果蠅族群中染色體逆位多態性和隱性不良因子累積之相關性

國立臺灣大學生命科學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

果蠅族群中染色體逆位多態性和隱性不良因子累積之相關性 Association between Chromosomal Inversion Polymorphisms and

Accumulation of Recessive Deleterious Mutations in Drosophila melanogaster

董詩凡

中華民國 101 年 7 月 July, 2012

指導教授： 丁照棣 方淑

Advisors: Chau-Ti Ting

Shu Fang

2

ii

誌謝

終於，我生命的旅途也來到了這一站。我從來就不是個聰明的人，若現在我 有那麼一點點小小的收穫成就，都要感謝身邊無數人的幫忙和扶持，造就我現在 的模樣。加入此實驗團隊滿三年，一路走來，我很幸運的認識所有學識豐富的老 師，相濡以沫的夥伴，他們都豐富了我的生命。

首先要感謝方淑老師和丁照棣老師。大學剛畢業的我，懷著熱情但卻對做研 究的真實情況不甚了解，誤打誤撞的在兩位老師的指導下進入演化的世界。從知 識的累積，到做實驗的方法方向，還有科學研究的態度，兩位老師都不吝惜的傾 囊相授，並引導著我。在我對生涯規劃徬徨的時候，對人生疲累抱怨的時候，老 師們都給我很大的協助和鼓勵。更重要的是兩位老師給我很多機會磨練砥礪自我 的能力極限和潛能。雖然有時候會懷疑自己做的成果到底能夠述說什麼，不過在 老師提供給我的機會之下，藉由即使是一點點正向的回饋和刺激，都讓我有所思 索成長。非常感謝方老師和丁老師的慷慨以及開闊的心胸，讓我這三年來在學習 的路上能夠羽翼漸豐，充實滿滿。

再來要感謝張慧羽老師，我總是私底下開玩笑稱其為「恩師們的恩師」。張老

師清楚的邏輯和口條，循序漸進的導引，將演化的抽象概念解釋得非常清楚，每 每結束張老師的課，都有如沐春風般的清爽，學術上的知識也得以充分的消化理 解。此外，張老師對於我的研究也不遺餘力的提供她的意見，指出我的錯誤和觀 念上要加強的地方。雖說被她問問題的時候總是感到很緊張，但也因為如此，我 能夠有更多層面的去思索整個研究的架構和邏輯。

此外要感謝曹順成老師，雖說剛認識曹老師的時候，常常分不太出來他開玩 笑和認真的話之間的差別，但相處久了之後就會發現他的風趣和幽默。感謝曹老 師幫我修改潤飾每次的英文稿件和海報，在老師的指導下也逐漸了解如何以英文 更精簡的表達。

感謝于宏燦老師和蔡玉真老師撥冗擔任我的口試委員，花時間評改我的論文，

iii

參與我的口試，並給予很多寶貴的意見。能夠得到兩位老師的建議和反思，對我 之後論文修改的方向有很大的幫助。再次謝謝兩位老師。

謝謝實驗室兩位身兼數職的人妻，忙碌的台灣果蠅遺傳資源中心助理詹文喬 和劉心彥。始終讓人家猜不出真實年紀的兩位女士，於公於私都幫了我相當大的 忙。我的實驗所需要的果蠅進口和訂購、實驗室相關事務的協助、心疼我所以自 願幫我分擔很多我應該要自己做的工作，也提供自身的經驗供我參考；特別是心 彥，不只照顧我的身心靈還有胃，這些點點滴滴我都銘記在心，同時很感激能得 到這樣的情誼。非常謝謝你們。

謝謝莞儒和昕儀，加入果蠅實驗室的第一年，在人生地不熟的台北，因為有 你們兩位的相伴，讓我逐漸適應這裡的生活，在中研院的工作也充滿笑聲和歡樂。

謝謝昕儀完成了這個實驗的前導工作，讓我後來得以進行並成為我的碩士論文。

謝謝其駿提供的腦力激盪和反思，在學術和生活中永遠都讓我有新的想法和 獲得。每當遇到困難時，總是想尋求你清晰的頭腦的協助。非常感謝。

謝謝實驗室一起留到很晚，一起討論作業，一起煩惱未來，一起同舟共濟的 成員們：家宣、家翔、佳蓉、宇謙、Ani、瑾瑜、映璇、佳豪、景賀、韋進、芷筠、

維中、淑媛。感謝這兩年的生活有你們大家，讓這實驗室充滿了歡笑和活力。常 常想著要是研究所生活沒有你們，那該會多麼無趣沉悶！這期間多多少少短暫的 別離，卻只加深更多的想念。原本只要一想到畢業就非常捨不得，但現在也逐漸 釋懷，因為我知道這些日子都會成為往後我珍貴的回憶。

最重要要感謝我親愛的家人：時刻刻想念女兒的媽媽，提供我經濟援助的爸 爸，三不五時很怕我畢不了業的弟弟。研究所這兩年回家的次數屈指可數，也非 常難向家人解釋到底每天每天都去實驗室是為了什麼，家裡的狀況也不能幫忙負 擔，知道我會操心也總是報喜不報憂。現在總算是一個階段的完成，你們的全力 支持讓我非常非常感謝。

最後要謝謝曾經愛我和我愛的人，因為有過去的你才有現在的我。只要全心

iv

付出過，就算最後分離了也沒有遺憾。

希望出現在我人生旅途上的所有人，都幸福。

v

摘要

在自然族群中，染色體逆位多態性被認為和物種的適應有高度關係。目前認 為族群維持染色體逆位多態性的機制，是透過其攜帶具選汰優勢的一組基因，因 逆位異型合子抑制遺傳重組，而被同時保留在族群裡造成共適應的現象。然而，

對於染色體逆位異型合子抑制重組造成的不良因子累積並增加突變負荷很少受到 討論。不良因子的累積會隨著逆位同結合型在族群中被消除，進而增加逆位異型 合子在族群裡的頻率。黃果蠅的熱帶非洲族群，同時具有高度的染色體逆位多態 性和高頻的隱性不良因子累積，此現象提供了一個相當適合的研究材料。藉由定 位隱性不良因子在染色體上累積的位置是否會真如預期在重組被高度抑制的位置，

例如染色體逆位的端點，來探討染色體逆位和不良因子之間的關係，我們利用重 組定位和基因缺失組定位出隱性不良因子累積的位置，結果在分別來自八個不同 單雌系的八條三號染色體上定位出十四個不同的隱性致死因子，並且所有的致死 因子和染色體逆位端點相距非常近，只有不到三個厘摩根。同時，隱性致死因子 分佈在染色體上的位置和族群內的逆位多態性相當一致。本研究顯示隱性不良因 子的累積受到染色體逆位多態性所抑制的重組率影響，在染色體上呈現不隨機的 分佈。同時，也暗示染色體逆位和隱性不良因子之間強烈的相關性。

