Odysseus基因對果蠅精子生成的影響

(1)

國立臺灣大學生命科學院生態學與演化生物學研究所碩士論文

Institute of Ecology and Evolutionary Biology College of Life Science

National Taiwan University Master Thesis

Odysseus基因對果蠅精子生成的影響

Characterization of Odysseus in Drosophila Spermatogenesis

吳家翔

Chia-Hsiang Wu

指導教授：丁照棣aa

中華民國102年1月 January, 2013

國立臺灣大學生態學與演化生物學研究所

碩士論文

O d ys seu s

基因對果蠅精子生成的影響

吳家翔撰

102 1

Advisor: Chau-Ti Ting

(2)

i

(3)

ii

Acknowledgements

I would like to thank my advisor, Chau-Ti Ting, for all her guidance and immense patience. I am grateful for every opportunity she gave me to learn in her lab. Her

infectious enthusiasm for science and truth always makes me wonder what it is like to be like that. I want to thank all the other members in Ting lab, especially Hsin-Yen Liu and Wen-Chiao Chan. They are always ready to help in every way and I do not think I could ever make it without them. My thanks also goes to the former members of Ting lab, Ya-Jen Cheng, Shi-Yow Yang, Chi-Chun Chen, Cheng-Lin Li, and Kevin Wei. I enjoyed and appreciated their company when I was too clumsy to know what to do.

I would like to thank Shu Fang and all the other members of Fang Lab at Academia Sinica. Without their help I would not be able to access facilities there (the Darkroom, especially). I also thank my committee members, Yu-Chen Tsai at Tunghai University and Haiwei Pi at Chang Gung University, for input and guidance to my thesis work. I thank the Core Facilities at ICOB, Academia Sinica, the DNA Sequencing Service at IBMS, Academia Sinica and the Technology Commons at National Taiwan University for technical support. I am grateful to Chuang-Yu Lin at ICOB for his help and advice about imaging, which is an important part of my work and I cannot thank him enough.

I thank Chung-I Wu for allowing me to visit his lab in spring and summer of 2011.

I thank Mao-Lien Wu for her help at the bench and in daily life. My special thanks goes to Huyen Vu Thanh and her lovely family, Kimmy and Eric, for their hospitality and kindness. My roommates at 6128 S Dorchester Ave. were wonderful people and I owe them so much. My friends Yian and Artur should be awarded the Key to the City for humanitarian work in Chicago. They are heroes who save the Windy City.

(4)

iii

I thank my college advisor, Shaw-Yhi Hwang, for all the cozy talks and nice coffee. I thank my friends for watching me and backing me up all the way: Jeff Lin, Tsung-Hsi Lin, Tzu-Wei Chan, Griffin Chang, Shi-Jun Chien, Wen-Che Tsai, Christy Lee, Eileen Lin, Flore Sun, Gennie Lin, and Tricia Lee. I am lucky and honored to meet long-time friends, Tien-Yin Huang, Cathy Hsu, and Wei-Che Hsu, in my early days when the friendship got started without self-interest. They are my most valuable assets.

Last but not least, I want to thank my family for their endless love and tolerance.

Difficult kids who want to be a rebel at age 30 can be reeeeeally troublesome. For this I owe my life to them.

(5)

iv

摘要

在物種形成的過程中，雜交不合（包括雜交不活與雜交不孕）能夠形成生殖 屏障。因此，了解造成雜交不合的基因，將有助於認識物種的起源。OdsH 是由 unc-4 基因複製而產生的一個基因。在秀麗隱桿線蟲中，unc-4 與神經發育有關。OdsH 是造成兩近緣種果蠅——擬黃果蠅與模里西斯果蠅——雜交子代雄性不孕的主要

因子，據信可能是因 OdsH^mau在精巢中異常表現、或因 OdsH^mau對擬黃果蠅的 Y

染色體異常結合所導致。在黃果蠅中，剔除 OdsH 使雄蟲生殖力下降；免疫染色實 驗顯示，缺少 OdsH 表現的精巢，其生殖細胞數目較少。在本研究中，我測試了六 株 OdsH RNAi 果蠅，其中四株來自果蠅品系中心，兩株為自行建立。我並使用了 在特定位置表現綠色螢光蛋白的果蠅，標記精巢內部不同細胞，觀察 OdsH 對黃果 蠅精子生成的影響。本研究展示六株 OdsH RNAi 果蠅的效果，有助於未來研究 OdsH 與 unc-4 的功能重疊。對黃果蠅精巢的觀察結果，則暗示 OdsH 可能參與精 子生成當中的減數分裂與後期分化。

關鍵字：果蠅、基因重複、雜交不合、OdsH 基因、RNA 干擾、精子生成、unc-4 基因

(6)

v

Abstract

Hybrid incompatibility, including hybrid inviability and hybrid sterility, can act as a reproductive barrier in the process of speciation. Therefore, genes involved in hybrid incompatibility may provide a window on how new biological species form. The gene Odysseus-site homeobox (OdsH) is duplicated from unc-4, a homolog of a neuron developmental gene in Caenorhabditis elegans. OdsH has been identified as a crucial element that triggers hybrid male sterility in reciprocal crosses between two sibling species, Drosophila simulans and D. mauritiana. Previous studies have suggested that misexpression of OdsH^mau in testes, or additional binding of OdsH^mau to the Y

chromosome of D. simulans, results in the hybrid sterility. In D. melanogaster, targeted gene knockout revealed that lack of OdsH expression reduces male fertility.

Immunostaining in OdsH null mutant (OdsH⁰) testes has shown a decrease in the number of germ cells. In this thesis, I generated two OdsH RNAi constructs and examined their efficiency with the other four available at stock centers. In addition, I used reporter lines that express enhanced green fluorescent protein to further

characterize this testis-specific gene at the cellular level. My results demonstrated the efficiency of these RNAi strains, which should facilitate future studies on the functional redundancy between OdsH and unc-4. Cytological observation implies that OdsH may play a role in meiosis and terminal differentiation during spermatogenesis.

