Causes and Consequences of Recombination Rate Variation in Drosophila
by
Laurie Sherie Stevison
Department of Biology
Duke University
Date:_______________________
Approved:
___________________________
Mohamed Noor, Supervisor
___________________________
Corbin D. Jones
___________________________
Thomas Mitchell-Olds
___________________________
Gregory Wray
___________________________
David MacAlpine
Dissertation submitted in partial fulfillment of
the requirements for the degree of Doctor of Philosophy in the Department of
Biology in the Graduate School
of Duke University
2011
ABSTRACT
Causes and Consequences of Recombination Rate Variation in Drosophila
by
Laurie Sherie Stevison
Department of Biology
Duke University
Date:_______________________
Approved:
___________________________
Mohamed Noor, Supervisor
___________________________
Corbin D. Jones
___________________________
Thomas Mitchell-Olds
___________________________
Gregory Wray
___________________________
David MacAlpine
An abstract of a dissertation submitted in partial
fulfillment of the requirements for the degree
of Doctor of Philosophy in the Department of
Biology in the Graduate School
of Duke University
2011
Copyright by
Laurie Sherie Stevison
2011
Abstract
Recombination occurs during meiosis to produce new allelic combinations in
natural populations, and thus strongly affects evolutionary processes. The model system
Drosophila has been crucial for understanding the mechanics underlying recombination
and assessing the association between recombination rate and several evolutionary
parameters. Drosophila was the first system in which genetic maps were developed using
recombination frequencies between genes. Further, Drosophila has been used to
determine genetic and environmental conditions that cause variation in recombination
rate. Finally, Drosophila has been instrumental in elucidating associations between local
recombination rate and nucleotide diversity, divergence and codon bias, as well as
helping determine the causes of these associations.
Here I present a fine-scale map of recombination rates across two major
chromosomes in Drosophila persimilis using 181 SNP markers spanning two of five
major chromosome arms. Using this map, I report significant fine-scale heterogeneity of
local recombination rates. However, I also observed ‘‘recombinational neighborhoods,’’
where adjacent intervals had similar recombination rates after excluding regions near
the centromere and telomere. I further found significant positive associations of finescale recombination rate with repetitive element abundance and a 13-bp sequence motif
known to associate with human recombination rates. I noted strong crossover
interference extending 5–7 Mb from the initial crossover event. Further, I observed that
fine-scale recombination rates in D. persimilis are strongly correlated with those
obtained from a comparable study of its sister species, D. pseudoobscura. I documented
a significant relationship between recombination rates and intron nucleotide sequence
iv
diversity within species, but no relationship between recombination rate and intron
divergence between species. These results are consistent with selection models
(hitchhiking and background selection) rather than mutagenic recombination models for
explaining the relationship of recombination with nucleotide diversity within species.
Finally, I found significant correlations between recombination rate and GC content,
supporting both GC-biased gene conversion (BGC) models and selection-driven codon
bias models.
Next, I looked at the role of chromosomal inversions in species maintenance by
examining the impact of inversions distinguishing species to disrupt recombination rates
within inverted regions, at inversion boundaries and throughout the remainder of the
genome. By screening nearly 10,000 offspring from females heterozygous for 3 major
inversions, I observed recombination rates within an inverted region in hybrids between
Drosophila pseudoobscura and D. persimilis to be ~10-4 (similar to rates of exchange for
inversion heterozygotes within species). However, despite the apparent potential for
exchange, I do not find empirical evidence of ongoing gene exchange within the largest of
3 major inversions in DNA sequence analyses of strains isolated from natural
populations. Finally, I observe a strong ‘interchromosomal effect’ with up to 9-fold
higher (>800% different) recombination rates along collinear segments of chromosome
2 in hybrids, revealing a significantly negative association between interchromosomal
effect and recombination rate in homokaryotypes, and I show that interspecies
nucleotide divergence is lower in regions with larger changes in recombination rates in
hybrids, potentially resulting from greater interspecies exchange. This last result
suggests an effect of chromosomal inversions on interspecies gene exchange not
considered previously.
v
Finally, I experimentally tested for a novel male-mediated effect on female
recombination rates by crossing males that differed by either induced treatment
variation or standing genetic variation to genetically identical females. After assaying
recombination frequency in the offspring of these genetic crosses, I fitted these data to a
statistical model where I showed no effect of male temperature treatment or male
genetic background on offspring recombination rate. However, I did observe a difference
of recombination rates of offspring laid 5-8 days post-mating between males treated with
Juvenile Hormone relative to control males. Environmental variation in male ability to
affect recombination rate in their mates suggests the potential for sexual conflict on
optimal proportion of recombinant offspring, perhaps leading to changes in populationlevel recombination rates with varying levels of sexual selection.
Overall, my map of fine-scale recombination rates allowed me to confirm findings
of broader-scale studies and identify multiple novel features that merit further
investigation. Furthermore, I have identified several similarities and differences between
inversions segregating within vs. between species in their effects on recombination and
divergence, and I have identified possible effects of inversions on interspecies gene
exchange that had not been considered previously. Finally, I have provided some
evidence that males may impact female recombination rates, although future work
should attempt to explore the range of male differences that impact this trait and the
mechanism through which males impact the outcome of female meiosis.
vi
Dedication
I dedicate the work presented here to the loving memory of Jesse Brit Lovvorn
(May 3, 1934 – October 13, 2010) and Sarah Purcello Seifert (October 8, 1920 – August
26, 2010).
vii
Contents
Abstract .............................................................................................................................. iv
Dedication ......................................................................................................................... vii
List of Tables ..................................................................................................................... xii
List of Figures .................................................................................................................. xiii
Acknowledgements ............................................................................................................ xv
1.