關鍵字：遺傳負荷、染色體逆位端點、穆勒氏棘輪、隱性致死因子、遺傳重組抑 制。

vi

Abstract

Chromosomal inversion polymorphisms have been demonstrated to play adaptive roles in natural populations by capturing local co-adapted alleles within

recombination-suppressed regions of inversion heterozygotes. On the other hand, a less studied but important role of inversion polymorphisms is that recombination suppression by heterozygous chromosomal rearrangement may accumulate recessive deleterious mutations and thus cause a great amount of mutation load. Deleterious mutations will be eliminated when homozygotes and in turn increase heterozygosity of various inversions. The Afrotropical population of Drosophila melanogaster with high chromosomal inversion heterozygosity and high ratio of recessive lethals provides an ideal material to test any association between them by examining the accumulation pattern of recessive lethals. Recessive lethals are predicted to locate near the inversion breakpoints where the recombination is greatly suppressed by inversion

heterokaryotypes. By using recombination and deficiency mappings, 14 recessive lethal alleles from eight lethal-bearing third chromosomes (each from eight distinct isofemale lines, respectively) were identified. All of recessive lethals were mapped into the regions close to (less than 3 centi-Morgan) the inversion breakpoints which were polymorphic in the African population. This result clearly shows that recessive lethal alleles are accumulated by recombination suppression and distributed

vii

non-randomly along inverted chromosomes. The data also provide the strong association between chromosomal inversions and the accumulation of recessive deleterious mutations.

Key words: genetic load, inversion breakpoint, Muller’s ratchet, recessive lethal, recombination-suppression

viii

Contents

口試委員審定書 ……….…….…i

致謝 ………..……….………..ii

摘要 ……….……….……...v

Abstract ………...……….vi

List of Figures ………...…...ix

List of Tables ………...………....xi

Introduction ……….…1

Materials and Methods ……….…6

Fly strains ……….6

Recessive lethal alleles mapping ………...…...7

Estimation of genetic distance of recessive lethals and breakpoints of inversions ...9

Statistical test ………..10

Results ……….………11

Discussion ……….………..33

High recombination suppression due to inversion heterozygotes ………..33

Accumulation pattern of recessive lethals ………..35

Chromosomal inversion polymorphisms vs. genetic load ………..38

References ………..40

Appendices ………...46

ix

List of Figures

Figure 1 The distribution of eight markers on the third chromosome of rucuca

strain …………..………..………..8

Figure 2 The cross scheme of recombination mapping ………...…………..8

Figure 3 The cross scheme of deficiency mapping ………9

Figure 4 The distribution of recessive lethal number on eight 3^rd chromosomes ...….13

Figure 5 Mapping results of the recessive lethal allele l^ZS2 ………..16

Figure 6 Mapping results of the recessive lethal allele l^ZS30-I ………...16

Figure 7 Mapping results of the recessive lethal allele l^ZS30-II ………..17

Figure 8 Mapping results of the recessive lethal allele l^ZS30-III ……….…18

Figure 9 Mapping results of the recessive lethal allele l^ZS30-IV ……….……19

Figure 10 Mapping results of the recessive lethal allele l^ZS53 ………..19

Figure 11 Mapping results of the recessive lethal allele l^ZH12 ……….….20

Figure 12 Mapping results of the recessive lethal allele l^ZH18-I ……….…...20

Figure 13 Mapping results of the recessive lethal allele l^ZH18-II ……….…..21

Figure 14 Mapping results of the recessive lethal allele l^ZH21-I ……….…...22

Figure 15 Mapping results of the recessive lethal allele l^ZH21-II ……….……..23

Figure 16 Mapping results of the recessive lethal allele l^ZH32 ……….…….23

x

Figure 17 Mapping results of the recessive lethal allele l^MW6-I……….……24 Figure 18 Mapping results of the recessive lethal allele l^MW6-II ……….……..25 Figure 19 The estimation of genetic positions of recessive lethals and inversion

breakpoints by interpolation of nearest published genetic markers ………...…....27 Figure 20 The distribution of recessive lethal alleles and inversions on the 3^rd

chromosome at three different levels ……….….30 Figure 21 Comparison between null distribution of genetic distance and true mean distances ………..32

xi

List of Tables

Table 1 Recombination mapping result using recombinants between the rucuca multimarker chromosome and the lethal-bearing chromosome ………..12 Table 2 Summary of 14 recessive lethal alleles ………….………...14 Table 3 Genetic distances (cM) between each recessive lethals and each nearest inversion breakpoint ………..…….…..28 Table 4 Genetic distances (GD) of recessive lethal alleles from the nearest

breakpoint of chromosomal inversions at different levels ………...…….…...29

1

Introduction

Chromosomal inversions have been recognized as an aberration in which the order of a chromosomal segment is reversed. Sturtevant (1921) showed and later Caron (1946) proved the mechanism that inversions suppress recombination by eliminating abnormal crossover products of meiosis (dicentric and acentric chromosomes) (Sturtevant 1921; Caron 1946). Many hypotheses were proposed to explain the evolutionary significance of chromosomal inversion polymorphisms in nature, and six main explanations for the spread of inversions in population (reviewed in Hoffmann and Rieseberg 2008). The most accepted explanation is the reduction of recombination by inverted heterozygotes, causing a set of locally adapted genes captured by inverted regions. Combinations of these alleles at loci within inversions are assumed to be co-adapted by having epistatic effects on fitness, and favored by natural selection toward heterozygotes (Dobzhansky 1937, 1970); or even without epistatic effects, only by migration-selection balance toward locally favored alleles in populations

(Kirkpatrick and Barton 2006). Previous studies, especially in Drosophila

pseudoobscura, showed that genetic frequencies of distinct chromosomal

rearrangements are involved in seasonal cycles and altitudes changes (Dobzhansky 1947). Schaeffer et al. (2003) proved the genomic evidence for maintaining positive epistatic relationships among loci within inversions of D. pseudoobscura. In other

2

Drosophila species, many traits, including viability, longevity, resistance to thermal

extremes, mating success and fecundity, have been demonstrated to be linked to

inversion polymorphisms (e.g., Barnes 1983; reviewed in Hoffmann et al. 2004). In D.

melanogaster, molecular markers linked to inversion In(3R)Payne are highly associated

with body size and heat resistance (Weeks et al. 2002; Anderson et al. 2003). Recent studies of other species also have revealed that selective advantages of specific locally adapted alleles are highly associated with various chromosomal rearrangements among different local environments, e.g., butterfly Heliconius numata (Joron et al. 2011), malaria mosquito Anopheles (Ayala et al. 2011; Rottschaefer et al. 2011), and yellow monkey flower Mimulus guttatus (Lowry and Wills 2010). Those studies strongly support that chromosomal inversion polymorphisms are maintained by adaptively selective advantages in local populations. In addition, the recombination suppression effect by inversion heterozygotes might cause genetic barrier within and/or between populations. Hence, chromosomal inversion polymorphisms have been suggested as a driving force on population divergence and even speciation process (reviewed in

Krimbas and Powell 1992; King 1993; Noor et al. 2001; reviewed in Rieseberg 2001;

Navarro and Barton 2003; Hoffmann et al. 2004). Stevison et al. (2011) tested the gene flow along the large inverted region of two closely related Drosophila species hybrids, and found that the nucleotide divergence differs in low recombination rate

3

regions, suggesting that inversion polymorphisms might drive the population

divergence.

Another mechanism, though less studied, was first described in Drosophila

pseudoobscura populations is that the genetic load might contribute to inversion

polymorphisms (Epling et al. 1961; Dobzhansky et al. 1963; Mayhew et al. 1966).