Keywords: Drosophila, gene duplication, hybrid sterility, OdsH, RNA interference, spermatogenesis, unc-4

(7)

vi

List of Figures

Figure 1. Synthesis of mir-6-1-based microRNA ... 10

Figure 2. Valium22 vector ... 15

Figure 3. Generation of pV22_OdsH^simAB ... 16

Figure 4. Generation of pV22_OdsH^mauAB ... 17

Figure 5. Drosophila eye images from wild-type and transgenic flies ... 21

Figure 6. Normalized values of eye area to w¹¹¹⁸ control ... 22

Figure 7. Normalized values of eye circumference to w¹¹¹⁸ control ... 23

Figure 8. RT-PCR detection of OdsH^mel expression in wild-type and transgenic male flies ... 24

Figure 9. His-GFP expression in testes of 0- and 32-day-old OdsH⁺ and OdsH⁰ flies ... 27

Figure 10. The average volume of His-GFP signal per testis ... 28

Figure 11. Sa-GFP expression in testes of 1-, 15-, and 55-day-old OdsH⁺ and OdsH⁰ flies ... 29

Figure 12. The average volume of Sa-GFP signals in spermatocytes ... 30

Figure 13. BamP-GFP expression in testes of 1-, 10-, and 32-day-old OdsH⁺ and OdsH⁰ flies ... 31

Figure 14. OdsH structure and the target regions of UAS-OdsH^RNAi strains used in this thesis... 33

Figure 15. Male germline differentiation in Drosophila melanogaster ... 37

(10)

ix

Figure 16. tMAC (aly-class genes) and tTAFs (can-class genes) network in Drosophila melanogaster spermatocytes ... 41 Figure 17. Design of pCaSpeR4_V22_OdsH^simAB, a pCaSpeR4-based RNAi vectors for

P-element transformation ... 48 Figure 18. Retrival of miRNA construct against OdsH^sim from pV22_OdsH^simAB into

attB-P[acman]-Cm^R by recombineering ... 50

(11)

x

List of Tables

Table 1. Fly strains used in this thesis ... 9 Table 2. Oligo names and sequences used in this thesis ... 11 Table 3. Summary of the introduction of mir-6-1-based RNAi constructs into

Drosophila melanogaster via C31 integrase-mediated recombination ... 20

Table 4. The target sequences of UAS-OdsH^RNAi strains obtained from stock centers .... 34 Table 5. Homology of the target sequences of mir-6-1-based RNAi constructs in

Drosophila melanogaster, D. simulans and D. mauritiana ... 34 Table 6. Oligo sequences for mir-6-1-based RNAi synthesis ... 52 Table 7. Synthesized miRNA sequences ... 53

(12)

1

Introduction

People have long been fascinated by the origin of species. From Aristotle (384 BC-322 BC) to Carl Linnaeus (1707-1778), philosophers strived to classify the creatures on Earth and assigned the order in which all lives are created. Conceivably, one of the earliest definitions of species pertains to the morphological similarity among individuals, because there is no more instinctive way to categorize lives than the observable and measureable traits by the naked eye. In face of vast numbers of species, people held different view on their origin. In the view of creationism, species today are just the same as they were when the Earth was formed. In contrast, some people doubted and

questioned the possibility of creating new species. In the pre-Darwinian era, challenges against the deep-rooted ecclesiastical perspective begun emerging already. Such as Linnaeus, based on his classification work, pointed that a new species could arise from hybridizing plants. Jean-Baptiste Lamarck (1744-1829) also advanced the idea of hybridization as a source of new species (Reviewed in MCCARTHY 2008). By the time Charles Darwin (1809-1882) was born, the Age of Enlightenment came to an end. It was a society where reason was upheaved to the level comparable to church traditions, a society where research fellows received supports from state-funded academies, and a society where salons, coffee houses and debate clubs permeated.

Several keys contributed to Darwin’s achievement in the theory of evolution: his immense curiosity and interest in natural history, the upper-class background and financially secure life that enabled him to be a gentleman scientist on HMS Beagle (1831-1836) (BROWNE 2009), and befriending his contemporary scientists, like Asa Gray (1810-1888), John Gould (1804-1881), John Herschel (1792-1871), Joseph Dalton Hooker (1817-1911) and Charles Lyell (1797-1875). Among them, Lyell’s

(13)

2

uniformitarian view of geology—the same laws and processes of nature have always operated in the past and at present on Earth in the same way— helped shape Darwin’s speculations on the “mystery of mysteries”(DARWIN 1859). Finally, a manuscript on natural selection from Alfred Russel Wallace (1823-1913) in 1858 gave impetus to the publication of Darwin’s painstaking work and the cornerstone of evolutionary biology as well in 1859. In On the Origin of Species, Darwin proposed natural selection as the fundamental mechanism of evolution. It occurs when descents, from a common ancestor, carry heritable traits with variation related to fitness. Individuals with certain traits may have better chance to survive and reproduce than the others without. This can

consequently lead to a specialized population, or a new species, that adapts to a particular way of life (DARWIN 1859).

Upon the arrival of Darwin’s theory, the community was not ready for this radical idea. One reason that it came under suspicion was the absence of evidence to support.

Although Gregor Mendel (1822-1884) published his seminal work on modern genetics in 1865, Darwin was not aware of it and failed to explicate his crossing experiment results with the blending inheritance and his pangenesis theory (Reviewed in

CHARLESWORTH and CHARLESWORTH 2009). It was not until the turn of the 20^th century that Mendel’s laws were rediscovered and advocated by Hugo de Vries (1848-1935), Carl Correns (1864-1933), Erich von Tschermak (1871-1962) and William Bateson (1861-1926) (Reviewed in KEYNES and COX 2008). Yet the evolution and species problems were not all solved happily ever after. For example, Bateson, an English biologist who coined the term “genetics,” did not lend support to Darwin’s idea. In fact, since the renaissance of Mendelian inheritance, there were dissenting voices from Darwin’s natural selection. These voices, including de Vries and Bateson, believed that

(14)

3

a new species arises from mutations that cause large, discontinuous modification of individuals. This theory—referred to as mutationism or Mendelism—was defied by the biometricians, who followed Darwin’s view and concentrated on small, continuous variations among individuals (Reviewed in AYALA and FITCH 1997). About the same time, explosive progress in genetics enabled scientists to trace the inheritance factors in cells. Thomas Hunt Morgan (1866-1945) used Drosophila melanogaster to demonstrate the chromosome theory of inheritance proposed by Walter Sutton (1877-1916) and Theodor Boveri (1862-1915). For the first time, a gene was showed physically on a specific chromosome as a carrier of traits. The Morgan group—including his students Alfred Sturtevant (1891-1970), Calvin Bridges (1889-1938) and Hermann Joseph Muller (1890-1967)—became the trailblazer for the further understanding of modern genetics. Together, they established what is common knowledge for geneticists today, such as genetic linkage, polytene chromosome map and X-ray mutagenesis (Reviewed in DAVIS 2004).

The mutationist-biometrician contradiction was reconciled by the introduction of mathematical statistics into genetics in the 1920s and 1930s. Ronald Fisher (1890-1962), Sewall Wright (1889-1988), and John Burdon Sanderson Haldane (1892-1964) were among the most influential figures that developed the theoretical framework of

population genetics. Their work demonstrated that natural selection could act on small, continuous variations that are passed on from parents to offspring through Mendelian inheritance and contribute to morphological and functional alterations. Furthermore, it led to the demise of mutationism and stimulated the forthcoming synthetic theory of evolution, the Modern Synthesis (Reviewed in AYALA and FITCH 1997; CHARLESWORTH

and CHARLESWORTH 2009).

