex.

[157] LEE, T., AND LUO, L. Mosaic analysis with a repressible cell marker for studies of gene function in neuronal morphogenesis. Neuron 22, 3 (Mar 1999), 451–461.

[158] LEES, A. Photoperiodic time measurement in the aphid Megoura viciae. Journal of Insect Physi- ology 19, 12 (1973), 2279 – 2316.

[159] LEES, A. Some effects of temperature on the hour glass photoperiod timer in the aphid Megoura viciae. Journal of Insect Physiology 32, 1 (1986), 79 – 89.

[160] LETUNIC,I.,DOERKS, T., AND BORK, P. SMART 7: recent updates to the protein domain annotation resource. Nucleic Acids Res 40, Database issue (Jan 2012), D302–D305.

[161] LEVINE,J.D.,FUNES, P., DOWSE,H.B., AND HALL, J. C. Advanced analysis of a cryptochrome mutation’s effects on the robustness and phase of molecular cycles in isolated peripheral tissues of Drosophila. BMC Neurosci 3 (Apr 2002), 5.

[162] LILLIEFORS, H. W. On the Kolmogorov-Smirnov test for the exponential distribution with mean unknown. Journal of the American Statistical Association 64 (1969), 387–389.

[163] LIM,C.,CHUNG, B. Y., PITMAN,J.L.,MCGILL,J.J.,PRADHAN,S.,LEE,J.,KEEGAN, K. P., CHOE,J., AND ALLADA,R. Clockwork orange encodes a transcriptional repressor important for circadian-clock amplitude in Drosophila. Curr Biol 17, 12 (Jun 2007), 1082–1089.

[164] LIM,C.,LEE,J.,CHOI,C.,KILMAN, V. L., KIM,J.,PARK,S.M.,JANG,S.K.,ALLADA,R., AND CHOE, J. The novel gene twenty-four deﬁnes a critical translational step in the Drosophila clock. Nature 470, 7334 (Feb 2011), 399–403.

[165] LIN,C., AND TODO, T. The cryptochromes. Genome Biol 6, 5 (2005), 220.

[166] LIN,G.G.-H.,LIOU, R.-F., AND LEE, H.-J. The period gene of the German cockroach and its novel linking power between vertebrate and invertebrate. Chronobiol Int 19, 6 (Nov 2002), 1023–1040.

[167] LIN,J.-M.,KILMAN, V. L., KEEGAN,K.,PADDOCK,B.,EMERY-LE,M.,ROSBASH,M., AND ALLADA, R. A role for casein kinase 2 alpha in the Drosophila circadian clock. Nature 420, 6917 (2002), 816–820. BIBLIOGRAPHY 153

[168] LIN, Y., HAN,M.,SHIMADA,B.,WANG,L.,GIBLER, T. M., AMARAKONE,A.,AWAD, T. A., STORMO,G.D.,VAN GELDER,R.N., AND TAGHERT, P. H. Inﬂuence of the period-dependent circadian clock on diurnal, circadian, and aperiodic gene expression in Drosophila melanogaster. Proc Natl Acad Sci U S A 99, 14 (Jul 2002), 9562–9567.

[169] LOESEL,R., AND HOMBERG, U. Histamine-immunoreactive neurons in the brain of the cockroach Leucophaea maderae. Brain Res 842, 2 (Sep 1999), 408–418.

[170] LUCAS-LLEDO´ ,J.I., AND LYNCH, M. Evolution of mutation rates: phylogenomic analysis of the photolyase/cryptochrome family. Mol Biol Evol 26, 5 (May 2009), 1143–1153.

[171] LUKAT,R., AND WEBER, F. The structure of locomotor activity in bilobectomized cockroaches (Blaberus fuscus). Experientia 35, 1 (1979), 38–39.

[172] LUKAT,R.,WEBER, F., AND WIEDENMANN, G. Cyclic layer deposition in the cockroach (Blaberus craniifer) endocuticle: A decentral circadian “clock”? . Journal of Insect Physiology 35, 4 (1989), 321 – 329.

[173] MAJERCAK,J.,SIDOTE,D.,HARDIN, P. E., AND EDERY, I. How a circadian clock adapts to seasonal decreases in temperature and day length. Neuron 24, 1 (Sep 1999), 219–230.

[174] MARCHLER-BAUER,A.,LU,S.,ANDERSON,J.B.,CHITSAZ, F., DERBYSHIRE,M.K., DEWEESE-SCOTT,C.,FONG,J.H.,GEER, L. Y., GEER,R.C.,GONZALES,N.R.,GWADZ, M.,HURWITZ,D.I.,JACKSON,J.D.,KE,Z.,LANCZYCKI,C.J.,LU, F., MARCHLER,G.H., MULLOKANDOV,M.,OMELCHENKO, M. V., ROBERTSON,C.L.,SONG,J.S.,THANKI,N., YAMASHITA,R.A.,ZHANG,D.,ZHANG,N.,ZHENG,C., AND BRYANT, S. H. CDD:aCon- served Domain Database for the functional annotation of proteins. Nucleic Acids Res 39, Database issue (Jan 2011), D225–D229.

[175] MARCHUK,D.,DRUMM,M.,SAULINO,A., AND COLLINS, F. S. Construction of T-vectors, a rapid and general system for direct cloning of unmodiﬁed PCR products. Nucleic Acids Res 19, 5 (Mar 1991), 1154.

[176] MARFORI,M.,LONHIENNE, T. G., FORWOOD,J.K., AND KOBE, B. Structural basis of high- afﬁnity nuclear localization signal interactions with importin-alpha. Trafﬁc 13, 4 (Apr 2012), 532– 548.

[177] MARKOVA, E. P., UEDA,H.,SAKAMOTO,K.,OISHI,K.,SHIMADA, T., AND TAKEDA,M. Cloning of Cyc (Bmal1) homolog in Bombyx mori: structural analysis and tissue speciﬁc distribu- tions. Comp Biochem Physiol B Biochem Mol Biol 134, 3 (Mar 2003), 535–542.

[178] MARTINEK,S.,INONOG,S.,MANOUKIAN,A.S., AND YOUNG, M. W. A role for the segment polarity gene shaggy/GSK-3 in the Drosophila circadian clock. Cell 105, 6 (Jun 2001), 769–779.

[179] MATHIAS,D.,REED,L.K.,BRADSHAW, W. E., AND HOLZAPFEL, C. M. Evolutionary divergence of circadian and photoperiodic phenotypes in the pitcher-plant mosquito, Wyeomyia smithii. J Biol Rhythms 21, 2 (Apr 2006), 132–139. 154 BIBLIOGRAPHY

[180] MATSUMOTO,A.,UKAI-TADENUMA,M.,YAMADA,R.G.,HOUL,J.,UNO,K.D.,KA- SUKAWA, T., DAUWALDER,B.,ITOH, T. Q., TAKAHASHI,K.,UEDA,R.,HARDIN, P. E., TANIMURA, T., AND UEDA, H. R. A functional genomics strategy reveals clockwork orange as a transcriptional regulator in the Drosophila circadian clock. Genes Dev 21, 13 (Jul 2007), 1687– 1700.

[181] MAZZOTTA,G.,ROSSI,A.,LEONARDI,E.,MASON,M.,BERTOLUCCI,C.,CACCIN,L.,SPO- LAORE,B.,MARTIN,A.J.M.,SCHLICHTING,M.,GREBLER,R.,HELFRICH-FORSTER¨ ,C., MAMMI,S.,COSTA,R., AND TOSATTO, S. C. E. Fly cryptochrome and the visual system. Proc Natl Acad Sci U S A 110, 15 (Apr 2013), 6163–6168.

[182] MAZZOTTA,G.M.,SANDRELLI, F., ZORDAN,M.A.,MASON,M.,BENNA,C.,CISOTTO, P., ROSATO,E.,KYRIACOU, C. P., AND COSTA, R. The clock gene period in the medﬂy Ceratitis capitata. Genet Res 86, 1 (Aug 2005), 13–30.

[183] MCDONALD,M.J., AND ROSBASH, M. Microarray analysis and organization of circadian gene expression in Drosophila. Cell 107, 5 (Nov 2001), 567–578.

[184] MEHNERT,K.I.,BERAMENDI,A.,ELGHAZALI, F., NEGRO, P., KYRIACOU, C. P., AND CAN- TERA, R. Circadian changes in Drosophila motor terminals. Dev Neurobiol 67, 4 (Mar 2007), 415–421.

[185] MEHNERT,K.I., AND CANTERA, R. A peripheral pacemaker drives the circadian rhythm of synaptic boutons in Drosophila independently of synaptic activity. Cell Tissue Res 334, 1 (Oct 2008), 103–109.

[186] MEISSNER,R.-A.,KILMAN, V. L., LIN,J.-M., AND ALLADA, R. TIMELESS is an important mediator of CK2 effects on circadian clock function in vivo. J Neurosci 28, 39 (Sep 2008), 9732– 9740.

[187] MERLIN,C.,FRANC¸ OIS,M.-C.,QUEGUINER,I.,MA¨IBECHE` -COISNE´,M., AND JACQUIN- JOLY, E. Evidence for a putative antennal clock in Mamestra brassicae: molecular cloning and characterization of two clock genes – period and cryptochrome – in antennae. Insect Mol Biol 15, 2 (Apr 2006), 137–145.