The recombination-suppressed region tends to have a higher probability to accumulate deleterious mutations, which could hide in inversion heterozygotes from elimination by natural selection. Some population surveys in Drosophila subobscura showed that a higher level of genetic load occurs in the population having a higher frequency of chromosomal rearrangements (Andjelković et al. 1998; Zivanović et al. 2000), while other studies revealed the opposite relationship that the frequency of chromosomal rearrangements in populations had a negative correlation with levels of genetic load (Watanabe 1969; Watanabe and Yamazaki 1976). On the other hand, some studies showed the non-random distribution of low frequency of recessive deleterious

mutations on inverted chromosomes in different Drosophila species (Crumpacker and Salceda 1969; Mestres et al. 1990; Yang et al. 2002). Although without direct

evidence, these studies suggested that the genetic load might contribute to the complex mechanism affecting the dynamics of genetic polymorphism in natural populations. In this study, I want to understand the distribution pattern on the chromosome of genetic

4

load and how genetic load contributes to the polymorphism of inversion populations.

High chromosomal inversion polymorphisms (Dobzhansky 1937) with

accumulation of recessive deleterious mutations (Sturtevant 1937; Dubinin 1946) were reported in Drosophila, including Afrotropical (including southern and eastern Africa) populations of D. melanogaster (Aulard et al. 2002). A recent study (S. Fang

unpublished data; also see Appendix A) on the inversion polymorphism of Afrotropical populations found that: (1) high inversion polymorphisms especially on right arms of 3^rd chromosomes; (2) high inversion heterozygosity; (3) high ratio of recessive

lethal-bearing chromosomes. These observations provide us a good material to study the genetic load mechanism and to test whether there is any association between accumulation deleterious mutations and inversion chromosomal inversion polymorphisms. If there is no association between chromosomal structural polymorphisms and genetic loads, mutations might distribute randomly on

chromosomes; otherwise mutations might distribute in a given pattern. From the feature of chromosomal inversions which reducing crossover events, it was predicted that recessive deleterious mutations on the third chromosome were to accumulate non-randomly on highly recombination-suppressed regions, and in turns, to increase inversion heterozygosity. With the advent of fine mapping tools, such as deficiency mapping in Drosophila, it is possible to answer this long-lasting unsolved question. In

5

this study, I used recombination and deficiency mappings to precisely map recessive lethal alleles to examine the relationship between the recombination-suppression by complex chromosomal inversions and the accumulation of recessive deleterious mutations in the Afrotropical population of D. melanogaster.

6

Materials and Methods

Fly strains

Drosophila melanogaster isofemale lines were collected and established from

Harare and Sengwa, Zimbabwe (referred hereafter as ZH and ZS, respectively) in 1990 (Wu et al.,1995) and from Malawi (MW) by J.W.O. Ballard (Iowa University) in 2002.

Third chromosomes from each isofemale line are extracted (Appendix A, see Figure A3) and balanced with the balancer chromosome, TM3, to keep recessive lethals and recessive sterile alleles in stocks. The multimarker line rucuca (stock number: 576) was obtained from Bloomington Drosophila Stock Center. This line contains eight visible markers on the third chromosome: roughoid (ru), hairy (h), thread (th), scarlet (st), curled (cu), stripe (sr), ebony (e), and claret (ca). The cytological and physical positions of these markers were described by Lindsey and Zimm (1992) and

summarized in Figure 1. Another similar multimarker line ruPrca (stock number 1711) also obtained from Bloomington Drosophila Stock Center, contains a marker

Prickly (Pr) addition to above eight. All the deficiency strains were obtained from

Drosophila Genomics Resource Center, Bloomington (http://dgrc.cgb.indiana.edu/) or Drosophila Genetic Resource Center, Kyoto

(http://kyotofly.kit.jp/cgi-bin/stocks/index.cgi) (Appendix B, Table B1). Those stocks contain different genomic deletions with accurate coordinates and span the entire third chromosome (details in FlyBase http://flybase.org/). Flies were reared on standard

7

cornmeal medium at 22ºC under a 12h-12h light-dark cycle.

Recessive lethal alleles mapping

To map recessive lethal alleles, eight third lethal-bearing chromosomes (four Straight and four In(3R)K) were chosen: ZS2-3, ZS30-2, ZH12-6, ZH32-1, ZS53-3, ZH18-6, ZH21-1, and MW6-3 (four former are straight, later four are In(3R)K). Each lethal-bearing chromosome was examined allelism by complementation test (Appendix A, Figure A4). Distinct recessive lethals from eight third lethal-bearing chromosomes were then mapped by recombination and deficiency mappings. For recombination mapping, the multimarker line rucuca was used to construct the recombinants to narrow down the regions of recessive lethal alleles (Figure 2). In the cross scheme of

recombination mapping, various genotypic recombinants (single male, for lack of recombination) at generation G2 were collected and then backcrossed with the original lethal-bearing line (Figure 2). By scoring the G₃ offspring, we could exclude the lethal-free chromosomal regions (Appendix B, Table B2).

For further deficiency mapping (Figure 3), all deficiency lines which cover the candidate region were chosen. By scoring F1 offspring numbers to examine whether or not recessive lethal alleles fail to complement deletions, the location of recessive lethal alleles can be mapped to the precise genomic regions of deletions which might only cover several genes (Appendix B, Table B3). Thus, recessive lethals can be

8

mapped to the gene level by deficiency mapping.

Figure 1 The distribution of eight markers on the third chromosome of rucuca strain.

Figure 2 The cross scheme of recombination mapping. G

0, the rucuca strain cross with lethal-bearing chromosome lines. G1, rucuca/lethal-bearing heterozygotes cross with ruPrca. Distinct recombinants male can be obtained from the G₁

9

rucuca/lethal-bearing female, then back cross with original lethal-bearing chromosome.

By examining whether or not recombinants fail to complement lethal alleles (G₃), the rough regions of recessive lethals could be mapped. Red cross: recessive lethal allele.

Figure 3 The cross scheme of deficiency mapping. These deletions with precise

coordinates (empty space on the black bar) were utilized to narrow down the location of recessive lethal alleles to the gene level. Red cross: recessive lethal allele.

Estimation of genetic distance of recessive lethals and breakpoints of inversions

The genetic positions of recessive lethals and breakpoints of inversions were marked by genetic distance (cM) and were estimated by interpolation method as

where M1 and M2 stand for genomic positions of the two nearest genetic markers, and R1 and R2 represent genetic positions of M1 and M2, respectively. X denotes the unknown genetic position of recessive lethals or inversion breakpoints. RX is the genomic position obtaining from experimental data. All the genomic position

10

information can be obtained from FlyBase.

Statistical test

With the help of Cheng-Ruei Lee, a R programming (Ihaka and Gentleman 1996) was used to test whether there is statistical significant in the distribution of recessive lethal alleles. The average true value of each distance between recessive lethals and nearest inversion breakpoints can be obtained by experimental results. I cut the 3^rd chromosome (102 cM) into 1020 bins, each bins is 0.1 cM long. I then randomly sampled 14 locations on 3^rd chromosome without replacement (1000 times), and also calculated the mean distance of each sample to nearest inversion breakpoints. I

obtained the normal distribution from null hypothesis (that is, random distribution), with which the true average distance from experimental data can be compared.