(15)

4

At the dawn of the Modern Synthesis era, two prominent scientists, Theodosius Dobzhansky (1900-1975) and Ernst Mayr (1904-2005), made tremendous contribution in combining the evolution theory and multiple scientific disciplines. Both of them proposed what is known as the biological species concept: species are natural populations that are able to interbreed and reproductively isolated from other such populations. Since then, the discussion about isolating mechanisms has been the core to decode speciation (DOBZHANSKY 1937; MAYR 1942). In 1936, Dobzhansky published his groundbreaking work in the genetics of speciation (DOBZHANSKY 1936). He crossed two species, D. pseudoobscura and D. persimilis, and backcrossed the fertile F₁ females to males of the two species. He thus obtained hybrids with many combinations of heterospecific chromosomes. With the help of visible markers, the fertility of these hybrids, which was measured by the testis size, indicated which region on whose chromosome caused hybrid sterility—a reproductive barrier that is required to create a new species. Dobzhansky gave empirical evidence that genes, instead of cytoplasm difference or chromosome rearrangement, cause hybrid sterility. Also, through this experiment, he provided an answer to Darwin’s paradox: how can sterility or inviability survive through natural selection? Dobzhansky proposed an interaction between (at least) two loci responsible for hybrid sterility: Assume two allopatric populations with the same genotype aabb. Then a mutation occurs: now one population comprises Aabb and AAbb, while the other comprises aaBb and aaBB. The new allele A is compatible with b, and B is compatible with a. In this scenario, allele A and B are never put to the test together. When the two populations meet, the incompatibility between A and B can lead the hybrid offspring to a dead end. This idea, also discussed and contributed by William Bateson (1909) and H. J. Muller (1942), has become known as the

(16)

5

Bateson-Dobzhansky-Muller Model (Reviewed in ORR 1996).

Dobzhansky’s pioneering experiment did not add zest to the seeking of hybrid incompatibility genes. Conversely, interest in this subject was not revived until early 1980s, when Jerry Coyne launched the study to answer the genetic basis of hybrid male sterility between D. simulans and D. mauritiana (COYNE 1984; JOHNSON 2008). Coyne and Charlesworth (1986) further traced the largest sterility effect to the X chromosome and found it closely linked to the marker, forked. Using molecular markers, Perez et al.

(1993) were able to associate the sterility effect to a major locus and named it Odysseus (Ods). The name “Odysseus”—the hero who devised the wooden horse that destroyed the city of Troy in the Homeric epic—is metaphorical: the D. mauritiana allele invades the D. simulans background and finally leads to a crash of the lineage. A subsequent investigation of Ods revealed that Ods^mau alone cannot cause sterility; only when Ods^mau and its distal region were cointrogressed, the sterility effect can be fully exerted (PEREZ

and WU 1995). Finally, an attempt of fine-scale genetic mapping successfully delineated the Ods locus to the nucleotide level (TING et al. 1998). The location of Ods was

defined by two introgression lines from a collection of 190 newly generated

recombinants: 99F (the fertile line that has the longest introgressed fragment from D.

mauritiana) and 193S (the sterile line that has the shortest introgressed fragment from D.

mauritiana), which differ only by a 3-kb region where reside the (later known) exons 3 and 4 of a gene. Since this gene contains a homeobox in the exons 2 and 3, it was named Odysseus-site homeobox gene (OdsH).

The discovery of OdsH has posed several questions: First, sequence analysis showed that OdsH, particularly its homeodomain, has undergone rapid evolution, which is asymmetric to its highly conserved paralog, unc-4 (TING et al. 1998). Second, the fact

(17)

6

that the male reproductive organ-specific OdsH duplicated from the neuron-specific unc-4 (TABUCHI et al. 1998) supports a hypothesis that a duplicated gene, which is liable to functional divergence and has not reached its stable maintenance, is a candidate

“speciation gene” (TING et al. 2004). Third, since the maladaptive defect induced by OdsH in hybrids is an evolutionary by-product, what is the normal function of OdsH? In D. melanogaster, OdsH is suggested to enhance sperm production, probably by speeding up sperm maturation (SUN et al. 2004) and/or maintaining germline stem cells (CHENG

et al. 2012). Fourth, what is the mechanism underlying the hybrid sterility? The homeodomain motif supports an educated guess that OdsH is a DNA-binding

transcription factor. The misregulated OdsH in D. simulans-D. mauritiana hybrid sterile line disrupts the expression of genes involved in spermatogenesis, which may be the cause of spermatogenic failure (LU et al. 2010; SUN et al. 2004). However, Bayes and Malik reported an additional heterochromatic localization of OdsH^mau on the Y

chromosome of D. simulans may cause chromosome decondensation in sterile hybrids, which suggests a “gain-of-function” protein-DNA interaction that echoes the

Bateson-Dobzhansky-Muller Model (BAYES and MALIK 2009).

In this thesis, I tested four OdsH RNAi strains available from stock centers and two that newly generated at my hands. I used reporter lines that express green

fluorescent protein (GFP) to characterize this testis-specific gene at the cellular level.

These efforts are intended to tackle two questions: (1) The function redundancy between OdsH and its paralog, unc-4, and (2) The specific role of OdsH in D. melanogaster spermatogenesis. Together, this thesis provides some new hints on genes involved in hybrid incompatibility and materials for future study.

(18)

7

Materials and Methods

Fly strains and husbandry

All Drosophila strains were raised at 25˚C on standard cornmeal medium.

his-GFP (CLARKSON and SAINT 1999), sa-GFP (CHEN et al. 2005) (a gift from Margaret T. Fuller at Stanford University), and bamP-GFP (CHEN and MCKEARIN 2003) (a gift from Haiwei Pi at Chang Gung University) were crossed to OdsH⁺ and OdsH⁰ flies (SUN et al. 2004) to visualize germ cells in different developmental stages of spermatogenesis. UAS-OdsH^RNAi strains were obtained from the Vienna Drosophila RNAi Center at the Campus Vienna Biocenter (GD51289 and KK103949) and the Transgenic RNAi Project at the Harvard Medical School (JF02198 and HMS01554).

GMR>OdsH^mel was used for testing RNAi efficiency. D. melanogaster y¹ sc¹ v¹

P{nos-phiC31\int.NLS} X; P{CaryP}attP2 was provided by Rainbow Transgenic Flies, Inc. (Camarillo, CA) and used for transgenesis. FM6 was used to make homozygous sa-GFP, OdsH⁺ and sa-GFP, OdsH⁰ flies. See Table 1 for strain information.