Introduction: Recombination Rates in Drosophila ...................................................... 1
1.1
Using Recombination Frequencies to Map the Genome ...................................... 2
1.2
Mechanics of Crossing Over ................................................................................. 4
1.2.1
Figuring out the steps ...................................................................................... 4
1.2.2
Additional modern recombination models – SDSA ....................................... 5
1.2.3
Unique features of recombination in Drosophila ............................................7
1.3
Variation in Crossing Over ................................................................................... 8
1.4
Evolutionary Effects of Recombination on Structure of the Genome ................. 9
1.4.1
Nucleotide diversity and divergence ..............................................................10
1.4.2
Causes .............................................................................................................10
1.4.3
Distinguishing between different scenarios ................................................... 11
1.4.4
Codon bias ...................................................................................................... 13
1.4.5
Other patterns ................................................................................................ 14
2. Genetic and Evolutionary Correlates of Fine-Scale Recombination Rate Variation in
Drosophila persimilis ........................................................................................................ 16
2.1
Introduction ........................................................................................................ 16
2.2
Methods ............................................................................................................. 20
2.2.1
Design of Recombination Markers ............................................................... 20
viii
2.2.2
Genotyping..................................................................................................... 21
2.2.3
Construction of the Recombination Map ..................................................... 22
2.2.4
Analyzing Correlates with Recombination Rate ........................................... 24
2.3
Results................................................................................................................ 28
2.3.1
Crossover Rates Along the 2nd and XR Chromosomes ................................ 28
2.3.2
Recombinational Neighborhoods ................................................................. 29
2.3.3
Motif Analysis ............................................................................................... 30
2.3.4
Crossover Interference.................................................................................. 33
2.3.5
Comparison of Crossover Rate Between Species .......................................... 33
2.3.6
Nucleotide Diversity Within Species and Divergence Between Species....... 34
2.3.7
Codon Bias and Biased Gene Conversion ..................................................... 36
2.4
Discussion .......................................................................................................... 38
2.4.1
Broad-Scale Patterning of Recombination Rates.......................................... 39
2.4.2
DNA Sequence Motifs Associated with Crossover Rate in Drosophila ........ 40
2.4.3
Recombination Rates Strongly Correlate Between Species........................... 41
2.4.4
Correlation to Diversity and Divergence at Varying Scales .......................... 42
2.4.5
Codon Bias Versus GC-Biased Gene Conversion (BGC) ............................... 43
2.4.6
Conclusion .................................................................................................... 45
3. The effect of chromosomal inversions on recombination in hybrids between
Drosophila pseudoobscura and D. persimilis .................................................................. 46
3.1
Introduction ....................................................................................................... 46
3.1.1 Expected rates of exchange within inverted regions based on within species
inversion polymorphisms ......................................................................................... 48
3.1.2
Recombination suppression of inversions extends beyond breakpoints ...... 50
3.1.3
Inversions as global recombination modifiers ............................................... 51
ix
3.2
Methods ............................................................................................................. 53
3.2.1
Single generation estimate of inversion crossover rate in hybrids ............... 53
3.2.2
Population genetics analysis of interspecies migration rates within inversion
53
3.2.3 High-throughput genotyping to analyze recombination rate changes
throughout the genome outside inversions .............................................................. 55
3.3
Results.................................................................................................................57
3.3.1
Single generation estimate of inversion crossover rate in hybrids ................57
3.3.2
Population genetics analysis of interspecies migration rates within inversion
58
3.3.3 Analysis of recombination rate reduction in single-generation hybrids at
inversion boundaries ................................................................................................. 61
3.3.4
3.4
Analysis of interchromosomal effect using high-throughput genotyping .... 62
Discussion .......................................................................................................... 64
3.4.1 Exchange across XR inversion detected in one generation, however
prolonged evidence of exchange not observed ......................................................... 64
3.4.2
Recombination suppression extends 2.5-3 Mb at inversion boundaries ..... 65
3.4.3 Interchromosomal effect highest in regions of low recombination along
chromosome 2........................................................................................................... 66
4.
Male-mediated effects on female recombination ...................................................... 69
4.1
Introduction ....................................................................................................... 69
4.2
Methods ..............................................................................................................75
4.2.1
Experiment 1 – Effect of male treatment on female recombination rate ...... 77
4.2.2
rate
Experiment 2 – Effect of male genetic background on female recombination
78
4.3
Results................................................................................................................ 79
4.4
Discussion .......................................................................................................... 86
4.4.1
Experiment 1 – Male hormone treatment alters female recombination rate 87
x
4.4.2 Experiment 2 – Variation in male genotype does not contribute to variation
in female recombination rate .................................................................................... 88
5.
Major Conclusions ..................................................................................................... 90
Appendix A........................................................................................................................ 97
Appendix B...................................................................................................................... 100
References ....................................................................................................................... 104
Biography ......................................................................................................................... 122
xi
List of Tables
Table 2.1: Corrected order of potential misassemblies in the D. pseudoobscura genome
based on D. persimilis recombination map...................................................................... 28
Table 2.2: Results of regression of recombination rate variation with various measures
on chromosome 2 and XR in D. persimilis....................................................................... 35
Table 2.3: Results of diversity within D. persimilis and divergence between D. persimilis
and D. miranda on chromosome 2 calculated at increasing scale (300 kb to 2 Mb). ..... 35
Table 3.1. Summary of diversity and divergence calculations for each of the 11 markers
along the XR chromosome arm. ....................................................................................... 60
Table 3.2. Summary of recombination suppression at inversion boundaries as compared
to Kulathinal et al. 2009. The refined range combines the results of both studies. ......... 61
Table B.1. Proteins Identified as differentially expressed between Zim6 and Zim30 using
Mass Spectrometry. …………………………………………………………………………………………… 102
xii
List of Figures
Figure 1.1: A genetic linkage map of the four chromosomes of Drosophila ....................... 3
Figure 1.2: Two major models of genetic recombination (a) Szostak (1983) DSBR model
(b) Allers and Lichten (2001) SDSA model. Modified from Haber et al. 2004 .................. 6
Figure 2.1: Total recombination map of chromosome 2 (left), and chromosome X, right
arm (XR) (right), each with 95% confidence intervals shown in grey. Centromere is
shown on the right in panel a and on the left in panel b. Map consists of 130 markers on
chromosome 2 and 51 markers on XR assayed in 1330 individuals................................. 24
Figure 2.2: Plot of association between crossover rate in D. persimilis and a 13-bp motif
abundant in human recombinational hotspots (a), and repetitive elements (b). ............ 32
Figure 2.3: Interference results on chromosome 2 in D. persimilis (A), in D.
pseudoobscura (B), and on the XR in D. persimilis (C). .................................................. 33
Figure 2.4: Comparison of chromosome 2 crossover rates between D. persimilis
(generated in this study) and D. pseudoobscura (data from Kulathinal et al. 2008).
Centromeric end is shown on the left. .............................................................................. 34
Figure 2.5: Small intron %GC versus cM/Mb which represents the effects of GC-biased
gene conversion (BGC) only (a), fourfold degenerate 3rd position %GC versus cM/Mb
which represents the combined effects of......................................................................... 37
Figure 3.1. Inter-chromosomal Effect. A plot of the log (base 10) fold change difference in
recombination rates along the 2nd chromosome between the published D. persimilis map
and the map of recombination rate in between species hybrids. ..................................... 63
Figure 4.1 Results of Redfield experiment showing changes in post-mating
recombination rates between markers Stubble (Sb–3R:11.9Mb) and scarlet (st–
3L:16.5Mb) on the 3rd chromosome (Modified from Redfield 1966). ............................... 71
Figure 4.2. Experimental Design. (A) Experiment 1: Females with homozygous lethal
dominant mutants were crossed to wild type males to produce heterozygous females.