[188] MERLIN,C.,LUCAS, P., ROCHAT,D.,FRANC¸ OIS,M.-C.,MA¨IBECHE` -COISNE,M., AND JACQUIN-JOLY, E. An antennal circadian clock and circadian rhythms in peripheral pheromone reception in the moth Spodoptera littoralis. J Biol Rhythms 22, 6 (Dec 2007), 502–514.

[189] MEYER, P., SAEZ,L., AND YOUNG, M. W. PER-TIM interactions in living Drosophila cells: an interval timer for the circadian clock. Science 311, 5758 (Jan 2006), 226–229.

[190] MORIYAMA, Y., KAMAE, Y., URYU,O., AND TOMIOKA,K. gb’clock is expressed in the optic lobe and is required for the circadian clock in the cricket Gryllus bimaculatus. J Biol Rhythms 27, 6 (Dec 2012), 467–477.

[191] MORIYAMA, Y., SAKAMOTO, T., KARPOVA,S.G.,MATSUMOTO,A.,NOJI,S., AND TOMIOKA, K. RNA interference of the clock gene period disrupts circadian rhythms in the cricket Gryllus bimaculatus. J Biol Rhythms 23, 4 (Aug 2008), 308–318. BIBLIOGRAPHY 155

[192] MORIYAMA, Y., SAKAMOTO, T., MATSUMOTO,A.,NOJI,S., AND TOMIOKA, K. Functional analysis of the circadian clock gene period by RNA interference in nymphal crickets Gryllus bimaculatus. J Insect Physiol 55, 2 (Feb 2009), 183–187.

[193] MOTE,M.I., AND BLACK, K. R. Action spectrum and threshold sensitivity of entrainment of circadian running activity In the cockroach Periplaneta americana. Photochemistry and Photobiology 34, 2 (1981), 257–265.

[194] MOTE,M.I., AND GOLDSMITH, T. H. Spectral sensitivities of color receptors in the compound eye of the cockroach Periplaneta. J Exp Zool 173, 2 (Feb 1970), 137–145.

[195] MUGURUMA, F., GOTO,S.G.,NUMATA,H., AND SHIGA, S. Effect of photoperiod on clock gene expression and subcellular distribution of PERIOD in the circadian clock neurons of the blow ﬂy Protophormia terraenovae. Cell Tissue Res 340, 3 (Jun 2010), 497–507.

[196] MURAD,A.,EMERY-LE,M., AND EMERY, P. A subset of dorsal neurons modulates circadian behavior and light responses in Drosophila. Neuron 53, 5 (Mar 2007), 689–701.

[197] MURRE,C.,MCCAW, P. S., VAESSIN,H.,CAUDY,M.,JAN, L. Y., JAN, Y. N., CABRERA, C. V., BUSKIN,J.N.,HAUSCHKA,S.D., AND LASSAR, A. B. Interactions between heterologous helix-loop-helix proteins generate complexes that bind speciﬁcally to a common DNA sequence. Cell 58, 3 (Aug 1989), 537–544.

[198] MYERS, M. P., WAGER-SMITH,K.,ROTHENFLUH-HILFIKER,A., AND YOUNG, M. W. Light- induced degradation of TIMELESS and entrainment of the Drosophila circadian clock. Science 271, 5256 (Mar 1996), 1736–1740.

[199] MYERS, M. P., WAGER-SMITH,K.,WESLEY,C.S.,YOUNG, M. W., AND SEHGAL, A. Po- sitional cloning and sequence analysis of the Drosophila clock gene, timeless. Science 270, 5237 (Nov 1995), 805–808.

[200] NAIDOO,N.,SONG, W., HUNTER-ENSOR,M., AND SEHGAL, A. A role for the proteasome in the light response of the timeless clock protein. Science 285, 5434 (Sep 1999), 1737–1741.

[201] NAKAJIMA,M.,IMAI,K.,ITO,H.,NISHIWAKI, T., MURAYAMA, Y., IWASAKI,H.,OYAMA, T., AND KONDO, T. Reconstitution of circadian oscillation of cyanobacterial KaiC phosphorylation in vitro. Science 308, 5720 (Apr 2005), 414–415.

[202] NAMBU,J.R.,LEWIS,J.O.,WHARTON,K.A., AND CREWS, S. T. The Drosophila single- minded gene encodes a helix-loop-helix protein that acts as a master regulator of CNS midline development. Cell 67, 6 (Dec 1991), 1157–1167.

[203] NANDA,K.K., AND HAMNER, K. C. Studies on the nature of the endogenous rhythm affecting photoperiodic response of biloxi soybean. Botanical Gazette 120, 1 (1958), pp. 14–25.

[204] NAWATHEAN, P., AND ROSBASH, M. The doubletime and CKII kinases collaborate to potentiate Drosophila PER transcriptional repressor activity. Mol Cell 13, 2 (Jan 2004), 213–223.

[205] NAWATHEAN, P., STOLERU,D., AND ROSBASH, M. A small conserved domain of Drosophila PERIOD is important for circadian phosphorylation, nuclear localization, and transcriptional repressor activity. Mol Cell Biol 27, 13 (Jul 2007), 5002–5013. 156 BIBLIOGRAPHY

[206] NELSON, W., TONG, Y. L., LEE,J.K., AND HALBERG, F. Methods for cosinor-rhythmometry. Chronobiologia 6, 4 (1979), 305–323.

[207] NGUYEN BA,A.N.,POGOUTSE,A.,PROVART,N., AND MOSES, A. M. NLStradamus: a simple Hidden Markov Model for nuclear localization signal prediction. BMC Bioinformatics 10 (2009), 202.

[208] NISHIITSUTSUJI-UWO,J.&PITTENDRIGH, C. S. Central nervous system control of circadian rhythmicity in the cockroach II. The pathway of light signals that entrain the rhythm. Zeitschrift fur¨ vergleichende Physiologie 58 (1968), 1–13.

[209] NISHIITSUTSUJI-UWO,J.&PITTENDRIGH, C. S. Central nervous system control of circadian rhythmicity in the cockroach III . The optic lobes, Locus of the driving oscillation? Zeitschrift fur¨ vergleichende Physiologie 58 (1968), 14–46.

[210] NITABACH,M.N., AND TAGHERT, P. H. Organization of the Drosophila circadian control circuit. Curr Biol 18, 2 (Jan 2008), R84–R93.

[211] NASSEL¨ ,D.R.,PERSSON,M.G., AND MUREN, J. E. Baratin, a nonamidated neurostimulating neuropeptide, isolated from cockroach brain: distribution and actions in the cockroach and locust nervous systems. J Comp Neurol 422, 2 (Jun 2000), 267–286.

[212] OHATA,K.,NISHIYAMA,H., AND TSUKAHARA, Y. Action spectrum of the circadian clock photoreceptor in Drosophila melanogaster. In: Touitou, Y. (Eds), Biological Clocks: Mechanisms and Applications. Elsevier, Amsterdam. (1998), 167–171.

[213] O’NEILL,J.S., AND REDDY, A. B. Circadian clocks in human red blood cells. Nature 469, 7331 (Jan 2011), 498–503.

[214] O’NEILL,J.S., VAN OOIJEN,G.,DIXON,L.E.,TROEIN,C.,CORELLOU, F., BOUGET, F.-Y., REDDY,A.B., AND MILLAR, A. J. Circadian rhythms persist without transcription in a eukaryote. Nature 469, 7331 (Jan 2011), 554–558.

[215] OZBER,N.,BARIS,I.,TATLICI,G.,GUR,I.,KILINC,S.,UNAL,E.B., AND KAVAKLI,I.H. Identiﬁcation of two amino acids in the C-terminal domain of mouse CRY2 essential for PER2 interaction. BMC Mol Biol 11 (2010), 69.

[216] PAGE, T. L. Interactions between bilaterally paired components of the cockroach circadian system. Journal of comparative physiology 124, 3 (1978), 225–236.

[217] PAGE, T. L. Effects of localized low-temperature pulses on the cockroach circadian pacemaker. Am J Physiol 240, 3 (Mar 1981), R144–R150.

[218] PAGE, T. L. Transplantation of the cockroach circadian pacemaker. Science 216, 4541 (Apr 1982), 73–75.

[219] PAGE, T. L. Regeneration of the optic tracts and circadian pacemaker activity in the cockroach Leucophaea maderae. Journal of comparative physiology 152, 2 (1983), 231–240.

[220] PAGE, T. L. Developmental manipulation of the circadian pacemaker in the cockroach: relationship between pacemaker period and response to light. Physiological Entomology 16 (1991), 243–248. BIBLIOGRAPHY 157

[221] PAGE, T. L., AND BARRETT, R. K. Effects of light on circadian pacemaker development. II. Responses to light. J Comp Physiol A 165, 1 (Apr 1989), 51–59.

[222] PAGE, T. L., AND BLOCK, G. D. Circadian rhythmicity in cockroaches: effects of early post- embryonic development and ageing. Physiological Entomology 5, 3 (1980), 271–281.

[223] PAGE, T. L., CALDAROLA, P. C., AND PITTENDRIGH, C. S. Mutual entrainment of bilaterally distributed circadian pacemaker. Proc Natl Acad Sci U S A 74, 3 (Mar 1977), 1277–1281.

[224] PAGE, T. L., AND KOELLING, E. Circadian rhythm in olfactory response in the antennae controlled by the optic lobe in the cockroach. J Insect Physiol 49, 7 (Jul 2003), 697–707.

[225] PAGE, T. L., MANS,C., AND GRIFFETH, G. History dependence of circadian pacemaker period in the cockroach. J Insect Physiol 47, 9 (Sep 2001), 1085–1093.