Synthesis of mir-6-1-based microRNA for RNAi constructs

The design of mir-6-1-based miRNA against OdsH was based on Chen et al.

(2007). A target site on OdsH^sim coding region was selected according the criteria as follows: (1) a 22-nucleotide sequence, (2) a 30-52% guanine-cytosine content, (3) at least three adenine-thymine within the 16^th to 20^th nucleotide, (4) an adenine at the 20^th nucleotide, (5) no guanine at the 13^th nucleotide, (6) an adenine at the 3^rd nucleotide (optional), and (7) an thymine at the 10^th nucleotide (optional). The resulting 22-nt sequence was then swapped to the Oligos 1 and 2 cassettes with the 20^th nucleotide in the Oligo 1 replaced by cytosine. Then, the last four nucleobases of the Oligo 2 were

(19)

8

changed to the reverse-complement counterpart of the last four nucleobases of the modified 22-nt sequence in Oligo 1. Anneal Oligos 1 and 2 by the first-round PCR using Expand High Fidelity^PLUS PCR System (Roche), which generated the mir-6-1 mimic sequence that later formed the miRNA precursors (Figure 1, A). This product was amplified by the second-round PCR with Oligos 3 and 4, which added the mir-6-1 flanking sequences and restriction recognition sites NotI/BglII and BamHI/XbaI at 5-prime and 3-prime ends respectively (Figure 1, B and C). The resulting

PCR-amplified fragment was confirmed by sequencing after cloned into pCR™4-TOPO® TA vector (Invitrogen) and the construct was named mir-6-1-OdsH^simA. By the same method the constructs mir-6-1-OdsH^simB, mir-6-1-OdsH^mauA and mir-6-1-OdsH^mauB were generated. See Table 2 for oligo sequences.

(20)

9

Table 1. Fly strains used in this thesis.

Strain Genotype Note

w¹¹¹⁸ w¹¹¹⁸

FM6 FM6, y¹ dm⁺ w¹/Dp(1;Y)y⁺ Bloomington stock 4327

OdsH⁺ w¹¹¹⁸ OdsH⁺ SUN et al. (2004)

OdsH⁰ w¹¹¹⁸ OdsH⁰ SUN et al. (2004)

GMR>OdsH^mel w^*; P{GAL4-ninaE.GMR}12; P{w⁺ UAS-OdsH^mel 9B} LIN (2009) his-GFP w¹¹¹⁸; P{His2AvT:Avic\GFP-S65T

}62A CLARKSON and SAINT (1999); Bloomington stock 5941

sa-GFP y w P{w⁺ sa-GFP} CHEN et al. (2005)

bamP-GFP hsFLP; Sp/Cyo; P{w⁺ -898/+133-bam:GFP} CHEN and MCKEARIN (2003)

GD51289 w¹¹¹⁸; P{w⁺ GD51289} UAS-OdsH^RNAi from VDRC

KK103949 w¹¹¹⁸; P{w⁺ KK103949} UAS-OdsH^RNAi from VDRC

JF02198 y¹ v¹; P{TRiP.JF02198}attP2 UAS-OdsH^RNAi from TRiP HMS01554 y¹ sc^* v¹; P{TRiP.HMS01554}attP2 UAS-OdsH^RNAi from TRiP

attP2 y¹ sc¹ v¹ P{nos-phiC31\int.NLS}X; P{CaryP}attP2 NI et al. (2008); Bloomington stock 25710

(21)

10

NotI BglII BamHI XbaI

(A)

(B)

(C)

(D)

Figure 1. Synthesis of mir-6-1-based microRNA. (A) Oligos 1 and 2, embedded a 22-nt sequence (yellow and red) that targets against the gene of interest, are annealed by the first-round PCR. (B) Oligos 3 and 4 are used to amplify the first-round PCR product, creating (C) a ~230-bp fragment flanked by NotI/BglII and BamHI/XbaI sites at 5-prime and 3-prime end, respectively. (D) As the miRNA template placed under the control of a promoter is transcribed, the hairpin-structured primary miRNA forms. This artificial primary miRNA then enters the endogenous RNAi pathway and acts as a functional miRNA.

promoter

Oligo 1

Oligo 2

Oligo 3

Oligo 4

(22)

11

Table 2. Oligo names and sequences used in this thesis.

Name Sequence (5′ to 3′)

“Oligo 1” for mir-6-1-based RNAi synthesis

miR6_OdsHsim_A1 GGCAGCTTACTTAAACTTAATCACAGCCTTTAATGTATAGCGGTTTGGTGTCAAACTATAAGTTAATATACCATATC miR6_OdsHsim_B1 GGCAGCTTACTTAAACTTAATCACAGCCTTTAATGTAAATCGTCCTTCAGCATCACATTAAGTTAATATACCATATC miR6_OdsHmau_A1 GGCAGCTTACTTAAACTTAATCACAGCCTTTAATGTGGAGAAAGCCTTCCAGGAACATTAAGTTAATATACCATATC miR6_OdsHmau_B1 GGCAGCTTACTTAAACTTAATCACAGCCTTTAATGTTCGAGGATATGGAAGTGGACGTTAAGTTAATATACCATATC

“Oligo 2” for mir-6-1-based RNAi synthesis

miR6_OdsHsim_A2 AATAATGATGTTAGGCACTTTAGGTACATAGCGGTTTGGTGTCAAAATATAGATATGGTATATTAACTTATAGT miR6_OdsHsim_B2 AATAATGATGTTAGGCACTTTAGGTACAAATCGTCCTTCAGCATCAAATTAGATATGGTATATTAACTTAATGT miR6_OdsHmau_A2 AATAATGATGTTAGGCACTTTAGGTACGGAGAAAGCCTTCCAGGAAAATTAGATATGGTATATTAACTTAATGT miR6_OdsHmau_B2 AATAATGATGTTAGGCACTTTAGGTACTCGAGGATATGGAAGTGGAAGTTAGATATGGTATATTAACTTAACGT

“Oligo 3” and “Oligo 4” for mir-6-1-based RNAi synthesis

mi6_5′ NotI_BglII GGCGCGGCCGCCGCCAGATCTTTTAAAGTCCACAACTCATCAAGGAAAATGAAAGTCAAAGTTGGCAGCTTACTTAAACTTA mi6_3′ BamHI_XbaI GGCCTCTAGAACGGATCCAAAACGGCATGGTTATTCGTGTGCCAAAAAAAAAAAAAATTAAATAATGATGTTAGGCAC