Genetic replicate heterozygous females were then crossed to either control or treated
males from two treatment groups: temperature (25°) or Juvenile Hormone (JHa). (B)
Experiment 2: Inbred lines were crossed to produce heterozygous females. Genetic
replicate heterozygous females were then crossed to 1 of 5 males from different genetic
backgrounds. For both experiments, proportion of recombinant offspring were
compared based on the identity of the male parent. ......................................................... 77
Figure 4.3. Results for experiment 1. Results of (A) male temperature treatment and (B)
male hormone treatment on female recombination rate between markers Stubble and
Glued on the 3rd chromosome (~17cM apart). The numbers next to each data point
xiii
indicate the sample size for each male treatment group per two-day collection period and
the bars represent standard error values. ..........................................................................81
Figure 4.4. Results from Experiment 2 prior to accounting for female variation.
Recombination rate variation as a function of male parent for sex-linked markers (A)
and autosomal markers (B) and a summary (C) showing the total results of the
experiment. The numbers next to each data point refers to the number of offspring
genotyped to obtain the mean and standard error values plotted. .................................. 83
Figure 4.5. Results for Experiment 2 accounting for variation in female genotype plotting
recombination rate of offspring based on the genotypes of both the male and female
parents. Numbers indicate the sample size of offspring genotyped at each category on
both the 3rd and X chromosomes. Female genotype was split based on the 4 possible
allelic combinations at the marker CAD (heterozygous in both parental stocks) on
chromosome 2L (20.76Mb). ............................................................................................. 85
Figure A.1. Preliminary results for experiment 2. Contingency tables of the number of
recombinant (R) and non-recombinant (NR) offspring of males from three different
Zimbabwe stocks split into offspring from (A) first 1-4 days post-mating and (B) next 5-8
days post-mating. These were analyzed separately because maternal age affects
recombination rate. .......................................................................................................... 98
Figure B.1. Individual protein results of top three candidates (from Table B.1) showing
variability between each of three replicates in total protein intensity for both zim6 and
zim30 samples. ................................................................................................................103
xiv
Acknowledgements
“If I have seen further than others, it is by standing upon the shoulders of giants” –Isaac Newton
All scientists should begin by acknowledging the vast body of knowledge collected
by researchers who have preceded their own work – it is this body of knowledge to which
I now have the honor of contributing. I would like to give immense thanks to my advisor
Mohamed Noor. He may never fully understand how much his willingness to believe in
me so many years ago has led me to where I am now, and has directed where I want to
go. May he forever be the research scientist in my head, and my dear friend!
Much of the research done during my dissertation would not have been possible
without the guidance, personal insights, mental support, and technical assistance of
several of my lab-mates throughout my time in the Noor lab – including, but not limited
to, Daniel Ortiz-Barrientos, Amanda Moehring, Sarah Bennett, Shannon McDermott,
Annie Boehling, Suzanne McGaugh, Caiti Smukowski, and Brenda Winkler.
My fellow grad students and friends – especially Irene Liu, Carrie OlsonManning, David Garfield, and Matt Johnson – have helped me along the way, have come
to my talks, and have provided me with much needed sanity over the past couple of
years! I would like to thank my family, especially my mom who always told me I could do
anything I wanted if I put my mind to it! I would like to thank my graduate committee
members who have taken the time to guide my progress by challenging me and providing
feedback not only on research, but also on fellowship/grant applications, and on my
career as a research scientist. Funding for the projects presented in this dissertation were
provided by the following grants: NSF-0909824 to myself, and NSF-0715484, NSF0509780, NIH-GM076051, and NIH-GM086445 to Mohamed Noor.
xv
1. Introduction: Recombination Rates in Drosophila
Population genetics is concerned with the inheritance of traits from generation to
generation within populations. Offspring generally resemble their parents, but the
inheritance of phenotypic combinations is not always in the same combinations as the
parents, which leads to ‘recombinant’ offspring. There are two distinct genetic
mechanisms that lead to recombinant offspring: independent assortment and crossing
over. Mendel’s second law of independent assortment explains how each parental
chromosome is equally likely to be segregated to the gametes, such that new
combinations of traits on different chromosomes arise each generation, based simply on
the principles of probability. The focus of this article is on the second mechanism of
producing recombinant offspring, known as crossing over. Although technically the term
‘recombination’ could refer to either independent assortment or crossing over, typical
usage focuses on the latter, and I use it in that context here. The underlying molecular
pathway involves physical exchange between chromosomes and leads to alleles on the
same chromosome being shuffled in the offspring. This process occurs in sexual species
during meiosis and is hypothesized to serve two main categories of functions: to stabilize
chromosomes during meiosis and to increase adaptability of sexual organisms. The
model system Drosophila has been crucial in the scientific exploitation of this process to
determine the order of genes on chromosomes, but some fungal systems have proven
more useful in the discovery of key enzymes in the molecular pathway. Further, surveys
of Drosophila have been instrumental in the discovery of factors that cause
recombination rate variation, as well as how recombination rate variation shapes the
genome.
1
1.1 Using Recombination Frequencies to Map the Genome
Pairs of genes that assort in a 1:1 ratio as expected due to Mendel’s law of
independent assortment are said to be in linkage equilibrium. However, when they
deviate from this expectation, scientists refer to the pair of genes as being in linkage
disequilibrium. One of the most influential genetic observations in the early 1900s was
the deviation of recombination fraction between certain genes from the expected ratio of
1:1. This observation led scientist Alfred Sturtevant to conclude that these genes were
physically linked on a chromosome (Sturtevant 1913). By measuring the recombination
fraction between series of genes in Drosophila, he constructed the first genetic linkage
maps, which are maps of the linear order of genes along a chromosome (Figure 1.1). This
analysis was possible before the development of molecular tools because of the many
different phenotypic mutant markers discovered in the Thomas Hunt Morgan fly lab.
Linkage maps are now generated using a variety of phenotypic or molecular
markers. By using a combination of markers with known relative genomic locations,
scientists can pinpoint where in the genome an unknown mutation resides based simply
on the recombination fraction between markers in a genetic cross and the phenotype of
interest. For species such as humans, genetic crosses between individuals are more
difficult. As a result, scientists either use pedigree information to reconstruct the
genotypes of individuals at various markers or, more recently, survey populations for a
million or more genetic markers at known locations throughout the genome. This
information is then used to infer the order of those genotypes in individuals, referred to
as haplotypes, or to map traits of interest.
2
Figure 1.1: A genetic linkage map of the four chromosomes of Drosophila
3
Although Drosophila played a central role in using recombination fractions to
construct the first genetic maps, fungal systems have been particularly useful in
elucidating the molecular pathway of recombination. One reason for this is the ease of
tetrad analysis with fungal systems such as yeast and Neurospora. Unlike the egg and
sperm products of meiosis in fruit flies and other animals, fungal tetrads constitute four
haploid gametic products of a single meiosis that are fused, and deviations from the
expected 1:1 ratio can be easily observed by dissecting and analyzing these tetrads
directly.