11

Results

To map the recessive lethals on the 3^rd chromosome of D. melanogaster, recombination mapping with rucuca multimarker line was applied to constrcut recombinants between the rucuca chromosome and the lethal-bearing chromosome.

By scoring G3 offspring produced by G2 recombinants crossed with the parental

lethal-bearing line (Appendix B, Table B3), the rough regions of recessive lethal alleles can be defined (Table 1). Because such mapping analyses can determine whether or not there is any recessive lethal but cannot determine how many recessive lethals

located at the specific region, these regions with lethal alleles may harbor more than one lethals.

12

Table 1 Recombination mappirng result using recombinants between the rucuca multimarker chromosome and the lethal-bearing chromosome G2 recombinant types G3 offspring lethality of each lethal-bearing chromosome

ZS30-2 ZS2-3 ZH32-1 ZH12-6 ZH18-6 ZH21-1 ZS53-3 MW6-3

ru

+ + + + + + + L L L L L L L L

ru h

+ + + + + + L L L L L L L L

ru h th st

+ + + + - L L L L L L L

ru h th st cu +

+ + A L L L - - - -

ru h th st cu sr

+ + A L L L - - - -

ru h th st cu sr e

+ A A A A - - - -

+ + + + + + + ca L L L L - L - -

+ + + + + +

e ca

L L L A - - - -

+ + + + +

sr e ca

L L L A - - - -

+ + + + cu sr

e ca

L A A A A A A A

+ +

th st cu sr e ca

L A A A A A A A

+

h th st cu sr e ca

L A A A A A A A

Rough regions of recessive lethals

ru-sr cu-ca cu-ca sr-ca st-ca st-ca st-ca st-ca

“L” denotes the lethal-bearing recombinant. “A” denotes the lethal-free recombinant. “-“denotes that the recombinant type was not obtained due to recombination suppression.

13

For fine mapping, specific deficiency lines were chosen (Figures 5-18). Those deficiency lines contain a series of genomic deletions which cover previous predicted regions of recessive lethals (Table 1). By scoring on F₁ offspring, whether or not genomic deletions are complement to recessive lethal alleles can be determined (Appendix B, Table B3). The result showed that 10/14 recesssive lethals are located on the right arm of the 3^rd chromosome, and the number of recessive lethal alleles on each eight chromosome ranging from one to four (Figure 4). Those recessive lethal alleles caused the flies died at different developmental stages, from the embryonic to pupal stages (Table 2). The precise genomic regions of recessive lethal alleles can be confirmed by intersection of overlapping smaller fragment of deletions which are both lack of complementation. In total, 14 recessive lethal alleles were mapped to the gene level on the 3^rd chromosome (Figures 5-18 and Table 2).

Figure 4 The distribution of recessive lethal number on eight 3

^rd chromosomes. X axis represents the number of recessive lethals of each distinct 3^rd chromosome. Y axis represents the number of 3^rd chromosomes bearing a specific number of lethals.

14

Table 2 Summary of 14 recessive lethal alleles

Lethal

allele

Chromosome rearrangements in isofemale lines

Cytological

position Genomic ranges Genetic position (cM)

Covering

genes Lethal stage

l

^ZH18-I

In(3R)K, In(3R)C

85D19-23 3R:5,338,742..5,380,704 48.5 11 No visible larval lethality

l

^ZH18-II

^96A23-25 3R:20,639,446..20,663,321 85.2 10 2

^nd

-3

^rd

instar larval lethality

l

^ZH21-I

In(3R)K, In(3R)P

76C5-6 3L:19,784,186..19,823,829 47.0 14 Eclosion failure

l

^ZH21-II

^82F6-7 3R:1,057,000..1,090,181 47.2 9 No visible larval lethality

l

^ZS53-I In(3R)K, In(3R)93;96

88E2-3 3R:10,983,368..11,075,682 56.5 12 1

^st

-2

^nd

instar larval lethality

l

^MW6-I

In(3R)K

73E5-F2 3L:17,025,406..17,142,023 45.6 9 No visible larval lethality

l

^MW6-II

^82B1-2 3R:254,982..279,012 47.1 5 1

^st

-2

^nd

instar larval lethality

l

^ZS30-I

In(3R)K, ST

61E1-2 3L:959651..1035182 0.0 6 Eclosion failure

l

^ZS30-II

78E1 3L:21597878..21631024 47.0 8 No visible larval lethality

l

^ZS30-III

^84B4-E11 3R:2916249..3919805 47.6 ~110 No visible larval lethality

l

^ZS30-IV

^87B11-C2 3R:8269738..8303300 51.6 9 No visible larval lethality

l

^ZS2 In(3R)K, ST

95E1 3R:19930781..19967091 84.5 10 2

^nd

-3

^rd

instar larval lethality

l

^ZH12 In(3R)K, ST

92B3-C1 3R:15662595..15716378 67.6 8 No visible larval lethality

l

^ZH32 In(3R)K, ST

95C13-14 3R:19747854..19768726 81.7 11 2

^nd

-3

^rd

instar larval lethality

15

Most of recessive lethal alleles can be mapped in the small genomic region containing less than dozens of genes, except the lethal allele

l

^ZS30-III mapped in the large

region covering more than one hundred genes (Figure 8 and Table 2). Interestingly, lethal allele

l

^ZS30-III showed lethal phenotype when mapped with a deficiency line

covering a large genomic region, but showed normal wild-type phenotypes when mapped with several lines covering a part of the large genomic region (Figure 8), suggesting that it was a synthetic lethal allele.

16

Figure 5 Mapping results of the recessive lethal allele l

^ZS2. Red line, the deletion region failed to complement the recessive lethal. Green line, the deletion region was complementary to the recessive lethal.

Figure 6 Mapping results of the recessive lethal allele l

^ZS30-I. Red line, the deletion region failed to complement the recessive lethal. Green line, the deletion region was complementary to the recessive lethal.

17

Figure 7 Mapping results of the recessive lethal allele l

^ZS30-II. Red line, the deletion region failed to complement the recessive lethal. Green line, the deletion region was complementary to the recessive lethal.

18

Figure 8 Mapping results of the recessive lethal allele l

^ZS30-III. Red line, the deletion region failed to complement the recessive lethal. Green line, the deletion region was complementary to the recessive lethal.

19

Figure 9 Mapping results of the recessive lethal allele l

^ZS30-IV. Red line, the deletion region failed to complement the recessive lethal. Green line, the deletion region was complementary to the recessive lethal.

Figure 10 Mapping results of the recessive lethal allele l

^ZS53. Red line, the deletion region failed to complement the recessive lethal. Green line, the deletion region was

20

complementary to the recessive lethal.

Figure 11 Mapping results of the recessive lethal allele l

^ZH12. Red line, the deletion region failed to complement the recessive lethal. Green line, the deletion region was complementary to the recessive lethal.

Figure 12 Mapping results of the recessive lethal allele l

^ZH18-I. Red line, the deletion region failed to complement the recessive lethal.

21

Figure 13 Mapping results of the recessive lethal allele l

^ZH18-II. Red line, the deletion region failed to complement the recessive lethal. Green line, the deletion region was complementary to the recessive lethal.