“LA_F” and “LA_R” for recombineering

VALIUM_GYPSY_L_F AGGCGCGCCTCGGTACACTAGTTG VALIUM_GYPSY_L_R CGCGGATCCTACTAGTGTTGTTGG

“RA_F” and “RA_R” for recombineering

VALIUM_GYPSY_R_F CGCGGATCCGCGTGCAGTGGCCAC VALIUM_GYPSY_R_R ACCTTAATTAAGAGCTCCTGCAGGTTG

“5′-Check-F” and “5′-Check-R” for recombineering

pacman_L for VGL GTGAGCGCGCGTAATACGACTCAC VALIUM20_22_R TAATCGTGTGTGATGCCTACC for OdsH^mel expression check by RT-PCR

OdsH c0+ GCAAAAGCTAAGACGAAAATGGAT

U8 4524- TGCTTAGCTAACCACCCGAAATCA

rp49F CAGTCGGATCGATATGCTAAGCTGT

rp49R TTACCGACCTTGGGCATCAGATACT

(23)

12

Generation of RNAi strains against OdsH^sim and OdsH^mau in D. melanogaster Two mir-6-1-based constructs, together with the vector Valium22 (NI et al. 2011) (Figure 2), were combined by a three-way ligation with appropriate restriction enzyme digestions, which created pV22_OdsH^simAB (Figure 3). pV22_OdsH^mauAB was generated by the same method (Figure 4). These two resulting constructs were then microinjected into embryos of attP flies (Bloomington Stock 25710) via ΦC31 integrase-mediated transformation by Rainbow Transgenic Flies, Inc. Hatched virgin adults (P₀) were sexed and mated with the parental attP virgin flies, and transformants were screened in F1 flies to create homozygous stocks.

Immunostaining

Immunostaining was performed as described (SINGH and HOU 2008) with minor modifications. Testes were dissected in Drosophila ringer’s solution and fixed in 4%

paraformaldehyde made in phosphate-buffered saline (PBS) for 20 minutes. Fixed samples were washed three times for 20 minutes each with 1 × PBST (0.1% Triton X-100 in 1 × PBS), and blocked in 1 min blocking buffer (Bio-Cando) at 4C overnight.

Samples were then incubated in primary antibody at 4C overnight, washed five times for 30 minutes each with 1 × PBST, and incubated in secondary antibody at room temperature for two hours in the dark. After 5 washes in 1 × PBST and another three washes in 1 × PBS, samples were mounted on slides with a drop of mounting medium (50% glycerol in 1 × PBS) and stored at 4C until imaging.

Primary and secondary antibodies were diluted in the blocking buffer. Primary antibodies used were: 1:50 mouse anti-Fasciclin III (Developmental Studies Hybridoma Bank, DSHB), and 1:10,000 rabbit anti-GFP (Torrey Pines Biolabs). Alexa fluorescence

(24)

13

conjugated secondary IgG (H+L) antibodies were used at 1:200 for goat anti-rabbit 488 and 568 (Molecular Probes).

Imaging and quantification

Fluorescence images of testes were obtained using the Leica TCS SP5 AOBS confocal microscope. Signals were analyzed using MetaMorph (Molecular Devices) software. Image stacks were converted into 3D objects by 4D viewer, and signal

intensity was estimated by summing the number of voxels within the regions of interest.

Light microscope images of compound eyes were obtained using Zeiss SteREO Lumar.V12. Compound eye size was measured using MetaMorph software. Scanning electron microscope images of compound eyes were obtained using the FEI Inspect S in low vacuum mode.

Statistical methods

Statistical significance of the differences in average volume of green fluorescent signals per testis and in average compound eye size and circumference of various genotypes was estimated using the Mann-Whitney U test.

RNA extraction and cDNA reverse transcription

Total RNA was extracted from five three-day-old adult male flies with TRIzol reagent (Invitrogen) following the manufacturer’s instructions. cDNA was synthesized using SuperScript® III First-Strand Synthesis System for RT-PCR (Invitrogen)

according to the manufacturer’s protocol with OdsH^mel-specific primers. Drosophila ribosomal protein 49 gene (rp49) was used as a loading control. PCR conditions

(25)

14

included one cycle of 2 min at 95°C followed by 25 cycles of 95°C for 30 s, 55°C for 90 s and 72°C for 30 s, and a final incubation at 72°C for 10 min. PCR products were separated on 0.8% agarose gels. See Table 2 for primer sequences.

(26)

15

Figure 2. Valium22 vector. Valium22 is a ΦC31-based transformation vector and specialized for expression in the female germline. ftz, the 3-prime UTR intron of fushi tarazu, contains the transport and localization signal. K10, the 3-prime UTR of female sterile (1) K10, contains the localization element. The whole design is flanked by two gypsy insulators to stabilize transgenic expression.

loxP 5xUAS

P-transposase promoter

Valium22

8973 bp

(27)

16

Figure 3. Generation of pV22_OdsH^simAB. Two synthetic miRNA templates and the vector Valium22 were digested with different restriction enzyme pairs: XbaI/BglII, BamHI/NotI, and XbaI/NotI. These three fragments were then combined in a three-way ligation to create pV22_OdsH^simAB for ΦC31 integrase-mediated transformation.

NotI

mir-6-1-OdsH^simA

BamHI BglII

mir-6-1-OdsH^simB

XbaI

NotI XbaI

Valium22

pV22_OdsH

^simAB

9224 bp

loxP 5xUAS

(28)

17

Figure 4. Generation of pV22_OdsH^mauAB. Two synthetic miRNA templates and the vector Valium22 were digested with different restriction enzyme pairs: XbaI/BglII, BamHI/NotI, and XbaI/NotI. These three fragments were then combined in a three-way ligation to create pV22_OdsH^mauAB for ΦC31 integrase-mediated transformation.

NotI

mir-6-1-OdsH^mauA

BamHI BglII

mir-6-1-OdsH^mauB

XbaI

NotI XbaI

Valium22

pV22_OdsH

^mauAB

9224 bp

loxP 5xUAS

(29)

18

Results

Generation of RNAi strains against OdsH^sim and OdsH^mau in Drosophila melanogaster

The mir-6-1-based RNAi constructs, pV22_OdsH^simAB and pV22_OdsH^mauAB, were introduced into D. melanogaster y¹ sc¹ v¹ P{nos-phiC31\int.NLS}X;

P{CaryP}attP2 embryos. For the pV22_OdsH^simAB construct, 63.1% of hatched larvae eclosed, 82.9% of the adults were fertile, and 26.5% of the fertile adults were

transformants. For the pV22_OdsH^mauAB construct, 60% of hatched larvae eclosed, 70.8% of the adults were fertile, and 29.4% of the fertile adults were transformants (Table 3). Two UAS-OdsH^RNAi strains were therefore generated and named

OdsH^RNAi_simAB and OdsH^RNAi_mauAB.