1.2 Mechanics of Crossing Over
1.2.1 Figuring out the steps
It has taken scientists decades to decipher the recombination pathway, and new
critical enzymes are still being discovered. Organisms that lack crucial enzymes offer
insight into these models and suggest additional pathways that can lead to the same
outcome. In 1964, Robin Holliday presented a model for recombination that is still
influential, at least in a modified form, today (Figure 1.2a) (Haber et al. 2004; Stahl
1994). This early model described a single-stranded nick in the deoxyribonucleic acid
(DNA) backbone; however, we learned through the experiments in 1981 by Orr-Weaver
and colleagues that recombination is initiated by a double-stranded break (DSB) in the
DNA (Orr-Weaver et al. 1981; Szostak et al. 1983). In 1997, Scott Keeney and colleagues
discovered the cause of this double-stranded break was an endonuclease protein called
Spo11 in yeast that breaks DNA and causes a series of subsequent recombination events
to occur (Keeney et al. 1997). These experiments were carried out in yeast, but
homologous proteins have been found in other systems including Drosophila.
Early models predicted many of the steps needed for recombination to take place,
however, the actual enzymes that perform these steps were discovered much later. For
4
example, scientists knew meiotic exchange between chromosomes required some type of
DNA strand invasion, but the double-stranded DNA ends needed to become singlestranded first. Then, scientists discovered the Mre11 exonuclease enzyme complex that is
involved in chopping DNA after DSB initiation to create single-stranded DNA (D'amours
and Jackson 2002). This DNA then invades the nearby homologous chromosome and
causes a displacement known as the D-loop formation (see Figure 1.2). The scissorshaped DNA complex that results is known as a double Holliday Junction (dHJ) since
this type of formation was predicted in 1964 in Dr. Holliday’s original model for
recombination. It is the cleavage of these Holliday junctions that decides whether the
recombination pathway results in crossover or non-crossover products. In a crossover
pathway, a horizontal and a vertical cleavage result in rearrangement of the flanking
sequence, whereas two horizontal cleavages result in a non-crossover.
Non-crossovers do not result in downstream genetic exchange, but they can allow
some very localized genetic exchange through mismatch repair. During recombination,
the formation of mismatched DNA resulting from exchange of heterozygous DNA
sequence yields heteroduplexed DNA, where non-complementary bases are paired in the
double helix. This mismatching targets mismatch repair mechanisms in the cell to fix the
strain in the double helix. This can lead either to no exchange or unequal exchange
where one chromosome is modified to have the same genetic material as the homologous
chromosome and the original information is lost. This process is known as gene
conversion, and unlike crossovers which can cause exchange over megabases of DNA,
gene conversion events are generally under a kilobase.
1.2.2 Additional modern recombination models – SDSA
The small scale of gene conversion events makes them difficult to detect without
very dense genetic mapping; however, crossovers are much easier to detect. Despite this,
5
experimental work shows that the majority of recombination products result in noncrossovers rather than crossovers. This observation has led to modification of the simple
Holliday junction resolution models which predicted either outcome to be equally likely.
One such advance is the arrival of synthesis dependent strand annealing (SDSA) models,
which rather than predicting a 50% crossover rate, predicts a majority gene conversion
rate, fitting empirical results better than previous models.
Figure 1.2: Two major models of genetic recombination (a) Szostak (1983)
DSBR model (b) Allers and Lichten (2001) SDSA model. Modified from
Haber et al. 2004
As shown in Figure 1.2, SDSA models differ from the classic double-strand break
repair (DSBR) models in a few critical ways (1) strand invasion and D-loop formation,
(2) location of dHJ relative to DSB and (3) outcome of crossovers versus non-crossovers
(Allers and Lichten 2001; Haber et al. 2004). First, strand invasion is seen as ratelimiting, such that initially only one strand participates until the second strand finally
captures the invading loop to result in a D-loop formation. Unlike DSBR models where
D-loop formation always occurs, D-loop formation is not always the result of SDSA
models. Once D-loop formation has occurred, the resulting dHJ is depicted as on one
6
side of the DSB. In DSBR models, the dHJ is always depicted as flanking the DSB.
Finally, this process results in a majority of non-crossover events as discussed earlier,
which contrasts the traditional DSBR models. This also means that recombination
proceeds in a fashion more reminiscent of transcription rather than replication since the
strand where the DSB formation occurs is the only strand that has newly synthesized
DNA whereas in DSBR, both strands receive some new genetic material.
1.2.3 Unique features of recombination in Drosophila
As discussed earlier, the ability to perform genetic crosses in Drosophila make
them very useful for determining the genomic location of unknown markers in the
reconstruction of genetic maps. The current availability of very dense genetic maps in
multiple species with a variety of markers and marker types make it a great model for
looking at fine scale recombination events. Further, haplotype reconstruction is often not
necessary in fruit flies since the haploid genotypes of individuals are known due to the
widespread use of inbred lines which are homozygous at most genetic loci.
When genetic crosses are performed, an extra layer of experimental control is
available to fruit fly geneticists due to a unique aspect of male Drosophila biology.
Drosophila males do not undergo crossing over, allowing for greater experimental
control over recombination events from generation to generation. The biological
phenomenon where the heterogametic sex, the male in Drosophila, does not undergo
crossing over is referred to as the Haldane–Huxley rule. This is not the case for all
species of Drosophila as Drosophila ananassae males do exhibit some limited crossing
over.
Further, many Drosophila researchers take advantage of ‘balancer’ chromosomes
that have large chromosomal inversions preventing recombination to maintain a known
haplotype within the stock (see Chapter 3: “The effect of chromosomal inversions on
7
recombination in hybrids between Drosophila pseudoobscura and D. persimilis”). Such
balancers are used for making segments of the genome homozygous or for maintaining
haplotypes that are homozygous sterile or lethal. Although the meiotic products of
crossovers within the chromosomal inversion are not viable, the viability of the stock is
not reduced as these meiotic products are shunted to the polar bodies during oogenesis.
This means that genetic crosses using these stocks do not suffer from reduced numbers
of offspring, another useful experimental feature of fruit flies.
Genetic manipulation of fruit flies is also possible to make them even more useful
as a model in studying recombination. As discussed earlier, fungal systems are unique in
their ability to do tetrad analysis; however, a few systems including Drosophila have
found ways to artificially create fused gametes. In Drosophila, irradiation can generate
fused X-chromosomes which are inherited as a single unit in female offspring. These
fused chromosomes, or half tetrads, can be used for tetrad analysis to determine the
location of actual genetic exchange similar to analysis in fungal systems (Hilliker et al.
1994). By using a combination of markers on the X-chromosome, genetic exchange over
very small genetic distances can be observed in females. This technique has been useful
for determining gene conversion rates in Drosophila.