[226] PARTCH,C.L.,GREEN,C.B., AND TAKAHASHI, J. S. Molecular architecture of the mammalian circadian clock. Trends Cell Biol (Jul 2013).

[227] PEARN, M. T., RANDALL,L.L.,SHORTRIDGE,R.D.,BURG,M.G., AND PAK, W. L. Molec- ular, biochemical, and electrophysiological characterization of Drosophila norpA mutants. J Biol Chem 271, 9 (Mar 1996), 4937–4945.

[228] PEIXOTO,A.A.,CAMPESAN,S.,COSTA,R., AND KYRIACOU, C. P. Molecular evolution of a repetitive region within the per gene of Drosophila. Mol Biol Evol 10, 1 (Jan 1993), 127–139.

[229] PENDERGAST,J.S.,NISWENDER,K.D., AND YAMAZAKI, S. Tissue-speciﬁc function of Pe- riod3 in circadian rhythmicity. PLoS One 7, 1 (2012), e30254.

[230] PESCHEL,N.,CHEN, K. F., SZABO,G., AND STANEWSKY, R. Light-dependent interactions between the Drosophila circadian clock factors cryptochrome, jetlag, and timeless. Curr Biol 19, 3 (Feb 2009), 241–247.

[231] PESCHEL,N.,VELERI,S., AND STANEWSKY, R. Veela deﬁnes a molecular link between Cryp- tochrome and Timeless in the light-input pathway to Drosophila’s circadian clock. Proc Natl Acad Sci U S A 103, 46 (Nov 2006), 17313–17318.

[232] PETRI,B.,HOMBERG,U.,LOESEL,R., AND STENGL, M. Evidence for a role of GABA and Mas-allatotropin in photic entrainment of the circadian clock of the cockroach Leucophaea maderae. J Exp Biol 205, Pt 10 (May 2002), 1459–1469.

[233] PETRI,B., AND STENGL, M. Pigment-dispersing hormone shifts the phase of the circadian pacemaker of the cockroach Leucophaea maderae. J Neurosci 17, 11 (Jun 1997), 4087–4093.

[234] PETRI,B.,STENGL,M.,WURDEN¨ ,S., AND HOMBERG, U. Immunocytochemical characterization of the accessory medulla in the cockroach Leucophaea maderae. Cell Tissue Res 282, 1 (Oct 1995), 3–19.

[235] PFAFFL, M. W. A new mathematical model for relative quantiﬁcation in real-time RT-PCR. Nucleic Acids Res 29, 9 (May 2001), e45. 158 BIBLIOGRAPHY

[236] PICOT,M.,CUSUMANO, P., KLARSFELD,A.,UEDA,R., AND ROUYER, F. Light activates output from evening neurons and inhibits output from morning neurons in the Drosophila circadian clock. PLoS Biol 5, 11 (Nov 2007), e315.

[237] PITMAN,J.L.,MCGILL,J.J.,KEEGAN, K. P., AND ALLADA, R. A dynamic role for the mushroom bodies in promoting sleep in Drosophila. Nature 441, 7094 (Jun 2006), 753–756.

[238] PITTENDRIGH, C. S. Circadian rhythms and the circadian organization of living systems. Cold Spring Harb Symp Quant Biol 25 (1960), 159–184.

[239] PITTENDRIGH, C. S. The circadian oscillation in Drosophila pseudoobscura pupae: a model for the photoperiodic clock. Zeitschrift fur¨ Pﬂanzenphysiologie 54 (1966), 275–307.

[240] PITTENDRIGH, C. S. Circadian surfaces and the diversity of possible roles of circadian organization in photoperiodic induction. Proc Natl Acad Sci U S A 69, 9 (Sep 1972), 2734–2737.

[241] PITTENDRIGH, C. S. Temporal organization: reﬂections of a Darwinian clock-watcher. Annu Rev Physiol 55 (1993), 16–54.

[242] PITTENDRIGH,C.S., AND DAAN, S. A functional analysis of circadian pacemakers in nocturnal rodents. Journal of comparative physiology 106, 3 (1976), 333–355.

[243] PLAUTZ,J.D.,STRAUME,M.,STANEWSKY,R.,JAMISON, C. F., BRANDES,C.,DOWSE, H.B.,HALL,J.C., AND KAY, S. A. Quantitative analysis of Drosophila period gene transcription in living animals. J Biol Rhythms 12, 3 (Jun 1997), 204–217.

[244] PREIBISCH,S.,SAALFELD,S., AND TOMANCAK, P. Globally optimal stitching of tiled 3D microscopic image acquisitions. Bioinformatics 25, 11 (Jun 2009), 1463–1465.

[245] PRICE,J.L.,BLAU,J.,ROTHENFLUH,A.,ABODEELY,M.,KLOSS,B., AND YOUNG, M. W. Double-time is a novel Drosophila clock gene that regulates PERIOD protein accumulation. Cell 94, 1 (Jul 1998), 83–95.

[246] QIU,J., AND HARDIN, P. E. Per mRNA cycling is locked to lights-off under photoperiodic conditions that support circadian feedback loop function. Mol Cell Biol 16, 8 (Aug 1996), 4182– 4188.

[247] REDDY, P., JACQUIER,A.C.,ABOVICH,N.,PETERSEN,G., AND ROSBASH, M. The period clock locus of D. melanogaster codes for a proteoglycan. Cell 46, 1 (Jul 1986), 53–61.

[248] REHN, J. A. G. African and Malagasy Blattidae, part III. Proc. Acad. Nat. Sci. Philadelphia 89 (1937), 17–123.

[249] REISCHIG, T., PETRI,B., AND STENGL, M. Pigment-dispersing hormone (PDH)- immunoreactive neurons form a direct coupling pathway between the bilaterally symmetric circadian pacemakers of the cockroach Leucophaea maderae. Cell Tissue Res 318, 3 (Dec 2004), 553–564.

[250] REISCHIG, T., AND STENGL, M. Morphology and pigment-dispersing hormone immunocytochemistry of the accessory medulla, the presumptive circadian pacemaker of the cockroach Leu- cophaea maderae: a light- and electron-microscopic study. Cell and Tissue Research 285 (1996), 305–319. BIBLIOGRAPHY 159

[251] REISCHIG, T., AND STENGL, M. Optic lobe commissures in a three-dimensional brain model of the cockroach Leucophaea maderae: a search for the circadian coupling pathways. J Comp Neurol 443, 4 (Feb 2002), 388–400.

[252] REISCHIG, T., AND STENGL, M. Ectopic transplantation of the accessory medulla restores circadian locomotor rhythms in arrhythmic cockroaches (Leucophaea maderae). J Exp Biol 206, Pt 11 (Jun 2003), 1877–1886.

[253] REISCHIG, T., AND STENGL, M. Ultrastructure of pigment-dispersing hormone-immunoreactive neurons in a three-dimensional model of the accessory medulla of the cockroach Leucophaea maderae. Cell Tissue Res 314, 3 (Dec 2003), 421–435.

[254] REPPERT,S.M., AND SAUMAN,I. Period and timeless tango: a dance of two clock genes. Neuron 15, 5 (Nov 1995), 983–986.

[255] REPPERT,S.M.,TSAI, T., ROCA,A.L., AND SAUMAN, I. Cloning of a structural and functional homolog of the circadian clock gene period from the giant silkmoth Antheraea pernyi. Neuron 13, 5 (Nov 1994), 1167–1176.

[256] RICHIER,B.,MICHARD-VANHEE´ ,C.,LAMOUROUX,A.,PAPIN,C., AND ROUYER, F. The clockwork orange Drosophila protein functions as both an activator and a repressor of clock gene expression. J Biol Rhythms 23, 2 (Apr 2008), 103–116.

[257] RIEGER,D.,SHAFER, O. T., TOMIOKA,K., AND HELFRICH-FORSTER¨ , C. Functional analysis of circadian pacemaker neurons in Drosophila melanogaster. J Neurosci 26, 9 (Mar 2006), 2531– 2543.

[258] RIEGER,D.,WULBECK¨ ,C.,ROUYER, F., AND HELFRICH-FORSTER¨ ,C. Period gene expression in four neurons is sufﬁcient for rhythmic activity of Drosophila melanogaster under dim light conditions. J Biol Rhythms 24, 4 (Aug 2009), 271–282.

[259] RITZ, T., ADEM,S., AND SCHULTEN, K. A model for photoreceptor-based magnetoreception in birds. Biophys J 78, 2 (Feb 2000), 707–718.

[260] ROBERTS, S. Circadian rhythms in cockroaches. Journal of comparative physiology 88, 1 (1974), 21–30.

[261] ROBERTS, S. K. Photoreception and Entrainment of Cockroach Activity Rhythms. Science 148, 3672 (May 1965), 958–959.

[262] RODRIGUEZ,J.,TANG,C.-H.A.,KHODOR, Y. L., VODALA,S.,MENET,J.S., AND ROSBASH, M. Nascent-Seq analysis of Drosophila cycling gene expression. Proc Natl Acad Sci U S A 110, 4 (Jan 2013), E275–E284.

[263] RODRIGUEZ-ZAS,S.L.,SOUTHEY,B.R.,SHEMESH, Y., RUBIN,E.B.,COHEN,M.,ROBIN- SON,G.E., AND BLOCH, G. Microarray analysis of natural socially regulated plasticity in circadian rhythms of honey bees. J Biol Rhythms 27, 1 (Feb 2012), 12–24.