RNAi knockdown rescues the OdsH^mel ectopic expression phenotype in Drosophila eye

A total of six UAS-OdsH^RNAi strains, including two newly generated and four obtained form stock centers, were investigated in the GMR-Gal4>UAS-OdsH^mel background. Observations on Drosophila eye phenotypic changes demonstrate that all the six RNAi strains yield the knockdown effect. Compared to the wild-type eyes (Figure 5, A-A′), GMR-driven OdsH^mel caused severe eye defects: reduction in size, and loss of ommatidia and bristles (Figure 5, B-B′). The two VDRC and two TRiP lines restored the eye defects evidently (Figure 5, C-F′). All of them displayed a

wild-type-like eye of normal size. However, only three of them showed regular

ommatidia and bristles (Figure 5, D-F′), except for the strain GD51289, which produced fused ommatidia and disorganized bristles (Figure 5, C-C′). The newly synthesized

(30)

19

OdsH^RNAi_simAB and OdsH^RNAi_mauAB, whose guide and passenger strands are based on the coding sequences of OdsH^sim and OdsH^mau, showed minor but palpable knockdown effect. In the GMR-Gal4>UAS-OdsH^mel background, OdsH^RNAi_simAB and OdsH^RNAi_mauAB partially restored not only the size of compound eyes but also the growth of ommatidia and bristles (Figure 5, G-G′ and H-H′). Measurements of eye area and circumference provide quantification results in strong agreement with phenotypic observations that the four strains from stock centers had higher efficiency than the two self-made (Figures 6 and 7).

To further confirm the rescue effect, I used RT-PCR to detect the OdsH^melmRNA level of the above-mentioned strains. The result showed that the VDRC and TRiP stocks obviously reduced OdsH^mel expression towards the wild-type (w¹¹¹⁸ and GMR-Gal4) level. It is expected that the decrease of OdsH^melmRNA was ambiguous in

OdsH^RNAi_simAB and OdsH^RNAi_mauAB, whose RNAi constructs were designed according to OdsH^sim and OdsH^mau coding sequences, because OdsH has experienced rapid evolution and diverged at the sequence level (Figure 8).

(31)

20

Table 3. Summary of the introduction of mir-6-1-based RNAi constructs into Drosophila melanogaster via C31 integrase-mediated recombination. *Only a part of the hatched larvae were delivered by Rainbow Transgenic Flies, Inc.

Symbol Genotype Hatching rate

(larvae/injected eggs)

Eclosing rate (P0 adults/larvae)

Fertility rate (fertile P0 adults/P0 adults)

Transformation rate (transformants/fertile P0 adults)

OdsH^RNAi_simAB y¹ sc¹ v¹; P{OdsH^RNAi_simAB}attP2 * 63.1% (41/65) 82.9% (34/41) 26.5% (9/34)

OdsH^RNAi_mauAB y¹ sc¹ v¹; P{OdsH^RNAi_mauAB}attP2 * 60% (24/40) 70.8% (17/24) 29.4% (5/17)

(32)

21

Figure 5. Drosophila eye images from wild-type and transgenic flies. (A-H) Scanning electron microscope images. (A′-H′) Light microscope images. (A-A′) w¹¹¹⁸. (B-B′) GMR>OdsH^mel. (C-C′) GMR>OdsH^mel; GD51289 (VDRC). (D-D′) GMR>OdsH^mel; KK103949 (VDRC). (E-E′) GMR>OdsH^mel; JF02198 (TRiP 1^st Gen.). (F-F′) GMR>OdsH^mel; HMS01554 (TRiP 2^nd Gen.). (G-G′) GMR>OdsH^mel; OdsH^RNAi_simAB. (H-H′) GMR>OdsH^mel; OdsH^RNAi_mauAB. Scale bars: 100 m.

A B C D E F G H

A′ B′ C′ D′ E′ F′ G′ H′

GMR>OdsH^mel GMR>OdsH^mel

w¹¹¹⁸

GD51289 KK103949 JF02198 HMS01554 OdsH^RNAi_simAB OdsH^RNAi_mauAB

(33)

22

Figure 6. Normalized values of eye area to w¹¹¹⁸ control. Genotypes are indicated below the graph. Error bars indicate the standard deviation of the mean. Statistical

significance was determined with the Mann-Whitney U test. *, P < 0.05; ***, P < 0.001;

ns, not significant. n = 10 testes for each data point.

***

ns

* *

***

Normalized value of eye area

(34)

23

Figure 7. Normalized values of eye circumference to w¹¹¹⁸ control. Genotypes are indicated below the graph. Error bars indicate the standard deviation of the mean.

Statistical significance was determined with the Mann-Whitney U test. *, P < 0.05; ***, P < 0.001; ns, not significant. n = 10 testes for each data point.

***

ns

ns * *

* *

Normalized value of eye circumference

(35)

24

Figure 8. RT-PCR detection of OdsH^mel expression in wild-type and transgenic male flies. Genotypes are indicated above the graph. rp49 was used as a loading control.

(36)

25

His-GFP expression in testes of OdsH⁺ and OdsH⁰ flies

To determine which testicular cells can be observed in the his-GFP background, it is necessary to clarify the construction of this transgenic line. his-GFP, synonymous with His2AvD-GFP, encodes a variant H2A.F/Z class histone of D. melanogaster fused with the green fluorescence protein of Aequorea victoria. Its localization to

chromosomes makes it a superb reporter gene to visualize nuclear activity during interphase, mitosis and meiosis (CLARKSON and SAINT 1999; WHITE-COOPER 2004). In post-meiotic spermatids, chromatin undergoes reorganization and condensation;

histones are modified, removed and replaced by protamines (JAYARAMAIAH RAJA and RENKAWITZ-POHL 2005; RATHKE et al. 2007). Therefore, in his-GFP testes, where GFP signal appears indicates an aggregate of chromosome in germ cells, including GSCs, spermatogonia, spermatocytes and early spermatids. The only somatic cells that can be easily observed are the dividing somatic stem cells (SSCs) and the newly divided somatic cyst cells. Unlike GSCs that undergo four rounds of mitosis and two rounds of meiosis, SSCs merely divide once during spermatogenesis, which makes them relatively few in number.

I first examined the His-GFP expression in the testes of OdsH⁺ and OdsH⁰ flies to uncover how OdsH affects the male germ cells in general. Immediately, I noted that His-GFP displayed various expression patterns in OdsH⁺ and OdsH⁰ testes at day 0. The distribution of His-GFP in testes was in the apical region (Figure 9, A and E), somewhat extending (Figure 9, B and F), or further extending towards the basal end (Figure 9, C and G). On the contrary, at day 32, His-GFP distribution was quite similar in OdsH⁺ and OdsH⁰ testes (Figure 9, D and H), from the apical tip towards the distal end.