1.3 Variation in Crossing Over
Recombination rates can vary dramatically between species, ranging from no
recombination in asexual systems to very high levels of allelic exchange across the
genomes of some sexual systems. Within any particular genome, meiotic recombination
rates are also variable, such that some regions have higher rates of crossing over than
others. Drosophila have been particularly useful in determining several factors which
cause recombination rate variation. For example, early work in the Morgan lab found
that there is variation with maternal age, where eggs developing later in life exhibit
8
higher numbers of crossovers than eggs generated earlier in life (Bridges 1927; Bridges
1929). Further attempts to examine plasticity in recombination rates have demonstrated
that factors such as temperature (Plough 1917; Plough 1921), nutrition (Neel 1941), age of
mating (Redfield 1966) and number of matings (Priest et al. 2007) also affect
recombination rates in Drosophila (see Chapter 4: “Male-mediated effects on female
recombination”). These factors seem to affect recombination rates in other systems as
well. The pattern that has emerged from the synthesis of these data is that stressful
conditions tend to trigger an increase in recombination rates (Agrawal et al. 2005;
Hadany and Beker 2003; Parsons 1988).
In addition to condition-dependent changes in recombination rates,
recombination within the genome can vary based on genotype. For instance, some alleles
can induce more double-strand breaks leading to higher rates of recombination in
particular regions of the genome (de Massy 2003; Symington 2002). Recent interest in
fine scale variation in recombination rates has led to the discovery of recombination
hotspots which are regions of high recombination limited to very small parts of the
genome. These findings are particularly surprising when summed over the whole
genome. For example, in humans, hotspots cover only 6% of the sequenced genome, but
approximately 60% of recombination in the genome occurs in these locations (Frazer et
al. 2007). Such hotspots seem to be associated with particular DNA sequence motifs in
some cases (Myers et al. 2008).
1.4 Evolutionary Effects of Recombination on Structure of the
Genome
Recombination is well-known for its evolutionary advantages of combining
advantageous alleles onto single haplotypes and removing deleterious alleles from
haplotypes that persist within a species. However, recombination rate variation is also
9
associated with various aspects of the structure of eukaryotic genomes, and the nature of
these associations is sometimes less clear. Three associations that have been studied
extensively in Drosophila species are those of local recombination rate with nucleotide
diversity within species, nucleotide divergence between species and codon usage bias.
1.4.1 Nucleotide diversity and divergence
The seminal compilation of data from 20 loci by Begun and Aquadro (1992)
elegantly demonstrated an overall strong positive association of recombination rate and
nucleotide diversity in this species. Subsequent studies have repeatedly confirmed this
observation in this and other Drosophila species, including whole-genome sequencing
efforts of its sister species Drosophila simulans (Begun et al. 2007), examinations of fine
scale crossover variation and sequence diversity along one chromosome in distantly
related species Drosophila pseudoobscura (Kulathinal et al. 2008) and studies of a few
loci in D. ananassae (Baines et al. 2004).
1.4.2 Causes
Such a positive association between recombination rate and nucleotide diversity
could be driven by a variety of mechanistic or evolutionary causes. Mechanistically, the
simplest explanation would be that the process of crossing over, or its DSB precursor, is
mutagenic. Consistent with this possibility, a few studies have found evidence for a
mutational effect of recombination in primates (Lercher and Hurst 2002), and
functional studies of yeast recombination support possible mutagenicity of DSBR
(Strathern et al. 1995). Mutation rate and recombination rate may also be associated
indirectly via common causes, such as base composition or rates of biased gene
conversion. For example, areas that are targets for DSBs may incidentally be regions that
experience higher mutation rates. Also, heteroduplexed DNA resulting from
10
recombination may be preferentially repaired to keep GC nucleotide bases over AT
nucleotides, creating a GC bias in the gene conversion process.
Multiple evolutionary processes can explain such an association of recombination
rate and nucleotide diversity as well. In regions of low recombination, the spread of a
new, advantageous allele will be associated with a reduction in nucleotide diversity
across a large window of the genome, a process called hitchhiking. In regions of high
recombination; however, the advantageous allele will dissociate from most of its
neighbors during its spread. Negative selection against deleterious mutations can also
generate a similar pattern through a process called background selection. As new
detrimental mutations arise within a population, other alleles at loci near these new
mutations are destined for eventual loss in regions of very low recombination (because
they cannot dissociate from them), whereas new mutations can be eliminated without
loss of much nearby variation in regions of high recombination. Other evolutionary
processes may also contribute to this pattern, such as interference among weakly
selected alleles in close proximity to one another.
1.4.3 Distinguishing between different scenarios
Simple mechanistic causes for an association of recombination rate and sequence
diversity predict that recombination rate should also be associated with sequence
divergence between species. However, the opposite has typically been observed:
Drosophila species often exhibit similar levels of sequence divergence in regions of high
and low recombination (e.g. Noor and Kliman 2003), demonstrating the lack of an
association of recombination rate with divergence (but see Chapter 2: “Genetic and
Evolutionary Correlates of Fine-Scale Recombination Rate Variation in Drosophila
persimilis”). Any association between recombination rate and divergence between
species is consistently weaker than the association of recombination rate and sequence
11
diversity within species. These observations argue against neutral explanations for the
associations in Drosophila.
However, complicating factors may obscure the association of recombination rate
to sequence divergence between species. For example, recombinational ‘hotspots’ arise
and disappear quickly over generations within some species, and such turnover of
recombination rates over time may obscure associations with interspecies divergence
(Spencer et al. 2006). Additionally, studies in Drosophila have assumed that mutations
would be associated with crossovers in particular, but mutations may instead be
associated with DSBs more generally, including ones that result in gene conversion
without crossing over. In regions of severely reduced recombination, such as adjacent to
centromeres, DSBs occur but do not resolve into crossovers. To date, no one has tested
directly for an association between interspecies divergence and DSB rate in Drosophila,
though one study failed to identify such an association among Saccharomyces species
(Noor 2008b). Hence, overall, although selective forces clearly contribute to the
association of recombination rate and sequence diversity within species, it is unclear if
there may also be a (lesser) contribution from possible mutagenicity of recombination.
Perhaps an even greater challenge is distinguishing the relative contributions of
hitchhiking versus background selection as selective forces driving the association of
recombination rate and nucleotide diversity within species. Looking across loci, the
process of hitchhiking should produce an excess of abundant new alleles, and this
frequency distortion of alleles is not predicted by background selection. This prediction
has been developed and applied using test statistics, and strong signatures of hitchhiking
have been identified in genome-wide comparisons. However, detecting hitchhiking from
the spread of adaptive alleles does not disprove the operation of background selection
and its effects on overall patterns of nucleotide diversity. As with mutational
12
contributions, we are left with support for one explanation but an ambiguous picture
regarding the relative contribution of others.