[264] ROGERS,A.S.,ROSATO,E.,COSTA,R., AND KYRIACOU, C. P. Molecular analysis of circadian clocks in Drosophila simulans. Genetica 120, 1-3 (Mar 2004), 213–222. 160 BIBLIOGRAPHY

[265] ROSATO,E.,CODD, V., MAZZOTTA,G.,PICCIN,A.,ZORDAN,M.,COSTA,R., AND KYR- IACOU, C. P. Light-dependent interaction between Drosophila CRY and the clock protein PER mediated by the carboxy terminus of CRY. Curr Biol 11, 12 (Jun 2001), 909–917.

[266] ROSEN´ , W. Q., HAN,G.-B., AND LOFSTEDT¨ , C. The circadian rhythm of the sex-pheromone- mediated behavioral response in the turnip moth, Agrotis segetum, is not controlled at the peripheral level. J Biol Rhythms 18, 5 (Oct 2003), 402–408.

[267] ROTHENFLUH,A.,ABODEELY,M.,PRICE,J.L., AND YOUNG, M. W. Isolation and analysis of six timeless alleles that cause short- or long-period circadian rhythms in Drosophila. Genetics 156, 2 (Oct 2000), 665–675.

[268] ROTHENFLUH,A.,YOUNG, M. W., AND SAEZ, L. A TIMELESS-independent function for PERIOD proteins in the Drosophila clock. Neuron 26, 2 (May 2000), 505–514.

[269] RUBIN,E.B.,SHEMESH, Y., COHEN,M.,ELGAVISH,S.,ROBERTSON,H.M., AND BLOCH, G. Molecular and phylogenetic analyses reveal mammalian-like clockwork in the honey bee (Apis mellifera) and shed new light on the molecular evolution of the circadian clock. Genome Res 16, 11 (Nov 2006), 1352–1365.

[270] RUTILA,J.E.,SURI, V., LE,M.,SO, W. V., ROSBASH,M., AND HALL, J. C. CYCLEisa second bHLH-PAS clock protein essential for circadian rhythmicity and transcription of Drosophila period and timeless. Cell 93, 5 (May 1998), 805–814.

[271] SAEZ,L.,DERASMO,M.,MEYER, P., STIEGLITZ,J., AND YOUNG, M. W. A key temporal delay in the circadian cycle of Drosophila is mediated by a nuclear localization signal in the timeless protein. Genetics 188, 3 (Jul 2011), 591–600.

[272] SAEZ,L., AND YOUNG, M. W. Regulation of nuclear entry of the Drosophila clock proteins period and timeless. Neuron 17, 5 (Nov 1996), 911–920.

[273] SAIFULLAH,A.S.M., AND PAGE, T. L. Circadian regulation of olfactory receptor neurons in the cockroach antenna. J Biol Rhythms 24, 2 (Apr 2009), 144–152.

[274] SAITOU,N., AND NEI, M. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 4, 4 (Jul 1987), 406–425.

[275] SAMBROOK,J., AND RUSSELL,D. Molecular cloning: a laboratory manual. No. Bd. 1 in Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, 2001.

[276] SANDRELLI, F., COSTA,R.,KYRIACOU, C. P., AND ROSATO, E. Comparative analysis of circadian clock genes in insects. Insect Mol Biol 17, 5 (Sep 2008), 447–463.

[277] SANDRELLI, F., TAUBER,E.,PEGORARO,M.,MAZZOTTA,G.,CISOTTO, P., LANDSKRON,J., STANEWSKY,R.,PICCIN,A.,ROSATO,E.,ZORDAN,M.,COSTA,R., AND KYRIACOU, C. P. A molecular basis for natural selection at the timeless locus in Drosophila melanogaster. Science 316, 5833 (Jun 2007), 1898–1900.

[278] SATHYANARAYANAN,S.,ZHENG,X.,XIAO,R., AND SEHGAL, A. Posttranslational regulation of Drosophila PERIOD protein by protein phosphatase 2A. Cell 116, 4 (Feb 2004), 603–615. BIBLIOGRAPHY 161

[279] SAUMAN,I.,TSAI, T., ROCA,A.L., AND REPPERT,S.M. Period protein is necessary for circadian control of egg hatching behavior in the silkmoth Antheraea pernyi. Neuron 17, 5 (Nov 1996), 901–909.

[280] SAUNDERS, D. S. The photoperiodic clock in the ﬂesh-ﬂy, Sarcophaga argyrostoma. J Insect Physiol 19, 10 (Oct 1973), 1941–1954.

[281] SAUNDERS, D. S. Thermoperiodic control of diapause in an insect: theory of internal coincidence. Science 181, 4097 (Jul 1973), 358–360.

[282] SAUNDERS, D. S. Evidence for ‘dawn’ and ‘dusk’ oscillators in the Nasonia photoperiodic clock. Journal of Insect Physiology 20, 1 (1974), 77 – 88.

[283] SAUNDERS, D. S. ‘Skeleton’ photoperiods and the control of diapause and development in the ﬂesh-ﬂy, Sarcophaga argyrostoma. Journal of comparative physiology 97, 2 (1975), 97–112.

[284] SAUNDERS, D. S. External coincidence and the photoinducible phase in the Sarcophaga photoperiodic clock. Journal of comparative physiology 132, 2 (1979), 179–189.

[285] SAUNDERS, D. S. Photoperiodic time measurement in Sarcophaga argyrostoma: an attempt to use daily temperature cycles to distinguish external from internal coincidence. Journal of Comparative Physiology A 154, 6 (1984), 789–794.

[286] SAUNDERS, D. S. Controversial aspects of photoperiodism in insects and mites. J Insect Physiol 56, 11 (Nov 2010), 1491–1502.

[287] SAUNDERS, D. S. Unity and diversity in the insect photoperiodic mechanism. Entomological Science 14, 3 (2011), 235–244.

[288] SAUNDERS, D. S. Insect photoperiodism: seeing the light. Physiological Entomology 37, 3 (2012), 207–218.

[289] SAUNDERS,D.S.,GILLANDERS,S., AND LEWIS, R. Light-pulse phase response curves for the locomotor activity rhythm in Period mutants of Drosophila melanogaster. Journal of Insect Physiology 40, 11 (1994), 957 – 968.

[290] SAUNDERS,D.S.,HENRICH, V. C., AND GILBERT, L. I. Induction of diapause in Drosophila melanogaster: photoperiodic regulation and the impact of arrhythmic clock mutations on time measurement. Proceedings of the National Academy of Sciences 86, 10 (1989), 3748–3752.

[291] SAUNDERS,D.S.,STEEL,C.,VAFOPOULOU,X., AND LEWIS, R. Insect Clocks. In Insect Clocks. Elsevier, Amsterdam, 2002, p. 1560.

[292] SAWYER,L.A.,HENNESSY,J.M.,PEIXOTO,A.A.,ROSATO,E.,PARKINSON,H.,COSTA, R., AND KYRIACOU, C. P. Natural variation in a Drosophila clock gene and temperature compensation. Science 278, 5346 (Dec 1997), 2117–2120.

[293] SAWYER,L.A.,SANDRELLI, F., PASETTO,C.,PEIXOTO,A.A.,ROSATO,E.,COSTA,R., AND KYRIACOU, C. P. The period gene Thr-Gly polymorphism in Australian and African Drosophila melanogaster populations: implications for selection. Genetics 174, 1 (Sep 2006), 465–480. 162 BIBLIOGRAPHY

[294] SCHIESARI,L.,KYRIACOU, C. P., AND COSTA, R. The hormonal and circadian basis for insect photoperiodic timing. FEBS Lett 585, 10 (May 2011), 1450–1460.

[295] SCHMID,B.,HELFRICH-FORSTER¨ ,C., AND YOSHII, T. A new ImageJ plug-in ”ActogramJ” for chronobiological analyses. J Biol Rhythms 26, 5 (Oct 2011), 464–467.

[296] SCHNEIDER,N.-L., AND STENGL, M. Gap junctions between accessory medulla neurons appear to synchronize circadian clock cells of the cockroach Leucophaea maderae. J Neurophysiol 95, 3 (Mar 2006), 1996–2002.

[297] SCHUCKEL,J.,SIWICKI,K.K., AND STENGL, M. Putative circadian pacemaker cells in the antenna of the hawkmoth Manduca sexta. Cell Tissue Res 330, 2 (Nov 2007), 271–278.

[298] SCHULTZ,J.,MILPETZ, F., BORK, P., AND PONTING, C. P. SMART, a simple modular architecture research tool: identiﬁcation of signaling domains. Proc Natl Acad Sci U S A 95, 11 (May 1998), 5857–5864.

[299] SCHULZE,J.,NEUPERT,S.,SCHMIDT,L.,PREDEL,R.,LAMKEMEYER, T., HOMBERG,U., AND STENGL, M. Myoinhibitory peptides in the brain of the cockroach Leucophaea maderae and colocalization with pigment-dispersing factor in circadian pacemaker cells. J Comp Neurol 520, 5 (Apr 2012), 1078–1097.

[300] SEHADOVA´ ,H.,MARKOVA, E. P., SEHNAL, F., AND TAKEDA, M. Distribution of circadian clock-related proteins in the cephalic nervous system of the silkworm, Bombyx mori. J Biol Rhythms 19, 6 (Dec 2004), 466–482.