22

Figure 14 Mapping results of the recessive lethal allele l

^ZH21-I. Red line, the deletion region failed to complement the recessive lethal. Green line, the deletion region was complementary to the recessive lethal.

23

Figure 15 Mapping results of the recessive lethal allele l

^ZH21-II. Red line, the deletion region failed to complement the recessive lethal. Green line, the deletion region was complementary to the recessive lethal.

Figure 16 Mapping results of the recessive lethal allele l

^ZH32. Red line, the deletion region failed to complement the recessive lethal. Green line, the deletion region was complementary to the recessive lethal.

24

Figure 17 Mapping results of the recessive lethal allele l

^MW6-I. Red line, the deletion region failed to complement the recessive lethal. Green line, the deletion region was complementary to the recessive lethal.

25

Figure 18 Mapping results of the recessive lethal allele l

^MW6-II. Red line, the deletion region failed to complement the recessive lethal. Green line, the deletion region was complementary to the recessive lethal.

26

To understand how those recessive lethals were close to inversion breakpoints, we measured their genetic distances to estimate their recombination rate (Figure 19). By the interpolation method, genetic positions of recessive lethal alleles and inversion breakpoints were estimated (Figure 19), and then the distance between recessive lethals and the nearest breakpoints was calculated (Table 3). The mapping result showed that most of the recessive lethal alleles were close to the nearest breakpoints of the

chromosomal inversions which were polymorphic in the Afrotropical D. melanogaster population. The extent of recombination reduction can be estimated by the distance between recessive lethal alleles and breakpoints of inversions which are polymorphic at three different levels (Table 4 and Figure 20). At the first level, the isofemale line level, the closest genetic distance is between l^ZS2 and the breakpoint of In(3R)93;96, 0.05 cM. When the level of inversion polymorphism enlarged the scale to the strong Z line level and even to the Afrotropical population level, all the recessive lethal alleles are accumulated nearby the breakpoints of inversions, and their genetic distances are less than 3.0 cM. The very short distances between recessive lethal alleles and breakpoints suggest that highly recombination-suppressed region in inversion heterozygotes might lead to the accumulation of recessive lethal alleles. Figure 20 shows the distribution pattern of recessive lethal alleles and chromosomal inversion polymorphism at three different levels. Most recessive lethal alleles (10/14) are located on the right arm of the 3^rd chromosomes, consistent with the higher inversion polymorphism on the right arm than on the left arm (Figure 20).

27

Figure 19 The estimation of genetic positions of recessive lethals and inversion

breakpoints by interpolation of nearest genetic markers. p, proximal breakpoint of the inversion. d, distal breakpoint of the inversion.

28

Table 3 Genetic distances (cM) between each recessive lethals and each nearest inversion breakpoint Lethal

allele

Genetic position

Distance from breakpoints of inversions in Zimbabwe races

In(3R)K In(3R)P In(3R)C In(3R)93;96 In(3R)93;98 In(3L)Ok In(3R)81;92 In(3L)M

50 cM;90.3cM 58.6 cM;85.1 cM 71.2 cM;102.0 cM ~70.55cM;~84.55 cM ~73.2 cM;~99.0 cM ~1.3 cM;~34.5 cM ~47.1 cM;~69.3 cM 26.0 cM;42.0 cM

l^ZH18-I 48.5 1.5 10.1

22.7

22.05

24.7

14.0 1.4 6.5

l^ZH18-II 85.2 5.1 0.1

14.0

0.65

12.0

50.7 15.9 43.2

l^ZH21-I 47.0 3.0 11.6

24.2

23.55

26.2

12.5 0.1 5.0

l^ZH21-II 49.0 1.0 9.6

22.2

21.55

24.2

14.5 1.9 7.0

l^ZS53 56.5 6.5 2.1

14.7

14.05

16.7

12.0 9.4 14.5

l^MW6-I 45.6 4.4 13.0

25.6

24.95

27.6

11.1 1.5 3.6

l^MW6-II 47.1 2.9 11.5

24.1

23.45

26.1

12.6 0.0 5.1

l^ZS30-I 0.0 50.0 58.6

71.2

70.55

73.2

1.3 47.1 26.0

l^ZS30-II 50.6 0.6 8.0

20.6

19.95

22.6

16.1 3.5 8.6

l^ZS30-III 47.6 2.4 11.0

23.6 22.95

25.6

13.1 0.5 5.6

l^ZS30-IV 51.6 1.6 7.0

19.6 18.95

21.6

17.1 4.5 9.6

l^ZS2 84.5 5.8 0.6

13.3

0.05

14.5

50.0 15.2 42.5

l^ZH12 67.6 17.6 9.0

3.6

2.95

5.6

33.1 1.7 25.6

l^ZH32 81.7 8.6 3.4

10.5

2.85

8.5

47.2 12.4 39.7

In(3L)Ok = In(3L)62D;68A. In(3L)M = In(3L)66D;71D. In(3R)K = In(3R)86E17;97A1-2. In(3R)P = In(3R)89B16;96A19. In(3R)C = In(3R)92D1-9;100F2-3.

29

Table 4 Genetic distances (cM) of recessive lethal alleles from the nearest breakpoint of chromosomal inversions at different levels

Lethal alleles

Cytological position

Lethal-bearing chromosomal rearrangements

Genetic distance from the breakpoints of nearest inversions

Isofemale line level Strong Z line level Afrotropical population level

(cM) Nearest inversion

type (cM) Nearest inversion

type (cM) Nearest

inversion type

l

^{MW6-I 73E5-F2}

^In(3R)K

4.40

In(3R)K

1.50

In(3R)81;92

1.50

In(3R)81;92

l

^{MW6-II 82B1-2}

^In(3R)K

2.90

In(3R)K

0.00

In(3R)81;92

0.00

In(3R)81;92

l

^{ZH12 92B3-C1 ST}

2.95

In(3R)93;96

1.70

In(3R)81;92

1.70

In(3R)81;92

l

^{ZH18-I 85D16-D18}

^In(3R)K

1.50

In(3R)K

1.40

In(3R)81;92

1.40

In(3R)81;92

l

^{ZH18-II 96A23-25}

^In(3R)K

0.10

In(3R)P

0.10

In(3R)P

0.10

In(3R)P

l

^{ZH21-I 76A1-D3}

^In(3R)K

3.00

In(3R)K

0.10

In(3R)81;92

0.10

In(3R)81;92

l

^{ZH21-II 82F6-7}

^In(3R)K

2.80

In(3R)K

0.10

In(3R)81;92

0.10

In(3R)81;92

l

^{ZH32 95C13-14 ST}

2.85

In(3R)93;96

2.85

In(3R)93;96

2.85

In(3R)93;96

l

^{ZS2 95E1 ST}

0.05

In(3R)93;96

0.05

In(3R)93;96

0.05

In(3R)93;96

l

^{ZS30-I 61E1-2 ST}

50.00

In(3R)K

1.30

In(3L)Ok

1.30

In(3L)Ok

l

^{ZS30-II 78E1 ST}

3.00

In(3R)K

3.00

In(3R)K

3.00

In(3R)K

l

^{ZS30-III 84B4-E11 ST}

2.40

In(3R)K

0.50

In(3R)81;92

0.50

In(3R)81;92

l

^{ZS30-IV 87B11-C2 ST}

1.60

In(3R)K

1.60

In(3R)K

1.60

In(3R)K

l

^{ZS53 88E2-3}

^In(3R)K

^2.10

^In(3R)P

^2.10

^In(3R)P

2.10

In(3R)P

In(3L)Ok = In(3L)62D;68A. In(3L)M = In(3L)66D;71D. In(3R)K = In(3R)86E17;97A1-2. In(3R)P = In(3R)89B16;96A19. In(3R)C = In(3R)92D1-9;100F2-3.