1.4.4 Codon bias
Although the genetic code permits multiple synonymous codons to produce the
same amino acid, many taxa exhibit strong biases for particular codons across their
genome. This ‘codon usage bias’ is thought to be driven largely by selection favoring
efficient translation of codons into proteins, though perhaps the pattern may also be
influenced by mutational biases (see later). Because natural selection is more effective in
regions of high recombination than low recombination as described earlier, selective
explanations for codon usage bias predict that it should be the most apparent in regions
of high recombination. This expectation has been validated, particularly in studies of
Drosophila (e.g. Hey and Kliman 2002 and see Chapter 2).
Many studies have suggested that codon bias is weak in regions of low
recombination because of the so-called Hill-Robertson (1966) effects. Essentially,
selection would be less efficient at particular sites in such regions because of interference
from selection acting on many other nearby sites. For example, a favored codon may be
near an unfavored one, and thus its spread is hampered, particularly in regions of low
recombination. If such interference is common, then looking across a gene, those codons
near the centre should exhibit the most interference (since they have many other codons
near them) whereas codons at the ends of genes should exhibit less interference. As a
result, one would predict that the center codons of long amino-acid coding sequences
should exhibit less codon bias than the codons near the edges, and this prediction has
been supported (Comeron and Kreitman 2002). With high recombination, interference
is reduced because the codons evolve more independently, and the association of codon
13
bias and recombination seems to disappear in Drosophila melanogaster when the
recombination rate exceeds 1.5 cM/Mb (Hey and Kliman 2002).
However, a mutational process may also explain some of the association of
recombination rate and codon bias. In Drosophila, preferred codons typically end in a C
or G. As mentioned earlier, in some recombination hotspots, the repair of mismatches
that arise during recombination is biased towards G+C richness, such that gene
conversion preferentially results in increased numbers of C or G nucleotides (Marais
2003; Marais et al. 2001). If such biased gene conversion solely contributes to the
relationship of recombination to codon bias, one predicts that recombination should be
equally strongly associated with GC content in third codon positions as in nearby
noncoding regions. This predicted relationship was not observed in D. melanogaster
(Hey and Kliman 2002; Kliman and Hey 2003), suggesting that natural selection for Cor G-ending codons is a driving force for causing codon bias.
1.4.5 Other patterns
Many other patterns in the genome associate with recombination rate. From an
evolutionary perspective, the reduced effectiveness of natural selection (through
interference or otherwise) in regions of low recombination should leave multiple other
signatures in the genome. Recombination increases the effectiveness of protein
adaptation throughout the D. melanogaster genome because genes residing in low
recombination regions are more often recombinationally linked with detrimental
mutations for multiple reasons. First, low recombination makes it more difficult to
remove the deleterious alleles and they accumulate in regions of low recombination, and
second, beneficial mutations are less likely to spread in regions of low recombination
than in regions with higher recombination rates and are thus lost due to random chance
more often (Presgraves 2005).
14
Additionally, high transposable element densities have been identified in low
recombination regions of multiple Drosophila species, but the exact form of selection
causing this pattern is not understood (Bartolome et al. 2002; Dolgin and Charlesworth
2008). Various DNA sequence motifs also associate with recombination rate in many
species (Myers et al. 2008), but whether these motifs are actual causes of recombination
rate variation or both parameters covary with another causative factor is not yet known
(see Chapter 2: “Genetic and Evolutionary Correlates of Fine-Scale Recombination Rate
Variation in Drosophila persimilis”). Clearly, we have only begun to scratch the surface
of understanding effects caused by or related to recombination rate variation within and
among genomes.
15
2. Genetic and Evolutionary Correlates of Fine-Scale
Recombination Rate Variation in Drosophila
persimilis
2.1 Introduction
Meiotic recombination rates vary both at coarse and fine scales across almost all
genomes assayed (Coop and Przeworski 2007; Copenhaver et al. 2002; Gerton et al.
2000; Kulathinal et al. 2008; Rockman and Kruglyak 2009; Stephan and Langley 1998;
Sturtevant 1913). Recombination rate variation correlates with various genomic features
such as particular sequence motifs, transposable elements, nucleotide diversity,
nucleotide divergence, GC content, and gene density (Marais 2003; Myers et al. 2005;
Nachman 2002; Petes 2001; Sella et al. 2009). Because these features are not
consistently correlated with recombination rate variation across every biological system,
researchers struggle to identify what causes recombination rates to be drastically higher
in regions known as hotspots of recombination and what the consequences of this
variation are within and across genomes. The various features that correlate with
recombination frequency may have cause/effect associations or may be incidentally
related. For example, sequence motifs associated with higher recombination could
contribute to higher recombination directly by serving as recruitment sites for proteins
involved in double-strand break repair (Baudat et al. 2010; Petes 2001). In contrast,
codon bias has been hypothesized to accumulate in regions of higher recombination due
to more efficient selection (Gordo and Charlesworth 2001; Marais 2003; but see
Marsolier-Kergoat and Yeramian 2009).
Several sequence motifs have been found to be associated with recombination
rate variation, in Drosophila, humans, and yeast (Cirulli et al. 2007; Myers et al. 2005;
Steiner et al. 2009). Recombination proteins bind to ‘open’ chromosomes to induce
16
double-stranded breaks (DSBs), therefore particular sequence motifs can serve as
binding sites for recombination proteins directly or indirectly bind transcription factors
that induce histone modifications and unwind chromatin (Petes 2001). In humans, a 13bp sequence motif associates with 40% of known hotspots of fine-scale recombination
using HapMap data (Myers et al. 2005; Myers et al. 2008). In addition, transposable
and repetitive elements have been highlighted as correlating with recombination rate
variation (Myers et al. 2005).
Over evolutionary time-scales average genomic recombination rates diverge
between species (Dumont and Payseur 2008). Fine-scale recombination rates, such as
the location of particular recombination hotspots, appear to diverge rapidly, as
documented among human populations (Baudat et al. 2010) and in human–chimpanzee
divergence (Ptak et al. 2005; Winckler et al. 2005), whereas broader-scale
recombination rates (~1 Mb) are highly correlated between these species (Duret and
Arndt 2008).
Different genetic processes may influence broad-scale and fine-scale variation in
recombination rates. For example, crossover rate is often reduced in broad regions
adjacent to centromeres (Gaut et al. 2007; Kliman and Hey 1993; Kulathinal et al.
2008). In addition, crossovers in one location of the genome prevent crossovers nearby,
a process known as crossover interference. The level of crossover interference has been
shown to decrease with distance from the initial crossover, with the size of total
interference varying across taxa (102–104 kb) (Copenhaver 2005; Copenhaver et al.
2002; Fitzpatrick et al. 2009; Foss et al. 1993; Hillers 2004; Housworth and Stahl 2003;
Stahl et al. 2004).