[301] SEHGAL,A.,PRICE,J.L.,MAN,B., AND YOUNG, M. W. Loss of circadian behavioral rhythms and per RNA oscillations in the Drosophila mutant timeless. Science 263, 5153 (Mar 1994), 1603– 1606.

[302] SEHGAL,A.,ROTHENFLUH-HILFIKER,A.,HUNTER-ENSOR,M.,CHEN, Y., MYERS, M. P., AND YOUNG, M. W. Rhythmic expression of timeless: a basis for promoting circadian cycles in period gene autoregulation. Science 270, 5237 (Nov 1995), 808–810.

[303] SELBY, C. P., AND SANCAR, A. The Second Chromophore in Drosophila Pho- tolyase/Cryptochrome Family Photoreceptors. Biochemistry 51, 1 (2012), 167–171.

[304] SHEEBA, V., SHARMA, V. K., GU,H.,CHOU, Y.-T., O’DOWD,D.K., AND HOLMES, T. C. Pigment dispersing factor-dependent and -independent circadian locomotor behavioral rhythms. J Neurosci 28, 1 (Jan 2008), 217–227.

[305] SHINDO,J.-I., AND MASAKI, S. Photoperiodic control of larval development in the semivoltine cockroach Periplaneta japonica (Blattidae: Dictyoptera). Ecological Research 10, 1 (1995), 1–12.

[306] SIEBERT, P. D., CHENCHIK,A.,KELLOGG,D.E.,LUKYANOV,K.A., AND LUKYANOV,S.A. An improved PCR method for walking in uncloned genomic DNA. Nucleic Acids Res 23, 6 (Mar 1995), 1087–1088.

[307] SIWICKI,K.K.,EASTMAN,C.,PETERSEN,G.,ROSBASH,M., AND HALL, J. C. Antibodies to the period gene product of Drosophila reveal diverse tissue distribution and rhythmic changes in the visual system. Neuron 1, 2 (Apr 1988), 141–150. BIBLIOGRAPHY 163

[308] SO, W. V., AND ROSBASH, M. Post-transcriptional regulation contributes to Drosophila clock gene mRNA cycling. EMBO J 16, 23 (Dec 1997), 7146–7155.

[309] SOKOLOVE, P. G. Localization of the cockroach optic lobe circadian pacemaker with microlesions. Brain Res 87, 1 (Apr 1975), 13–21.

[310] STANEWSKY,R.,KANEKO,M.,EMERY, P., BERETTA,B.,WAGER-SMITH,K.,KAY,S.A., ROSBASH,M., AND HALL, J. C. The cryb mutation identiﬁes cryptochrome as a circadian photoreceptor in Drosophila. Cell 95, 5 (Nov 1998), 681–692.

[311] STEHL´IK,J.,ZAVODSK´ A´ ,R.,SHIMADA,K.,SAUMAN,I., AND KOSTAL´ , V. Photoperiodic induction of diapause requires regulated transcription of timeless in the larval brain of Chymomyza costata. J Biol Rhythms 23, 2 (Apr 2008), 129–139.

[312] STENGL,M., AND HOMBERG, U. Pigment-dispersing hormone-immunoreactive neurons in the cockroach Leucophaea maderae share properties with circadian pacemaker neurons. J Comp Phys- iol A 175, 2 (Aug 1994), 203–213.

[313] STOLERU,D.,NAWATHEAN, P., FERNANDEZ´ ,M. D. L. P., MENET,J.S.,CERIANI, M. F., AND ROSBASH, M. The Drosophila circadian network is a seasonal timer. Cell 129, 1 (Apr 2007), 207–219.

[314] STOLERU,D.,PENG, Y., AGOSTO,J., AND ROSBASH, M. Coupled oscillators control morning and evening locomotor behaviour of Drosophila. Nature 431, 7010 (Oct 2004), 862–868.

[315] STOLERU,D.,PENG, Y., NAWATHEAN, P., AND ROSBASH, M. A resetting signal between Drosophila pacemakers synchronizes morning and evening activity. Nature 438, 7065 (Nov 2005), 238–242.

[316] SUH,J., AND JACKSON, F. R. Drosophila ebony activity is required in glia for the circadian regulation of locomotor activity. Neuron 55, 3 (Aug 2007), 435–447.

[317] SUN, W. C., JEONG,E.H.,JEONG,H.-J.,KO, H. W., EDERY,I., AND KIM, E. Y. Two distinct modes of PERIOD recruitment onto dCLOCK reveal a novel role for TIMELESS in circadian transcription. J Neurosci 30, 43 (Oct 2010), 14458–14469.

[318] SUNDRAM, V., NG, F. S., ROBERTS,M.A.,MILLAN´ ,C.,EWER,J., AND JACKSON, F. R. Cellular requirements for LARK in the Drosophila circadian system. J Biol Rhythms 27, 3 (Jun 2012), 183–195.

[319] SYROVA´ ,Z.,DOLEZEL,D.,SAUMANN,I., AND HODKOVA´ , M. Photoperiodic regulation of diapause in linden bugs: are period and clock genes involved? Cell Mol Life Sci 60, 11 (Nov 2003), 2510–2515.

[320] SOHLER¨ ,S.,NEUPERT,S.,PREDEL,R.,NICHOLS,R., AND STENGL, M. Localization of leucomyosuppressin in the brain and circadian clock of the cockroach Leucophaea maderae. Cell Tissue Res 328, 2 (May 2007), 443–452.

[321] SOHLER¨ ,S.,NEUPERT,S.,PREDEL,R., AND STENGL, M. Examination of the role of FMRFamide-related peptides in the circadian clock of the cockroach Leucophaea maderae. Cell Tissue Res 332, 2 (May 2008), 257–269. 164 BIBLIOGRAPHY

[322] SOHLER¨ ,S.,STENGL,M., AND REISCHIG, T. Circadian pacemaker coupling by multi- peptidergic neurons in the cockroach Leucophaea maderae. Cell Tissue Res 343, 3 (Mar 2011), 559–577.

[323] TAGAYA,J.,NUMATA,H., AND GOTO, S. G. Sexual difference in the photoperiodic induction of pupal diapause in the ﬂesh ﬂy Sarcophaga similis. Entomological Science 13, 3 (2010), 311–319.

[324] TAMURA,K.,PETERSON,D.,PETERSON,N.,STECHER,G.,NEI,M., AND KUMAR,S. MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol Evol 28, 10 (Oct 2011), 2731–2739.

[325] TANAKA,S., AND ZHU, D.-H. Presence of three diapauses in a subtropical cockroach: control mechanisms and adaptive signiﬁcance. Physiological Entomology 28, 4 (2003), 323–330.

[326] TANOUE,S.,KRISHNAN, P., KRISHNAN,B.,DRYER,S.E., AND HARDIN, P. E. Circadian clocks in antennal neurons are necessary and sufﬁcient for olfaction rhythms in Drosophila. Curr Biol 14, 8 (Apr 2004), 638–649.

[327] TAUBER,E.,ZORDAN,M.,SANDRELLI, F., PEGORARO,M.,OSTERWALDER,N.,BREDA, C.,DAGA,A.,SELMIN,A.,MONGER,K.,BENNA,C.,ROSATO,E.,KYRIACOU, C. P., AND COSTA, R. Natural selection favors a newly derived timeless allele in Drosophila melanogaster. Science 316, 5833 (Jun 2007), 1895–1898.

[328] TENG, S.-W., MUKHERJI,S.,MOFFITT,J.R., DE BUYL,S., AND O’SHEA, E. K. Robust circadian oscillations in growing cyanobacteria require transcriptional feedback. Science 340, 6133 (May 2013), 737–740.

[329] TOMIOKA,K.,URYU,O.,KAMAE, Y., UMEZAKI, Y., AND YOSHII, T. Peripheral circadian rhythms and their regulatory mechanism in insects and some other arthropods: a review. J Comp Physiol B 182, 6 (Aug 2012), 729–740.

[330] TOMIOKA,K., AND YOSHII, T. Entrainment of Drosophila circadian rhythms by temperature cycles. Sleep and Biological Rhythms 4, 3 (2006), 240–247.

[331] TOMITA, T., MIYAZAKI,K.,ONISHI, Y., HONDA,S.,ISHIDA,N., AND OISHI, K. Con- served amino acid residues in C-terminus of PERIOD 2 are involved in interaction with CRYP- TOCHROME 1. Biochim Biophys Acta 1803, 4 (Apr 2010), 492–498.

[332] TYSHCHENKO, V. P. Two-oscillatory model of the physiological mechanism of insect photoperiodic reaction. Zhurnal Obshchei Biolgii 27 (1966), 209–222.

[333] UEDA,H.R.,MATSUMOTO,A.,KAWAMURA,M.,IINO,M.,TANIMURA, T., AND HASHIMOTO, S. Genome-wide transcriptional orchestration of circadian rhythms in Drosophila. J Biol Chem 277, 16 (Apr 2002), 14048–14052.

[334] URYU,O.,KAMAE, Y., TOMIOKA,K., AND YOSHII, T. Long-term effect of systemic RNA interference on circadian clock genes in hemimetabolous insects. J Insect Physiol (Feb 2013).

[335] URYU,O.,KARPOVA,S.G., AND TOMIOKA, K. The clock gene cycle plays an important role in the circadian clock of the cricket Gryllus bimaculatus. J Insect Physiol 59, 7 (Jul 2013), 697–704. BIBLIOGRAPHY 165

[336] URYU,O., AND TOMIOKA, K. Circadian oscillations outside the optic lobe in the cricket Gryllus bimaculatus. J Insect Physiol 56, 9 (Sep 2010), 1284–1290.