Quantification assay showed a significantly higher GFP signal in OdsH⁺ testes than in

(37)

26

OdsH⁰ testes at day 0, but not at day 32 (Figure 10).

Sa-GFP expression in testes of OdsH⁺ and OdsH⁰ flies

Since I observed a significant difference of GFP expression between OdsH⁺; his-GFP and OdsH⁰; his-GFP, I investigated the expression pattern of Sa-GFP, a primary spermatocyte marker (CHEN et al. 2005), in testes of different ages by

immunostaining (Figure 11, A-F). Quantification assay showed at days 1, 15 and 55, no significant difference of GFP signal in sa-GFP, OdsH⁺ and sa-GFP, OdsH⁰ testes (Figure 12).

BamP-GFP expression in testes of OdsH⁺ and OdsH⁰ flies

Next, I characterized the BamP-GFP expression pattern in testes of different ages.

Immunostaining shows that, in both OdsH⁺; bamP-GFP and OdsH⁰; bamP-GFP testes, bam promoter-driven GFP expression either resembled the endogenous bam expression, which starts from the 4-cell cyst stage until the entry into premeiotic G2 phase and peaks in 8-cell cysts (Figure 13, A-C and G-I), or displayed somewhat misimpression pattern that spread extensively from the apical region to the distal end (Figure 13, D-F and J-L).

No quantification was done due to inconsistency in expression pattern.

(38)

27

Figure 9. His-GFP expression in testes of 0- and 32-day-old OdsH⁺ and OdsH⁰ flies.

Testes from (A-C) 0- and (D) 32-day-old flies of OdsH⁺; his-GFP and from (E-G) 0- and (H) 32-day-old flies of OdsH⁰; his-GFP were shown. Testes were stained with anti-GFP (green). Scale bars: 100 m.

OdsH0 ; his-GFP

Day 32

OdsH+ ; his-GFP

Day 0

A B C D

E F G H

(39)

28

Figure 10. The average volume of His-GFP signal per testis. The histogram depicts the average volume of His-GFP signal per testis from OdsH⁺; his-GFP and OdsH⁰; his-GFP flies at 0 and 32 days of age. Numbers on the bars indicate the sample size for each genotype. Error bars indicate the standard deviation of the mean. Statistical significance was determined with the Mann-Whitney U test. ***, P < 0.001; ns, not significant.

ns

(14) (15) (4) (3)

*******

(40)

29

Figure 11. Sa-GFP expression in testes of 1-, 15-, and 55-day-old OdsH⁺ and OdsH⁰ flies. Testes from (A) 1- (B) 15- and (C) 55-day-old flies of sa-GFP, OdsH⁺ and from (D) 1- (E) 15- and (F) 55-day-old flies of sa-GFP, OdsH⁰ were shown. Testes were stained with anti-GFP (green) and anti-Fasciclin III (*, hub cells). Scale bars: 100 m.

A B C

D E F

sa-GFP, OdsH+ sa-GFP, OdsH0

Day 15 Day 55

Day 1

*

* * *

* *

(41)

30

Figure 12. The average volume of Sa-GFP signals in spermatocytes. The histogram depicts the average volume of Sa-GFP signal in spermatocytes per testis from sa-GFP, OdsH⁺ and sa-GFP, OdsH⁰ flies at 1, 15 and 55 days of age. Numbers on the bars indicate the sample size for each genotype. Error bars indicate the standard deviation of the mean. Statistical significance was determined with the Mann-Whitney U test. ns, not significant.

ns

(17) (16) (12) (14) (14) (10)

ns

(42)

31

Figure 13. BamP-GFP expression in testes of 1-, 10-, and 32-day-old OdsH⁺ and OdsH⁰ flies. Testes from (AD) 1- (BE) 10- and (CF) 32-day-old flies of OdsH⁺; bamP-GFP and from (GJ) 1- (HK) 10- and (IL) 32-day-old flies of OdsH⁰; bamP-GFP were shown. Testes were stained with anti-GFP (green) and anti-Fasciclin III (*, hub cells). Scale bars: 100 m.

A B C

D E F

G H I

J K L

Day 10 Day 32

Day 1

OdsH+ ; bamP-GFPOdsH0 ; bamP-GFP

*

* *

*

* *

(43)

32

Discussion

OdsH RNAi strains should facilitate future studies on the functional divergence of duplicate genes and the normal function of OdsH^sim

Four OdsH RNAi strains from VDRC and TRiP were tested in GMR>OdsH^mel background. All of them rescued the ectopic expression phenotype to different levels, which indicates their efficiency. It seems unlikely to associate their efficiency with their targeting regions; all of the targeting regions fall within the exon 4 of OdsH, with one (JF02198 from TRiP) including the 3′ UTR (Figure 14 and Table 4). The observed less efficiency of GD51289 from VDRC is probably because the RNAi construct is

generated via the P-element transformation, and its expression is subject to the position effect. Ultimately, these strains should be examined in a fertility assay to see whether their phenotype—males produce less offspring—shows consistency with OdsH null mutant. These strains should facilitate the simultaneous knockdown of OdsH and unc-4 for future studies on the functional divergence between the two paralogs.

For the newly generated OdsH^RNAi_simAB and OdsH^RNAi_mauAB, two target sites for each construct fall within the homeobox (mir-6-1-OdsH^simA and mir-6-1-OdsH^mauA) and the exon 4 (mir-6-1-OdsH^simB and mir-6-1-OdsH^mauB), respectively (Figure 14).

Considering the sequence divergence of OdsH (Table 5), it is expected to see the incomplete rescue effect. Although the phenotypic rescue effect were observed, these two strains need to be further tested in the GMR>OdsH^sim and GMR>OdsH^mau

background to determine the knockdown efficiency.

(44)

33

Figure 14. OdsH structure and the target regions of UAS-OdsH^RNAi strains used in this thesis. Self-made RNAi constructs

(mir-6-1-OdsH^simA, mir-6-1-OdsH^simB, mir-6-1-OdsH^mauA and mir-6-1-OdsH^mauB) are denoted as simA, simB, mauA and mauB. Note that target regions of GD51289, HMS01554 and simB do not overlap.

OdsH

HMS01554

simA mauB simB

mauA

KK103949 GD51289

JF02198

Intron

Homeobox UTR

Coding sequence Exon 3

Exon 1 Exon 2 Exon 4

(45)

34

Table 4. The target sequences of UAS-OdsH^RNAi strains obtained from stock centers. OdsH 3′ UTR is underlined.