The most consistent predictor of recombination rate variation is nucleotide
variation across a genomic region, which has been observed in humans, mice, tomatoes,
17
fruit flies, etc. (Begun and Aquadro 1992; Hellmann et al. 2003; Nachman 1997; for
review see Nachman 2002; Stephan and Langley 1998). Further, this correlation has
been shown to vary with the scale at which recombination is assayed, with stronger
associations at finer scales (Kulathinal et al. 2008; Spencer et al. 2006). Several
hypotheses have been proposed to explain this association (see section “Evolutionary
Effects of Recombination on Structure of the Genome” of Chapter 1), one of which
suggests that the mechanism of crossing over increases local genetic diversity via
mutagenicity of the process itself (Lercher and Hurst 2002). A more widely accepted
hypothesis proposes that recombination rates play a role in how selection shapes the
genome. When non-neutral mutations occur, hitchhiking and background selection
reduce local genetic diversity in linked regions. Recombination mitigates these processes
by reducing the size of the region effectively linked to the mutation (Charlesworth et al.
1993; Maynard Smith and Haigh 1974).
One can distinguish these neutral and selective hypotheses by evaluating the
correlation between recombination rate and nucleotide divergence between species. If
mutations occur more readily in or near the sites of recombination events (e.g., Strathern
et al. 1995), then nucleotide diversity within species and divergence between species will
be elevated at neutral sites in regions of high recombination. In contrast, selective
processes will decrease diversity, but not affect mean divergence at linked neutral sites
(Birky and Walsh 1988; Sella et al. 2009). To test for such hypotheses, for example,
synonymous coding or non-coding sites have been used under the assumption that they
are evolving neutrally or are under only very weak selection (de Procé et al. 2009).
Results of this test vary across systems, with recombination rates correlating with
nucleotide diversity but only sometimes with nucleotide divergence (Begun and Aquadro
18
1992; Begun et al. 2007; Bussell et al. 2006; Coop and Przeworski 2007; Kulathinal et al.
2008; Noor 2008b; Roselius et al. 2005; Tenaillon et al. 2004).
Natural selection or mutational processes may cause associations of
recombination rate with various other genomic parameters, including codon bias, GC
content, and gene density (Birdsell 2002; but see Drouaud et al. 2006; Gerton et al.
2000; Hey and Kliman 2002; Kliman and Hey 2003; Kong et al. 2002) and see section
“Evolutionary Effects of Recombination on Structure of the Genome” of Chapter 1.
Because selection is expected to be more efficient in regions of higher recombination
rate, codon bias should positively correlate with recombination rates (Hey and Kliman
2002; Marais et al. 2001). In Drosophila, more frequently used (‘‘preferred’’) codons
typically end in GC, possibly due to tRNA abundance, and therefore GC-content of third
positions in codons (hereafter GC3) is often used as a measure of codon bias (Akashi and
Schaeffer 1997; Bachtrog 2007). However, such an association of GC3 with
recombination rate may also be driven by biased mutational repair during gene
conversion events. GC-biased gene conversion (BGC) occurs when heteroduplexed
recombination intermediates with A:C and G:T mismatches are repaired preferentially
into the G or C bases, resulting in increased GC-content (presumably also at GC3) across
regions of high recombination (Marais 2003). Distinguishing neutral from selective
explanations for the recombination-GC3 correlation has proven challenging (Marais et
al. 2001).
To examine the various correlates of recombination rates described above, I have
generated a fine-scale recombination map for the second chromosome and right arm of
the X chromosome (hereafter, XR) of Drosophila persimilis, a North American fruit fly
in the D. pseudoobscura species group. A broader-scale linkage map of the second
chromosome was previously generated for D. pseudoobscura (Kulathinal et al. 2008),
19
which diverged from D. persimilis approximately 0.5–1 million years ago. I examined
how recombination rates correlated with GC content, codon bias, repetitive elements,
and specific sequence motifs in D. persimilis. I compared the fine-scale recombination
maps between D. persimilis and D. pseudoobscura, as well as the extent of crossover
interference within these species. Kulathinal et al. (2008) found a strong correlation
with nucleotide sequence diversity within D. pseudoobscura and divergence to D.
persimilis (Kulathinal et al. 2008). However, Noor (2008a) suggested that due to the
recent species divergence, ancestral polymorphism could have complicated the
divergence correlation (see “Discussion” section). Here, I analyzed how fine-scale
recombination rate varies with nucleotide diversity within D. persimilis and divergence
to a more distant relative, D. miranda, to reduce the impact of segregating ancestral
polymorphism (but see Charlesworth et al. 2005).
2.2 Methods
2.2.1 Design of Recombination Markers
Two inbred lines of Drosophila persimilis, Mount Saint Helena (MSH) 1993 and
MSH 3, were crossed to generate heterozygous F1 females. These females were
backcrossed to MSH 1993 males, and 1,440 progeny were genotyped to assay regional
recombination rates. The genome sequence of the MSH 3 strain was published
previously (Clark et al. 2007), and I obtained partial genomic sequence (0.5x) of MSH
1993 using Roche/454 technology (NCBI SRA accession SRA010365) (Margulies et al.
2005). The new 454 sequences were aligned to the MSH3 strain assembly, the D.
pseudoobscura genome sequence assembly 2.0 (with gene annotations from FlyBase
Release 2.3) (Richards et al. 2005), and available 454 sequence for D. miranda (NCBI
SRA accession SRA008268) following the methods outlined in Kulathinal et al. (2009),
simultaneously with the sequence alignments described therein.
20
Design of single nucleotide polymorphism (SNP) based genotyping markers
consisted of selecting markers to correspond to the SNP locations from Kulathinal et al.
(2008) and then defining the remainder of the allotted 384 markers to be evenly spaced
across the two target chromosomes based on the D. pseudoobscura sequence assembly.
Next, the genomic sequence of each region (~10–40 kilobases [kb]) was extracted from
the MSH 3 genome, and the heuristic alignment FASTA (Pearson and Lipman 1988) was
used to search for raw MSH 1993 Roche 454 reads that matched to portions of the target
sequence. I then used a series of custom Perl scripts (available from the Dryad repository
at doi:10.5061/dryad.1877) to parse the FASTA outputs and find SNPs which
distinguished the two inbred lines and contained 100% matching sequence 50 bp
upstream and downstream of the SNP. I used reciprocal BLAST (Altschul et al. 1997) to
determine that the marker was unique in both genomes. When more than one marker
was available for a given target region, I used quality scores and/or the coverage of a
given SNP in the MSH 1993 genome. For a subset of markers, I confirmed that the SNPs
were diagnostic via Sanger sequencing in both lines of D. persimilis to confirm the
sequence. Eight of nine markers confirmed that the putative SNPs differentiated the
lines, and the single error was in the published MSH 3 genome sequence assembly. I
then used the NCBI trace archives to confirm the quality scores of a subset of markers in
the MSH 3 genome. For these markers, the base call was confirmed to be of high quality
and coverage, thus I assumed resequencing errors to be rare.