[337] VOSSHALL,L.B.,PRICE,J.L.,SEHGAL,A.,SAEZ,L., AND YOUNG, M. W. Block in nuclear localization of period protein by a second clock mutation, timeless. Science 263, 5153 (Mar 1994), 1606–1609.

[338] WASSMER, G. T., AND PAGE, T. L. Photoperiodic time measurement and a graded response in a cockroach. J Biol Rhythms 8, 1 (1993), 47–56.

[339] WEBER, F. Cyclic layer deposition in the cockroach (Blaberus craniifer) endocuticle: a circadian rhythm in leg pieces cultured in vitro. Journal of Insect Physiology 41, 2 (1995), 153 – 161.

[340] WEI,H., AND STENGL, M. Light affects the branching pattern of peptidergic circadian pacemaker neurons in the brain of the cockroach Leucophaea maderae. J Biol Rhythms 26, 6 (Dec 2011), 507– 517.

[341] WEI,H., AND STENGL, M. Ca(2+) -dependent ion channels underlying spontaneous activity in insect circadian pacemaker neurons. Eur J Neurosci (Jul 2012).

[342] WEN,C.-J., AND LEE, H.-J. Mapping the cellular network of the circadian clock in two cockroach species. Arch Insect Biochem Physiol 68, 4 (Aug 2008), 215–231.

[343] WEN, H. W., AND LEE, H. J. Unequal coupling between locomotor pacemakers of the German cockroach, Blattella germanica (L.). J Insect Physiol 46, 1 (Jan 2000), 89–97.

[344] WEVER, R. Inﬂuence of weak electromagnetic ﬁelds on the circadian periodicity of humans. Naturwissenschaften 55, 1 (Jan 1968), 29–32.

[345] WHEELER,D.A.,KYRIACOU, C. P., GREENACRE,M.L.,YU,Q.,RUTILA,J.E.,ROSBASH, M., AND HALL, J. C. Molecular transfer of a species-speciﬁc behavior from Drosophila simulans to Drosophila melanogaster. Science 251, 4997 (Mar 1991), 1082–1085.

[346] WIEDENMANN. Weak and strong phase shifting in the activity rhythm of the cockroach Leu- cophaea maderae (Blaberidae) after light pulses of high intensity. Zeitschrift fur¨ Naturforschung 32C (1977), 464–465.

[347] WIEDENMANN, G. Two Activity Peaks in the Circydian Rhythm of the cockroach Leucophaea maderae. J. interdiscipl. Cycle Res. 8 (1977), 378–383.

[348] WIEDENMANN,G.,LUKAT,R., AND WEBER, F. Cyclic layer deposition in the cockroach endocuticle: a circadian rhythm? . Journal of Insect Physiology 32, 12 (1986), 1019 – 1027.

[349] WILLIAMS,J.A.,SU,H.S.,BERNARDS,A.,FIELD,J., AND SEHGAL, A. A circadian output in Drosophila mediated by neuroﬁbromatosis-1 and Ras/MAPK. Science 293, 5538 (Sep 2001), 2251–2256.

[350] WILLS,S.A.,PAGE, T. L., AND COLWELL, C. S. Circadian rhythms in the electroretinogram of the cockroach. J Biol Rhythms 1, 1 (1986), 25–37. 166 BIBLIOGRAPHY

[351] WINFREE, A. T. Integrated view of resetting a circadian clock. Journal of Theoretical Biology 28, 3 (1970), 327 – 374.

[352] YAMADA,H., AND YAMAMOTO, M.-T. Association between circadian clock genes and diapause incidence in Drosophila triauraria. PLoS One 6, 12 (2011), e27493.

[353] YILDIZ,O.,DOI,M.,YUJNOVSKY,I.,CARDONE,L.,BERNDT,A.,HENNIG,S.,SCHULZE, S.,URBANKE,C.,SASSONE-CORSI, P., AND WOLF, E. Crystal structure and interactions of the PAS repeat region of the Drosophila clock protein PERIOD. Mol Cell 17, 1 (Jan 2005), 69–82.

[354] YOSHII, T., FUNADA, Y., IBUKI-ISHIBASHI, T., MATSUMOTO,A.,TANIMURA, T., AND TOMIOKA,K. Drosophila cryb mutation reveals two circadian clocks that drive locomotor rhythm and have different responsiveness to light. J Insect Physiol 50, 6 (Jun 2004), 479–488.

[355] YOSHII, T., RIEGER,D., AND HELFRICH-FORSTER¨ , C. Two clocks in the brain: an update of the morning and evening oscillator model in Drosophila. Prog Brain Res 199 (2012), 59–82.

[356] YOSHII, T., WULBECK¨ ,C.,SEHADOVA,H.,VELERI,S.,BICHLER,D.,STANEWSKY,R., AND HELFRICH-FORSTER¨ , C. The neuropeptide pigment-dispersing factor adjusts period and phase of Drosophila’s clock. J Neurosci 29, 8 (Feb 2009), 2597–2610.

[357] YU,Q.,JACQUIER,A.C.,CITRI, Y., HAMBLEN,M.,HALL,J.C., AND ROSBASH, M. Molec- ular mapping of point mutations in the period gene that stop or speed up biological clocks in Drosophila melanogaster. Proc Natl Acad Sci U S A 84, 3 (Feb 1987), 784–788.

[358] YU, W., AND HARDIN, P. E. Circadian oscillators of Drosophila and mammals. J Cell Sci 119, Pt 23 (Dec 2006), 4793–4795.

[359] YUAN,Q.,METTERVILLE,D.,BRISCOE,A.D., AND REPPERT, S. M. Insect cryptochromes: gene duplication and loss deﬁne diverse ways to construct insect circadian clocks. Mol Biol Evol 24, 4 (Apr 2007), 948–955.

[360] ZENG,H.,QIAN,Z.,MYERS, M. P., AND ROSBASH, M. A light-entrainment mechanism for the Drosophila circadian clock. Nature 380, 6570 (Mar 1996), 129–135.

[361] ZERR,D.M.,HALL,J.C.,ROSBASH,M., AND SIWICKI, K. K. Circadian ﬂuctuations of period protein immunoreactivity in the CNS and the visual system of Drosophila. J Neurosci 10, 8 (Aug 1990), 2749–2762.

[362] ZHANG, Y., LIU, Y., BILODEAU-WENTWORTH,D.,HARDIN, P. E., AND EMERY, P. Light and temperature control the contribution of speciﬁc DN1 neurons to Drosophila circadian behavior. Curr Biol 20, 7 (Apr 2010), 600–605.

[363] ZHENG,X.,KOH,K.,SOWCIK,M.,SMITH,C.J.,CHEN,D.,WU,M.N., AND SEHGAL,A. An isoform-speciﬁc mutant reveals a role of PDP1 epsilon in the circadian oscillator. J Neurosci 29, 35 (Sep 2009), 10920–10927.

[364] ZHU,D.-H., AND TANAKA, S. Photoperiod and temperature affect the life cycle of a subtropical cockroach, Opisoplatia orientalis: seasonal pattern shaped by winter mortality. Physiological Entomology 29, 1 (2004), 16–25. BIBLIOGRAPHY 167

[365] ZHU,D.-H., AND TANAKA, S. Different photoperiodic responses in two geographic populations of a subtropical cockroach, Periplaneta japanna Asahina. Acta Ecologica Sinica 27 (2007), 3687– 3693.

[366] ZHU,H.,SAUMAN,I.,YUAN,Q.,CASSELMAN,A.,EMERY-LE,M.,EMERY, P., AND REP- PERT, S. M. Cryptochromes deﬁne a novel circadian clock mechanism in monarch butterﬂies that may underlie sun compass navigation. PLoS Biol 6, 1 (Jan 2008), e4.

[367] ZHU,H.,YUAN,Q.,BRISCOE,A.D.,FROY,O.,CASSELMAN,A., AND REPPERT, S. M. The two CRYs of the butterﬂy. Curr Biol 15, 23 (Dec 2005), R953–R954.

[368] ZIMMERMANN, J. Westernblotanalyse des Uhrproteins PERIOD bei der Schabe Leucophaea maderae. Diploma Thesis, Philipps-University Marburg (2002), 130pp.

[369] ZOLTOWSKI,B.D.,VAIDYA, A. T., TOP,D.,WIDOM,J.,YOUNG, M. W., AND CRANE,B.R. Structure of full-length Drosophila cryptochrome. Nature 480, 7377 (Dec 2011), 396–399.

[370] ZUCKERKANDL,E., AND PAULING, L. Evolutionary divergence and convergence in proteins. Edited in Evolving Genes and Proteins by V. Bryson and H.J. Vogel, Academic Press, New York. (1965), 97–166.