30

Figure 20 The distribution of recessive lethal alleles and inversions on the 3

^rd chromosome at three different levels. Black lines, inversions found in isofemale lines. Gray lines, inversions found in Strong-Z lines. Light gray lines, inversions found in the Afrotropical population.

Reverse triangle, the centromere.

31

A R programming was run to test whether these recessive lethal alleles distribute randomly on the chromosome or not. The null hypothesis here is that all recessive lethals distribute randomly on the 3^rd chromosome. Two data groups were compared.

One is the true mean genetic distance between each 14 lethals and each nearest

inversion breakpoint (from Table 4). The other is the expected mean genetic distance between 14 random-sampling locations on the 3^rd chromosome to each nearest inversion breakpoint. With random sampling 1000 times, the null distribution was generated as a probability distribution. The statistical P value can be obtained by comparing the true mean distance to the null distribution. All P values from three different

population levels are less than 0.05 (isofemale line level, P = 0.012; strong Z line level,

P = 0.001; Afrotropical population level, P = 0.001), meaning that the null hypothesis

was rejected (Figure 21). The statistical analyses strongly supported that highly non-random distribution of recessive lethals on the 3^rd chromosome was the consequence of high recombination suppression by polymorphic inversions.

32

Figure 21 Comparison between null distribution of genetic distance and true mean

distances. Red arrows are true mean values. (A) The isofemale line level, the true mean is 5.56 cM. (B) The strong Z level, the true mean is 1.02 cM. (C) The Afrotropical level, the true mean value is 1.02 cM.

33

Discussion

My data clearly shows that highly recombination-suppressed regions are likely to accumulate deleterious mutations (recessive lethal alleles in this case). All the 14 recessive lethal alleles mapped are very close to the nearest breakpoints of inversions found in Afrotropical populations. Here, I describe how inversions affect

recombination rates along chromosomes, then possible pattern of deleterious mutations accumulating on the chromosomes with polymorphic inversions, and the association between recessive lethal alleles and chromosomal inversion polymorphisms in natural populations.

High recombination suppression due to inversion heterozygotes

Recombination rates in inversion heterozygotes are highly variable along the inverted chromosomal region. Outside the inversion, they decrease far from chromosomal inversion, and reach the lowest rate nearby breakpoints; inside the inversion, they go up gradually from the inversion breakpoints, and reach the highest peak in the middle region of the inversion (Navarro et al. 1997). My result is the first study to show the fine-scale mapping of recessive lethal alleles to be located nearby breakpoints of chromosomal inversion (Table 4), providing the evidence that deleterious mutation accumulation is in the low recombination region of inverted heterozygotes, as previous natural population surveys suggested (Epling et al. 1961; Dobzhansky et al.

34

1963; Zivanović et al. 2000). In addition, the reduced recombinational effects of the chromosomal inversions might extend to several megabases beyond the breakpoints of inversions. The inhibition ranges of recombination by chromosomal inversions have been described in some Drosophila species. In D. melanogaster, recombination suppression might extend to about over 1 Mb from the breakpoints of inversion In(2L)t, (Depaulis et al. 1999); similarly, recent genomic study showed large regions of linkage disequilibrium on both sides of inversion breakpoints of In(3R)Mo (0.5 Mb and 1.5 Mb), and the LD at long distances is overall greater for 3R then any other chromosome arms (Langley et al. 2012). From my experimental data, the

recombination-suppressed region restricts to3 cM, roughly equal to 0.5Mb long, showing the consistence with the genomic study by Langley et al. (2012). In D.

subobscura, the extents of recombination suppression in the regions outside the distal

breakpoints are similar in different inversions (3.5 and 3.9 Mb for inversions O6 and

O7, respectively) (Pegueroles et al. 2010). In the F

₁ hybrids of D. pseudoobscura and

D. persimilis, the recombination suppression varies between 2.1 - 2.8 Mb outside the

XR inversion (Kulathinal et al. 2009). Meanwhile, in my experiments, the

recombination suppression is larger than 20 cM in In(3R)K/ST heterozygotes (Figure B1, Table B4). Though markers are limited, recombination is suppressed between two markers, cu and sr, which on both sides of In(3R)K proximal breakpoint (Table B4), also two markers, e and ca for distal breakpoints. Because those

35

recombination-suppressed regions contain a large amount of genes, it may have a greater effect on genetic loads once deleterious mutations occur. It was suggested that the size of inversion affected the recombination rate; large inversions (genetic distances

> 20 cM) are expected to achieve an observable rate of double-crossover inside the inversion (Navarro et al. 1997). In the largest inversion in this study, In(3R)K which covers 40.3 cM, I observed no double-crossover inside the inversion and only observed one outside the inversion among 468 offspring (Table B4, grey). The recombination rate and amount of recombinant types in this inversion heterozygote is much lower and less than in the homozygote (Appendix B, Table B5), indicating that that recombination rate along the chromosome is dramatically inhibited by chromosomal inversions. The difference between the large region of recombination suppression revealed by my double crossing-over data and the small region genomic data suggests that gene conversion might play a role to increase the gene exchange inside the inversion (Chovnick 1973; Navarro et al. 1997).

Accumulation pattern of recessive lethals

The processes of deleterious mutation accumulation along the genome have been suggested in some theoretical works. Muller (1964) proposed a hypothesis that in an asexual population, due to no recombination, distinct deleterious mutations at various loci create a spectrum of genotypes carrying increment mutations from 0 to many.

36

Individuals in this population can carry more mutations than their ancestors did (Muller 1964). Moreover, because of genetic drift in a finite population, the lowest among class of deleterious mutations may be lost by chance from the population. At the same time, with no recombination, all remaining individuals acquire more mutations than last generation. Thus, individuals in this population undergo the irreversible process and cannot escape from bearing some among of deleterious mutations. This situation can also be applied to the non-recombination diploid sexual population, such as the Y chromosome in Drosophila species, which lacks of recombination in males

(Charlesworth 1978). Similarly, chromosomal inversion heterozygotes could generate highly recombination-suppressed regions of the genome (breakpoints of inverted regions in this case), and initiate the Muller’s ratchet as well. According to the distribution of the recessive lethal numbers on the eight 3^rd chromosomes in my study (Figure 4), even with small sample size, the accumulation of deleterious mutations reveals that the population is undergoing Muller’s ratchet process. The rate of

Muller’s ratchet turns faster when small population size because of stronger genetic drift effect. The simulation study of Charlesworth and Charlesworth (1997) also revealed the same situation in the non-recombination diploid sexual population. Combining with the empirical data of recessive lethal alleles, there are two possibilities for the accumulation of recessive lethal. One is that those recessive lethals are accumulated in laboratory stocks with the extreme small population; another is that the effective

37

population size in nature is small. For the first possibility, it is likely that very few, if not none, recessive lethals were accumulated in the isofemale line based on the

following reasons. The simulation studies by Charlesworth et al. (1993) and

Charlesworth and Charlesworth (1997) revealed that it took a very long time (more than 2000 generations) for deleterious mutations fixation with non-recombination and/or inbreeding populations (Charlesworth et al. 1993; Charlesworth and Charlesworth 1997). Taking the established time of our isofemale lines into consideration, 20 years (i.e., 200-400 generations) is not long enough for all 14 recessive lethals occurring after collected from the wild. In addition, complex inversion In(3LR)TM3, which has been used as a balancer since first described in 1958 (Mitchell 1958), bears two to five recessive lethal alleles which was identified by my deficiency mapping (Table B3) as a by-product of my experiments (Table B6, Figure B). Compared to the small number of recessive lethal alleles accumulated in the extreme small laboratory population for over 50 years, most of the 14 recessive lethal alleles are more likely to be from the wild.