Strain Length (bp) Sequence (5′ to 3′)

GD51289 328 CGCAGCAAGCGGATGAAGAAAGCCATCGATCGGCAGGCGAAGAAGCTACAGGACAAGGGATTGGAAGTGGACTATGCCCGTCTGGAGGCCGAGTACCTAG CTGCCCACCAGGAGAACGGAGTGGATGAGAATAACTGGCTGGATGATGATGGCTACGATGATCTGCACATCGATGTGGTGGGCGTTGAACCGGAGTACGT GACCGGCGACAGTTTGGATCACTCGTTCTGCTCCTCCAGGACGTACCAGACGAAGAGCACCAGCAGCGAACTGGATTCCAATGATATGGGACTCCAGGGA AGAGTGGAAACTCCGCCACCACCACAGC

KK103949 331 GACAAGGGATTGGAAGTGGACTATGCCCGTCTGGAGGCCGAGTACCTAGCTGCCCACCAGGAGAACGGAGTGGATGAGAATAACTGGCTGGATGATGATG GCTACGATGATCTGCACATCGATGTGGTGGGCGTTGAACCGGAGTACGTGACCGGCGACAGTTTGGATCACTCGTTCTGCTCCTCCAGGACGTACCAGAC GAAGAGCACCAGCAGCGAACTGGATTCCAATGATATGGGACTCCAGGGAAGAGTGGAAACTCCGCCACCACCACAGCCACCGATGCAGAATAAGACCCTC TACAATTCGCCCTTCAGCATCGAATCCCTGT

JF02198 405 AGCTACAGGACAAGGGATTGGAAGTGGACTATGCCCGTCTGGAGGCCGAGTACCTAGCTGCCCACCAGGAGAACGGAGTGGATGAGAATAACTGGCTGGA TGATGATGGCTACGATGATCTGCACATCGATGTGGTGGGCGTTGAACCGGAGTACGTGACCGGCGACAGTTTGGATCACTCGTTCTGCTCCTCCAGGACG TACCAGACGAAGAGCACCAGCAGCGAACTGGATTCCAATGATATGGGACTCCAGGGAAGAGTGGAAACTCCGCCACCACCACAGCCACCGATGCAGAATA AGACCCTCTACAATTCGCCCTTCAGCATCGAATCCCTGTTGGGATCGTAACGAATATCGAATGTTGTACTATAATTCTACTATGATTTCGGGTGGTTAGC TAAGC

HMS01554 21 CAGCAGCGAACTGGATTCCAA

Table 5. Homology of the target sequences of mir-6-1-based RNAi constructs in Drosophila melanogaster, D. simulans and D. mauritiana.

The designed sequences are shown in bold letters. Mismatched nucleotides are indicated in red. Gaps are denoted as “-.” Percentages in parentheses refer to the level of sequence homology in comparison with the designed sequence (number of identical matches divided by the sequence length).

mir-6-1-OdsH^mauA mir-6-1-OdsH^simA mir-6-1-OdsH^mauB mir-6-1-OdsH^simB

D. melanogaster GGAGAGAGTCTTCCAGGGAAGT(86.4%) ATAGCGGTTTGGTTCCAGAATC(81.8%) ATATGGGACTCCAGG--GAAGA(63.6%) AATTCGCCCTTCAGCATCGAAT(86.4%) D. simulans GGAGAGAGCCTTCCAGGACAAT(90.9%) ATAGCGGTTTGGTGTCAAAATA ATATGGAAGTCCAGG--GAATA(77.3%) AAATCGTCCTTCAGCATCAAAT

D. mauritiana GGAGAAAGCCTTCCAGGAAAAT ATAAAGATATGGTTCCAAAATA(72.7%) ATATGGAAGTGGAAGTGGAATA AAATCGTCCTTCAGCATCAAAT (100%)

(46)

35

His-GFP expression implies GSC loss or disruption of subsequent spermatogenic development in young OdsH⁰ flies

Spermatogenesis in D. melanogaster is a series of cellular divisions and

metamorphosis events. At the apex of a testis tub lies a group of somatic support cells termed hub cells, which are surrounded by 6-12 germline stem cells (GSCs) (SPRADLING

et al. 2011). Each GSC is flanked by a pair of cyst progenitor cells (i.e. somatic stem cells, SSCs); both GSCs and SSCs are connected to the hub cells. As a GSC divides asymmetrically into two daughter cells, one maintains the GSC identity, and the other becomes a spermatogonium encapsulated by two somatic cyst cells. While these two cyst cells undergo no further division, the spermatogonium enters four rounds of synchronous mitotic division with incomplete cytokinesis, producing a cyst of 16 interconnected spermatogonia. The spermatogonia then embark on premeiotic DNA replication, and switch to an extended G2 phase for cell growth as spermatocytes (LIM et al. 2012). During the G2 phase, each spermatocyte substantially increases 25 times in volume. This premeiotic G2 phase continues for more than 3 days before two meiotic divisions proceed, which generates 64 round spermatids (WHITE-COOPER 2010). Still interconnected by a cytoplasmic bridge (ring canal), these round spermatids

subsequently enter the elongation program and become individual, mature sperms (Figure 15).

At day 0, His-GFP expression patterns were various in OdsH⁺ and OdsH⁰ testes.

This is probably due to the individual variation. Because the development of testes starts in the late larval stage (CHENG et al. 2008), it is likely that flies collected on the same day differ in sexual maturation. The quantification difference of GFP signal indicates there are more cells (predominantly germ cells) in OdsH⁺ testes. One explanation

(47)

36

(“explanation A”) is that OdsH⁰ flies fail to maintain as many GSCs as OdsH⁺ flies do, and GSC loss in OdsH⁰ testes causes a decrease in germ cell number. An alternative (“explanation B”) is that the GSCs are intact, but the subsequent development in OdsH⁰ testes is disrupted. However, at day 32, the quantification of GFP shows no difference between OdsH⁺ and OdsH⁰ testes. How does the age factor coordinate with these two explanations? One hypothesis is that the age effect on GSC maintenance, or the subsequent developmental stages, is accumulative and relatively minor than OdsH⁰ in young flies. In aged flies, OdsH⁰ effect is masked by the increasing age effect, so the germ cell number decreases in OdsH⁺ and OdsH⁰ testes without significant difference.

Both explanations are contradictory to the previous discovery that the GSC number is significantly less in 10-day-old, but not in 3-day-old, OdsH⁰ testes (CHENG et al. 2012).

(48)

37

Spermatogonium Spermatocyte Spermatid

Figure 15. Male germline differentiation in Drosophila melanogaster. The schema is not in scale. Expression periods of three GFP reporter lines used in this study, His2AvD-GFP, sa-GFP and bamP-GFP, are indicated. Note that the various spermatid “stages” are defined by the nuclear shape, not mitochondrial.

GSC Sperm

His2AvD-GFP

bamP-GFP

sa-GFP