[371] ZAVODSK´ A´ ,R.,SEHADOVA´ ,H.,SAUMAN,I., AND SEHNAL, F. Light-dependent PER-like proteins in the cephalic ganglia of an apterygote and a pterygote insect species. Histochem Cell Biol 123, 4-5 (Jun 2005), 407–418. 168 BIBLIOGRAPHY 6 Acknowledgements

I dearly thank everyone who made this dissertation possible. First of all I would like to thank Prof. Monika Stengl for introducing me to the very interesting topic of insect chronobiology, providing the basis for my work and for her support. All members of the animal physiology labs of the Universities of Marburg and Kassel, and the Institute for Integrative Neuroanatomy, Charit´eBerlin, that I had the pleasure of getting to know, I thank for their support and many interesting conversations. The members of the examination committee I thank for taking the time to review my work and their patience. Special thanks goes to Dr. Christian Derst for his collaboration, introducing me to various molecular biology techniques, and all the leisure activities he took me to during my stay in Berlin. The department members of the University of Kassel institute of biology, especially the department of biochemistry, I wish to thank for their congenial support and help. My ofﬁce roommates in Marburg and the submarine crew in Kassel I thank for their company and the nice atmosphere. My friends and my family I thank for their constant support. There were times I would have given up without you. Finally I would dearly like to thank everyone whom I forgot. In case I did forget YOU, im sorry. I really am.

Credits

Prof. Monika Stengl (1st reviewer) Bianca Backasch Martina Kern Prof. Mireille Schäfer (2nd reviewer) El-Sayed Baz Michaela Keuper Prof. Charlotte Förster Jonas Benzler Steffen Klingenhöfer Prof. Friedrich Herberg Daniela Bertinetti Prof. Martin Klingenspor ...members of the examination committee. Oliver Bertinetti Gisela Kaschlaw Prof. Horst Bohn Matthias Knape Florian Bolze Holger Knieps Christian Derst Luzie Braulke Martin Kollmann Azar Massah Janis Brusius Christopher König Andreas Arendt Tony Burmeister Sarah Körte ...contributed to this work. Stefan Dippel Steffi Krannich Mandy Diskar Jürgen Krieger Marie-Christine Donath Michael Krug Nico Funk David Dreyer Angela Kurylas Sarah Körte Christian Eichler Alexia Loste Christiane Werckenthin Basil El-Jundi Robin Lorenz Andreas Nolte Cornelia Exner Azar Massah ...proofread my writings. Sandy Fastner Kathrin Muda Julia Fischer Christine Nowack Christian Flecke Andreas Nolte Andreas Nolte Nico Funk Rebecca Olkrug¨ Christian Flecke Roman Fricke Keram Pfeiffer El-Sayed Baz Tobias Fromme Dominik Piston Hanzey Yasar Petra Gawalek Anke Prinz Ildefonso Atienza López Thomas Graeff Ursula Reichert Nico Funk Karin Große-Mohr Thomas Reischig Sarah Körte Ewald Grosse-Wilde Frank Richter Sandra Söhler Svenja Hampl Katrin Riedinger Sandy Fastner Prof. Bill Hansson Stefanie Rulla Three coffee machines Prof. Paul Hardin Prof. Joachim Schachtner Various ornamental fish Stanley Heinze Nils-Lasse Schneider ...had to stand me as roommate. Michael Helwig Julia Schendzielorz Jennifer Hermann Thomas Schendzielorz Penelope Higgs Julia Schuckel Many thanks to all people I met during my phd thesis Verena Hirschberg Wolfgang W. Schwippert and that helped me by one means or another (In Prof. Uwe Homberg Norman Seger alphabetical order) Wolf Hütteroth Christin Sender Andreas Arendt Susanne Jaguttis Jutta Seyfarth Thordis Arnold Martin Jastroch Sandra Söhler Ildefonso Atienza López Timo Kanzleiter Ralf Stanewsky

169 170 CHAPTER 6. ACKNOWLEDGEMENTS

Regina Stieber Martina Mappes Prof. Christian Wegener Kai Stieger Alexander Tups HongYing Wei Sigrid Stöhr Christa Uthof Christina Wollenhaupt Gábor Szerencsi Maike Vetter Hanzey Yasar Magdalene Trzcionka Matthias Vömel Mechthild Zissel Bertan Türemis Isabel Walther A Appendix

5’-TGCTATCCAGGCTGTGCTTTCTCTGTACGCTTCTGGTAGAACCACTGGTATTGTGCTGG ACTCTGGTGATGGCGTCTCGCATACAGTACCAATTTATGAAGGTTATGCTCTTCCCCAT GCCATCCTACGTCTGGATCTGGCTGGCCGTGACTTGACTGACTACCTCATGAAGATCCT GACTGAGCG TGGTTACAGC-3’

Figure A.1: Part of a Rhyparobia maderae actin coding sequence.

Gene Tissue Light Regime one-way ANOVA Cosinor zero- amplitude test PFP rmPer brain(-AMe) LD12:12 0.0174 4.34 0.0561 AMe LD12:12 0.3997 1.12 0.3018 Malpighi LD12:12 0.0802 2.61 0.0875 brain LD12:12 0.2001 1.74 0.0553 brain LD6:18 0.009 5.22 0.2546 brain LD18:6 5.7598e-05 16.15 0.0010 brain DD6:18 0.1067 2.33 0.0724 brain DD18:6 0.0024 7.28 0.1359 rmTim1 brain(-AMe) LD12:12 2.961e-07 43.06 0.0054 AMe LD12:12 0.0005 10.32 0.0010 Malpighi LD12:12 0.1213 2.21 0.8067 brain DD12:12 0.0017 7.91 0.0072 brain LD6:18 3.180e-05 18.15 0.0023 brain LD18:6 6.3010e-06 24.69 6.2e-05 brain DD6:18 0.0024 7.24 0.8686 brain DD18:6 8.04643e-06 23.58 1.2e-07 rmCry2 brain(-AMe) LD12:12 2.4712e-07 44.47 0.0297 AMe LD12:12 4.09e-05 17.27 0.0526 Malpighi LD12:12 0.4756 2.61 0.6493 brain DD12:12 0.0014 8.18 0.00016 brain LD6:18 0.0157 4.47 0.0771 brain LD18:6 0.0035 6.64 0.0619 brain DD6:18 0.2386 1.58 0.0108 brain DD18:6 0.0385 3.39 0.8560

Table A.1: Statistics of the qPCR experiments. All statistics were carried out as described in Materials and Methods.

171 172 APPENDIX A. APPENDIX ai rerio Danio musculus Mus castaneum Tribolium bullata Sarcophaga pedestris Riptortus americana Periplaneta humanus Pediculus melanogaster Drosophila nigrofasciatus Dianemobius plexippus Danaus quinquefasciatus Culex mori Bombyx impatiens Bombus germanica Blattella bisignata Blattella tryoni Bactrocera mellifera Apis pernyi Antheraea gambiae Anopheles pisum Acyrthosiphon aegypti Aedes Species aoi vitripennis Nasonia domestica Musca bimaculatus Gryllus al A.2: Table ceso ubr fteisc oefebc opgnsue o euneainet n hlgntcanalysis. phylogenetic and alignments sequence for used genes loop feedback core insect the of numbers Accession per2 per3 AAM77468 per1 NP BAG07407 AAA64677 EEB13563 NP AAO48719 CPIJ007193 NP Q8MMG2 ABL61512 XP NP AAA64675 AAY40757 NP AAEL008141 Period AAD3916 BAG48878 0163 NP 001106933 52505 NP 001036975 FN377569 FN377570 cry2-1: 001011596 CAX37109 001164576 NP : NP : NP : 035197 035196 001152839 XP NP Cry-DASH Photolyase 6-4 1 Cryptochrome 2 Cryptochrome 1 Timeless BAJ16356 0163 NP 001106934 EEB17807 002432421 722912 NP 001037622 A45 AAD39548 AAD4656 ACJ08741 BAF45421 ABO31112 AAU14170 XP BAG07408 ABO38436 AAX5859 ABA62409 EDS29383 AAK11644 NP ABO38435 FN377571 cry2-2: XP 001606405 NP 001182627 XP 001077099 313179 001648498 001076794 NP 347NP 732407 001182628 321104 477188 NM 205686 173

Figure A.2: Alignment of PERIOD sequences from different insect species. Numbers at the beginning and end of each alignment box indicate the position of the ﬁrst and last residues in respect to the N- terminus. Abbreviations: C1-C6, conserved sequence blocks 1-6; C-Domain, PAS C-terminal domain; CCID, CLK:CYC inhibition domain; CLD, cytoplasmic localization domain; NLS, nuclear localization signal; PAS, PER-ARNT-SIM domain. 174 APPENDIX A. APPENDIX

Figure A.3: Alignment of PERIOD sequences continued. 175

Figure A.4: Alignment of PERIOD sequences continued. 176 APPENDIX A. APPENDIX

Figure A.5: Alignment of CRYPTOCHROME 2 sequences from different insect species. Numbers at the beginning and end of each alignment box indicate the position of the ﬁrst and last residues in respect to the N-terminus. Abbreviations: RD1, repressive domain 1 ); RD2a, repressive domain 2a; RD2b, repressive domain 2b; NLS, nuclear localization signal. 177

Figure A.6: Alignment of CRYPTOCHROME 2 sequences continued. 178 APPENDIX A. APPENDIX

Figure A.7: Alignment of TIMELESS 1 sequences from different insect species. Numbers at the beginning and end of each alignment box indicate the position of the ﬁrst and last residues in respect to the N-terminus. Abbreviations: PER-1: PER binding domain 1; PER-2: PER binding domain 2.