Moreover, if recessive lethals occur after established in the vial with a small population size, they might be quickly eliminated by genetic drift. From the distribution data of these 14 recessive lethals (Figure 20), some recessive lethal alleles are located far from

In(3R)K, such as l

^ZS30-I , which still exist in the stock without the shelter by inversion

breakpoints. This data strengthen that most of the recessive lethal alleles accumulated

38

in the wild. The last point, it is worth mentioning that mutations occur in the extreme small laboratory population and deleterious mutations accumulate around highly recombination-suppressed regions are not mutually exclusive. For second possibility, Chang and colleagues (Chang and Lin 1995; Chang et al. 1996; Yang et al.2002) hypothesized that in D. albomicans, deleterious mutation accumulation resulted from the bottleneck effect due to the population shrinking between seasons. The D.

melanogaster populations used in this study are from Zimbabwe and the nearby regions

and have strong sexual isolation with other populations (Wu et al. 1995; Hollocher et al.

1997). This strong behavior isolation may reduce the effective population size.

Meanwhile, the high chromosomal inversion polymorphism in natural populations could build complex heterokaryotype combinations, which might cause a great variation of recombination rates along the chromosome and thus result in different mutation accumulation hotspots. In that case, distinct recessive lethal alleles will be sheltered in different chromosomal regions with great recombination suppression.

Chromosomal inversion polymorphisms vs. genetic load

Chromosomal inversion polymorphisms in natural populations may be maintained mainly by positive selection toward local co-adapted genes (Lowry and Wills 2010;

Ayala et al. 2011; Joron et al. 2011; Rottschaefer et al. 2011) and also possibly by genetic load which eliminates the homozygous deleterious mutations, at least in the case

39

of some Drosophila species (Andjelković et al. 1998; Zivanović et al. 2000). In my study with the Afrotropical population of D. melanogaster, lethal mutations are accumulated in highly recombination-suppressed regions by inversion

heterokaryotypes, providing a great amount of genetic loads in populations. With distinct chromosome rearrangements, deleterious mutations are produced independently by genetic drift at first, and then accumulate in different recombination-suppressed regions. Once the Muller’s ratchet starts to work, because of the various mutation hotspots along the chromosome caused by various inversion heterozygotes, genetic loads are hardly fully eliminated from populations. However, the irreversible process of mutation accumulation could increase the effect of inbreeding depression and decrease the fitness in assortative mating populations; some deleterious mutations finally get loss in populations. In this situation, deleterious mutations are neither completely eliminated nor fixed in the population, but are sheltered by inverted

heterozygotes in the population at low to moderate frequency. Meanwhile, the amount of genetic loads might keep a dynamic equilibrium in populations. By knowing the recombination-suppressed effects on deleterious mutation accumulation, my results provide the strong evidence that the accumulation of recessive lethal alleles is associated with high inversion polymorphism in the Afrotropical D. melanogaster population.

40

References

Anderson, A. R., J. E. Collinge, A. A. Hoffmann, M. Kellett, and S. W. McKechnie, 2003 Thermal tolerance trade-offs associated with the right arm of chromosome 3 and marked by the hsr-omega gene in Drosophila melanogaster. Heredity 90:

195-202.

Andjelković, M., G. Zivanović, and M. Milanov, 1998 Genetic loads and coadaptivity of chromosomal inversions I. Inversion polymorphism and genetic load on

chromosome O in a Drosophila subobscura population from Petnica. J. Zool. Syst.

Evol. Res. 36: 123-128.

Aulard, S., J. R. David, and F. Lemeunier, 2002 Chromosomal inversion

polymorphism in Afrotropical populations of Drosophila melanogaster. Genet.

Res. 79: 49-63.

Ayala, D., M. C. Fontaine, A. Cohuet, D. Fontenille, R. Vitalis et al., 2011

Chromosomal inversions, natural selection and adaptation in the malaria vector

Anopheles funestus. Mol. Biol. Evol. 28: 745-758.

Barnes, P. T., 1983 Balancing selection, inversion polymorphism and adaptation in DDT-resistant populations of Drosophila melanogaster. Genetics 105: 87-104.

Carson, H. L., 1946 The selective elimination of inversion dicentric chromatids during meiosis in the eggs of Sciara impatiens. Genetics 31: 95-116.

Chang, H., and F.-J. Lin, 1995 The interaction between chromosomal inversion and recessive lethal alleles in Drosophila albomicans. Zool. Stud. 34: 47-54.

Chang, H., S.-F. Lan, and F.-J. Lin, 1996 Population significance of high frequency recessive lethals in Drosophila albomicans. Zool. Stud. 35: 138-145.

Charlesworth, B., 1978 Model for evolution of Y chromosomes and dosage compensation. Proc. Natl. Acad. Sci. USA 75: 5618-5622.

41

Charlesworth, B., and D. Charlesworth, 1997 Rapid fixation of deleterious alleles can be caused by Muller’s ratchet. Genet. Res. 70: 63-73.

Charlesworth, D., M. T. Morgan, and B. Charlesworth, 1993 Mutation accumulation in finite outbreeding and inbreeding populations. Genet. Res. 61: 39-56.

Chovnick, A., 1973 Gene conversion and transfer of genetic information within the inverted region of inversion heterozygotes. Genetics 75: 123-131.

Crumpacker, D. W., and V. M. Slaceda, 1969 Chromosomal polymorphism and genetic load in Drosophila pseudoobscura. Genetics 61: 859-873.

Depaulis, F., L. Brazier and M. Veuille, 1999 Selective sweep at the Drosophila

melanogaster Suppressor of Hairless locus and its association with the In(2L)t

inversion polymorphism. Genetics 152: 1017-1024.

Dobzhansky, T. G., 1937 Genetics and the Origin of Species. Columbia University Press, New York.

Dobzhansky, T. G., 1947 Adaptive changes induced by natural selection in wild populations of Drosophila. Evolution 1: 1-16.

Dobzhansky, T. G., 1970 Genercis of Evolutionary Process. Columbia University Press, New York

Dobzhansky, T. G., and M. L. Queal, 1937 Genetics of natural populations. I.

Chromosome variation in populations of Drosophila pseudoobscura inhabiting isolated mountain ranges. Genetics 23: 239-251.