Muscle Malpighi Brain TAG kDa 6 18 6 18 6 18 6 18 -

200 150

120 < PER 100

85 70 60

Figure A.8: The anti rmPER (mouse) antibody detected a protein of the same size as the anti rmPER (rabbit) antibody in R. maderae nervous tissue, which most probably represents rmPER. A western blot of different R. maderae tissues is shown. The rmPER (mouse) antibody was diluted 1:1:000. 10 µg total protein were used on a 5 % gel. 179

ZT kDa 0 4 8 12 16 20

200 150

120 < PER 100 85 70

Figure A.9: The anti Antheraea pernyi PER-S peptide antibody detected a protein of the same size as the anti rmPER (rabbit) antibody in R. maderae nervous tissue, which most probably represents rmPER. No apparent differences between ZT (LD 12:12) could be observed. A western blot time series of R. maderae brain tissues is shown. The anti Antheraea pernyi PER-S peptide antibody was diluted 1:1000, 10 µg total protein were used on a 5 % gel.

ZT kDa 0 4 8 12 16 20

260 250

140 130 < PER

100 100 70 70 < TUB

Figure A.10: Different western blots suggested different maxima using the anti rmPER (rabbit) antibody. Note that the background also changes between ZTs (LD 12:12) and that there are no differences in the loading control. A western blot time series of R. maderae brain tissues is shown. The anti rmPER (rabbit) antibody was diluted 1:10000, the TYR-TUBULIN antibody 1:10000. 2 µl homogenate were used on a 7.5 % gel. 180 APPENDIX A. APPENDIX

ZT kDa 0 4 8 12 16 20

260 250

140 130 < PER 100 100

70 70

55 50 < TUB

Figure A.11: Another example of a ZT (LD 12:12) series using the anti rmPER (rabbit) antibody. A western blot time series of R. maderae brain tissues is shown. The anti rmPER (rabbit) antibody was diluted 1:10000, the TYR-TUBULIN antibody 1:10000. 2 µl homogenate were used on a 7.5 % gel.

A B C

Figure A.12: Anti rmPER (mouse, A), anti rmTIM1 (rabbit, B) and anti rmCRY2 (rabbit, C) antibodies did not recognize any epitopes in ﬂuorescent immunohistochemical stainings. Maximum projections of CLSM image stacks (Z-direction: 23 µm) of tissue near the AMe are shown. The primary antibodies were diluted 1:500. Scalebar: 50 µm. 181

Figure A.13: Actograms of R. maderae males kept in LD 6:18, 100 lx, trial I. Information above blots indicates the number of the animal within the trial and the highest count of wheel turns per minute. White background indicates light, gray darkness and red that no data was recorded. Abbreviations: rpm, revolutions per minute 182 APPENDIX A. APPENDIX

Figure A.14: Actograms of R. maderae males kept in LD 6:18, 100 lx, trial II. Information above blots indicates the number of the animal within the trial and the highest count of wheel turns per minute. White background indicates light, gray darkness and red that no data was recorded. Abbreviations: rpm, revolutions per minute 183

Figure A.15: Actograms of R. maderae males kept in LD 18:6, 100 lx, trial I. Information above blots indicates the number of the animal within the trial and the highest count of wheel turns per minute. White background indicates light, gray darkness and red that no data was recorded. Abbreviations: rpm, revolutions per minute 184 APPENDIX A. APPENDIX

Figure A.16: Actograms of R. maderae males kept in LD 18:6, 100 lx, trial II. Information above blots indicates the number of the animal within the trial and the highest count of wheel turns per minute. White background indicates light, gray darkness and red that no data was recorded. Abbreviations: rpm, revolutions per minute 185

Figure A.17: Actograms of R. maderae males kept in LD 6:18, 1000 lx. Information above blots indicates the number of the animal within the trial and the highest count of wheel turns per minute. White background indicates light, gray darkness and red that no data was recorded. Abbreviations: rpm, revolutions per minute 186 APPENDIX A. APPENDIX

Figure A.18: Actograms of R. maderae males kept in LD 6:18, 1000 lx, continued. Information above blots indicates the number of the animal within the trial and the highest count of wheel turns per minute. White background indicates light, gray darkness and red that no data was recorded. Abbreviations: rpm, revolutions per minute 187

Figure A.19: Actograms of R. maderae males kept in LD 18:6, 1000 lx. Information above blots indicates the number of the animal within the trial and the highest count of wheel turns per minute. White background indicates light, gray darkness and red that no data was recorded. Abbreviations: rpm, revolutions per minute 188 APPENDIX A. APPENDIX

[ht] Figure A.20: Actograms of R. maderae males kept in LD 18:6, 1000 lx, continued.Information above blots indicates the number of the animal within the trial and the highest count of wheel turns per minute. White background indicates light, gray darkness and red that no data was recorded. Abbreviations: rpm, revolutions per minute Zusammenfassung

Der Wechsel von Tag und Nacht erzeugt einen regelmäßigen Rhythmus von verschiedenen Um- weltreizen, allen voran Licht und Temperatur. Fast jedes bis zum heutigen Tage untersuchte Lebewesen besitzt einen endogenen Mechanismus zur Zeitwahrnehmung, und diese ”innere Uhr” befähigt Lebewesen dazu, sich vorausschauend an rhythmische Umwelt-Anderungen¨ anzupassen. Circadiane Rhythmen bestehen auch ohne jegliche äußere Reize und basieren auf einem molekularen Rückkopplungs-Mechanismus, der Rhythmen in Genexpression und Proteinkonzen- tration von etwa 24 Stunden erzeugt. Obwohl sich die grundsätzlichen Mechanismen und Kom- ponenten dieses molekularen Uhrwerks in allen Insekten ähneln, zeigte sich jedoch immer mehr, dass es im Detail doch wesentliche Unterschiede zwischen verschiedenen Insektengruppen gibt. Während das molekulare Uhrwerk der Fruchtfliege Drosophila melanogaster inzwischen sehr gut untersucht ist, fehlen bei den meisten Insektengruppen immernoch eingehende Untersuchun- gen. Fast nichts ist über die molekulare Basis von circadianen Rhythmen bei der Schabe Rhy- parobia maderae bekannt, obwohl diese Art bereits seit Langem als Modellorganismus in der Chronobiologie dient. Um mit der Forschung am molekularen, circadianen System von R. maderae zu beginnen, wurde die Struktur und das Expressionsprofil der core feedback loop Gene per, tim1 und cry2 analysiert. Mittels degenerierten Primern und RACE konnte das vollständige offene Leseraster (OLR) von rmPer und rmCry2, und ein Teil des rmTim1 OLR kloniert werden. Eine phylogenetische Analyse gruppierte rmPER und rmCRY2 gemeinsam mit den Orthologa hemimetaboler Insekten. Viele bei D. melanogaster funktionell charakterisierte Domänen sind bei diesen Proteinen konserviert, was auf eine ähnliche Funktion in der inneren Uhr von R. maderae hinweist. Mittels quantitativer PCR konnte gezeigt werden, dass die mRNA von rm- Per, rmTim1 und rmCry2 in verschiedenen Lichtregimen in der gleichen Phasenlage Tageszeit- abhängig schwankt. Die Phasenlage stellte sich bei unterschiedlichen Photoperioden jeweils relativ zum Beginn der Skotophase ein, mit Maxima in der ersten Hälfte der Nacht. Auch im Dauerdunkel zeigen sich Rhythmen in der rmTim1 und rmCry2 Expression. Die Amplitude der rmPer Expressionsrhythmen war jedoch so gering, dass keine signifikanten Unterschiede zwischen den einzelnen Zeitgeberzeiten (ZT) festgestellt werden konnten. Mittels Laufrad-Assays wurde untersucht wie Kurz- und Langtag Lichtregime die Verhaltensrhythmen beeinflussen. Es konnten nur Unterschiede in der Periodenlänge unter freilaufenden Bedingungen festgestellt werden, wenn höhere Lichtintensitäten (1000 lx) zur Synchronisation (entrainment) genutzt wurden. Die Periode des freilaufenden Rhythmus war bei Tieren aus dem Kurztag länger. Die photoperiodische Plastizität zeigte sich also auch auf Verhaltensebene, obwohl höhere Lichtintensitäten notwendig waren um einen Effekt zu beobachten. Basierend auf den Sequenzen der zuvor klonierten OLR wurden gegen rmPER, rmTIM1 und rm- CRY2 gerichtete Antikörper hergestellt. Die Antikörper gegen rmPER und rmTIM1 erkannten in western blots sehr wahrscheinlich spezifisch das jeweilige Protein. Zeitreihen von Gehirngewebe- Homogenisaten zeigten keinen offensichtlichen circadianen Rhythmus in der Proteinkonzen- tration, wahrscheinlich auf Grund einer Oszillation mit niedriger Amplitude. In Immunhis- tochemischen Färbungen konnte nur mit dem gegen rmPER gerichteten Antikörper aus Kan-

189 190 APPENDIX A. APPENDIX inchen ein Signal beobachtet werden. Beinahe jede Zelle des Zentralnervensystems war rmPER- immunreaktiv im Zellkern. Es konnten keine Unterschiede zwischen den untersuchten ZTs festgestellt werden, ähnlich wie bei den western blot Zeitreihen. In dieser Studie konnten erstmals molekulare Daten der circadianen Uhr von R. maderae erfasst und dargestellt werden. Die Uhrgene per, tim1 und cry2 werden in dieser Schabenart exprimiert und ihre Domänenstruktur sowie das circadiane Expressionsmuster ähneln dem hypothetischen ursprünglichen Insektenuhrwerk, welches der circadianen Uhr von Vertebraten nahesteht. Das molekulare Uhrwerk von R. maderae kann sich an unterschiedliche Photoperioden anpassen, und diese Anpassungen manifestieren sich im Expressionsprofil der untersuchten Uhrgene ebenso wie im Verhalten.