Dobzhansky, T. G., B. Spassky, and T. Tidwell, 1963 Genetics of natural populations.

XXXII. Inbreeding and the mutational and balanced genetic loads in natural populations of Drosophila pseudoobscura. Genetics 48: 61-373.

Dubinin, N. P., 1946 On lethal mutations in natural populations. Genetics 31: 21-38.

Epling, C., V. E. Tinderholt, and R. H. T. Matonni, 1961 Frequencies and allelism of

42

lethal factors within and between gene arrangements. I. San Jacinto mountains.

Evolution 15: 447-454.

Hoffmann, A. A., and L. H. Rieseberg, 2008 Revisiting the impact if inversions in evolution: from population genetic markers to drivers of adaptive shifts and speciation? Annu. Rev. Ecol. Evol. Syst. 39: 21-42.

Hoffmann, A. A., C. M. Sgrò, and A. R. Weeks, 2004 Chromosomal inversion polymorphisms and adaptation. Trends Ecol. Evol. 19: 482-488.

Hollocher, H., C.-T. Ting, F. Pollack, and C.-I Wu, 1997 Incipient speciation by sexual isolation in Drosophila melanogaster: variation in mating preference and correlation between sexes. Evolution 51:1175-1181.

Ihaka, R and R. Gentleman, 1996 R: A Language for Data Analysis and Graphics. J.

Comp. Graph. Stat 5: 299-314.

Joron, M., L. Frezal, R. T. Jones, N. L. Chamberlain, S. F. Lee et al., 2011 Chromosomal rearrangements maintain a polymorphic supergene controlling butterfly mimicry. Nature 477: 203-206.

King, M., 1993 Species Evolution: The Role of Chromosome Change. Cambridge University Press, Cambridge.

Kirrkpatrick, P., and N. Barton, 2006 Chromosomal inversions, local adaptation, and speciation. Genetics 173: 419-434.

Krimbas, C. B., and J. R. Powell, 1992 Drosophila Inversion Polymorphism. CRC Press, Boca Raton, FL.

Kulathinal, R. J., L. S. Stevison, and M. A. F. Noor, 2009 The genomics of speciation in Drosophila: diversity, divergence, and introgression estimated using

low-coverage genome sequencing. PLoS Genet. 5: e1000550.

Langley, C. H., K. Stevens, C. Cardeno, Y. C. G. Lee, D. R. Schrider et al., 2012

43

Genomic variation in natural populations of Drosophila melanogaster. Genetics doi: 10.1534/genetics.112.142018.

Lindsey, D. L., and G. G. Zimm, 1992 The Genome of Drosophila melanogaster.

Academic Press, New York.

Lowry, D. B., and J. H. Wills, 2010 A widespread chromosomal inversion

polymorphism contributes to a major life-history transition, local adaptation, and reproductive isolation. PLoS Biol. 8: e1000500.

Mayhew, S. H., S. K. Kato, F. M. Ball, and C. Epling, 1966 Comparative Studies of arrangements within and between populations of Drosophila pseudoobscura.

Evolution 20: 646-662.

Mestres, F., C. Pegueroles, A. Prevosti, and L. Serra, 1990 Colonization of America by Drosophila subobscura: Lethal genes and the problem of the O5 inversion.

Evolution 44: 1823-1836.

Mitchell D. F. 1958 Inversion heterozygosity, recombination, and variability of quantitative traits. Cold Spring Harb. Symp. Quant. Biol. 23: 279-290.

Muller, H. J. 1964 The relation of recombination to mutational advance. Mutat. Res.

106:2-9.

Navarro, A., and N. H. Barton, 2003 Accumulating postzygotic isolation genes in parapatry: a new twist on chromosomal speciation. Evolution 57: 447-459.

Navarro, A., E. Betrán, A. Barbadilla, and A. Ruiz, 1997 Recombination and gene flux caused by gene conversion and crossing over in inversion heterokaryotypes.

Genetics 146: 695-709.

Noor, M. A. F., K. L. Grams, L. A. Bertucci, and J. Reiland, 2001 Chromosomal inversions and the reproductive isolation of species. Proc. Natl. Acad. Sci. USA 98:

12084-12088.

44

Payne, F. 1924 Crossover modifiers in the third chromosome of Drosophila

melanogaster. Genetics 9: 327-342.

Pegueroles, C., V. Ordóñez, F. Mestres, and M. Pascual, 2010 Recombination and selection in the maintenance of the adaptive value of inversions. J. Evol. Biol. 23:

2709-2717.

Rottschaefer, S. M., M. M. Riehle, B. Coulibaly, M. Sacko, O. Niaré et al., 2011 Exceptional diversity, maintenance of polymorphism, and recent directional selection on the APL1 malaria resistance genes of Anopheles gambiae. PLoS Biol.

9: e1000600.

Rieseberg, L. H., 2001 Chromosomal arrangements and speciation. Trends Ecol. Evol.

16: 351-358.

Schaeffer, S. W., M. P. Goetting-Minesky, M. Kovacevic, J. R. Peoples, J. L. Graybill

et al., 2003 Evolutionary genomics of inversions in Drosophila pseudoobscura:

Evidence for epistasis. Proc. Natl. Acad. Sci. USA 100: 8319-8324.

Stevison, L. S., K. B. Hoehn, and M. A. F. Noor, 2011 Effects of inversions on within- and between-species recombination and divergence. Genome Biol. Evol. 3:

830-841.

Sturtevant, A. H., 1921 A case of rearrangement of genes in Drosophila. Proc. Natl.

Acad. Sci. USA 7: 235-237.

Sturtevant, A. H., 1937 Autosomal lethals in wild populations of Drosophila

pseudoobscrura. Biol. Bull. 73: 542-551.

Sturtevant, A. H., and K. Mather, 1938 The interrelations of inversions, heterosis and recombination. Am. Nat. 72: 447-452.

Watanabe, T. K., 1969 Frequency of deleterious chromosomes and allelism between lethal genes in Japanese natural populations of Drosophila melanogaster. Japan. J.

45

Genetics 44: 171-187.

Watanabe, T. K., and T. Yamazaki, 1976 Evidence for coadaptation: negative correlation between lethal genes and polymorphic inversions in Drosophila

melanogaster. Genetics 82: 697-702.

Weeks, A. R., S. W. McKechnie1, and A. A. Hoffmann, 2002 Dissecting adaptive clinal variation: markers, inversions and size ⁄stress associations in Drosophila

melanogaster from a central field population. Ecol. Letters 5: 756-763.

Wu, C.-I, H. Hollocher, D. J. Begun, C. F. Aquadro, Y. Xu, et al., 1995 Sexual

isolation in Drosophila melanogaster: A possible case of incipient speciation. Proc.

Natl. Acad. Sci. USA 92: 2519-2523.

Yang, Y.-Y., F.-J. Lin, and H. Chang, 2002 Comparison of recessive lethal

accumulation in inversion-bearing and inversion-free chromosomes in Drosophila.

Zool. Stud. 41: 271-282.

Zivanović, G., M. Andjelković, and D. Marinkovic, 2000 Genetic load and coadaptation of chromosomal inversions. II. O-chromosomes in Drosophila

subobscura populations. Hereditas 133: 105-113.