2.2.2 Genotyping
DNA was isolated from 1,440 backcross offspring using the Gentra PureGene
protocol (Qiagen, Hilden, Germany). SNP markers were screened in all offspring using
the Illumina BeadXpress platform (Fan et al. 2003) (Illumina, Inc. San Diego, CA) at the
Genomic Analysis Facility within the Duke University Center for Human Genome
21
Variation. The output consisted of raw genotypes at all markers for all individuals. A
subset of individuals (106 on chromosome 2 and 110 on X chromosome) did not provide
genotypes at all markers and were not used. A subset of markers did not work on any
individuals (58 markers). Another subset of markers were homozygous in all individuals
and thus represented sequencing errors in one of the two genomes. These sequencing
errors were represented equally between the two genome sequences (64 in MSH 3; 47 in
MSH 1993) suggesting that there was no bias in the quality of the genomes. Finally, 34
markers were excluded which either produced no genotype in 20% or more of the
individuals surveyed or were not heterozygous in the 6 F1 individuals included in the
dataset. The final recombination map on the right arm of the metacentric X chromosome
(XR) consisted of 51 markers, with 672 crossovers in 1,330 individuals over the last 10
Mb of the 30 Mb chromosome arm. The final map of the second chromosome consisted
of 1,334 individuals assayed at 130 markers over the whole 31 Mb of the acrocentric
chromosome with 1,692 crossovers observed. The X chromosome coverage is much less
due to the lack of continuity in the assembly on the centromeric end. I chose to focus on
the 2nd chromosome because it is the only chromosome that both is in a single
assembled scaffold (unlike the X-chromosome arms or 4th chromosome) and does not
bear within-species inversions (unlike chromosome 3). Recombination data on XR was
of interest due to other ongoing related projects (see Chapter 3: “The effect of
chromosomal inversions on recombination in hybrids between Drosophila
pseudoobscura and D. persimilis”).
2.2.3 Construction of the Recombination Map
The centiMorgans per interval were calculated using the number of crossovers
and the total sample size (recombination fraction) and applying a correction for multiple
crossovers and interference of closely linked crossover events (Kosambi 1944). The
22
physical distance between markers was based on the D. persimilis assembly reference
positions, except where intervals spanned multiple scaffolds when the D. pseudoobscura
distance was used as a substitute. Physical distances calculated using the D.
pseudoobscura and D. persimilis reference positions were consistently within 50 kb
except at 2 of 179 intervals (interval 69 and 96 on Chr 2). These intervals were removed
from the scale analyses described below. Distances were used to calculate centiMorgans
per megabase (cM/Mb) for the total map of recombination for each chromosome (Figure
2.1). 95% Confidence intervals were calculated as in Cirulli et al. (2007); I tested whether
the distribution of crossovers across regions deviated from a random distribution of
crossovers via computational resampling by randomly placing the observed number of
crossovers across all intervals, accounting for the sizes of the intervals.
To be conservative, two intervals of anomalously high recombination were
excluded from further analyses. The interval on the second chromosome with
recombination rate 30.1 cM/Mb may be anomalously inflated, as the interval size used in
the denominator could be underestimated since the interval spanned multiple scaffolds
in D. persimilis. On the XR chromosome, the interval at the telomeric end had an
unusually high recombination rate (16.2 cM/Mb) that could not be confirmed by distal
markers, since it was at the end of the map.
23
Figure 2.1: Total recombination map of chromosome 2 (left), and
chromosome X, right arm (XR) (right), each with 95% confidence intervals
shown in grey. Centromere is shown on the right in panel a and on the left in
panel b. Map consists of 130 markers on chromosome 2 and 51 markers on
XR assayed in 1330 individuals.
2.2.4 Analyzing Correlates with Recombination Rate
Centromeric and telomeric regions were included in generating the
recombination map. These regions, however, consistently undergo very little crossing
over even though they experience double-strand breaks and repair with gene conversion.
Thus, they can potentially disrupt correlations of recombination rates to other genomic
features and were excluded from all subsequent analyses (Andolfatto and Wall 2003;
Kulathinal et al. 2008). For this exclusion, 5 Mb upstream and downstream from the
centromere and telomere on both chromosomes were removed from the dataset for
analyses described below.
To compare whole chromosome recombination rates, the total number of
crossovers across the whole chromosome were summed and divided by the total physical
distance. To analyze auto-correlation of recombination rate across adjacent intervals
along my map, I adjusted for size by using the 1 Mb scale recombination map (see
24
below). The recombination rates were regressed against the recombination rate of the
adjacent intervals to determine if recombination rates were dependent on neighboring
regions and how far the auto-correlation extended. To determine particular sequence
motifs that correlate with regions of high recombination rates, I extracted the genomic
sequence of three of the highest and three of the lowest recombination rates (excluding
intervals with no crossovers and intervals with greater than 0.5% N nucleotides). These
were then parsed for frequency of all possible 5-bp motif combinations and summed
over all three high recombination regions and all three low recombination regions
(adjusting for differences in size). I chose 5-bp motifs so as to be consistent with similar
motif searches in D. pseudoobscura (Cirulli et al. 2007), to avoid too many comparisons
reducing statistical power, and because smaller sized fragments have in the past been
useful in later discovery of larger motifs (Myers et al. 2005; Myers et al. 2008). To
determine which motifs were most predictive of recombination rate, motifs were sorted
based on the difference between the frequencies of occurrence of each motif in the high
versus low recombination sequences. The ten motifs with the greatest differences were
chosen for subsequent analyses. A custom Perl script was then used to compute the
frequency of these and other described motifs in each interval in order to regress it
against recombination rate of the interval (see Cirulli et al. 2007). Repetitive element
abundance was calculated using the Tandem Repeats Finder (Benson 1999) and the
percent abundance for each interval was regressed to cM/Mb for each chromosome.
Interference was assayed at the 1 Mb scale, with all recombination intervals
condensed to this scale. If there was a gap in the recombination map >1.5 Mb, the
interference analysis of the region was split. Chromosome 2 had a large gap in the map
(approximately 4 Mb) because of a series of adjacent markers that did not amplify and
was therefore analyzed as two separate regions of approximately 15 Mb each. Coefficient
25
of coincidence was calculated using two different methods. First, a partial measure of
coincidence was calculated as in Weinstein (1918) using the equation:
xn/az
where a is the number of crossovers in the 1st crossover interval being assayed, z
is the number of crossovers in the 2nd crossover interval being assayed, n is the total
number of individuals genotyped, and x is the number of individuals with a crossover in
both a and z. Second, the Schweitzer measure which uses the fraction realized as the
measure of coincidence was calculated (Schweitzer 1934; Schweitzer 1935). For this
measure, I used the equation:
xn/(a+x)(z+x)
where all values are defined the same as above except n, which is the sum of a, z,
x and the total number of individuals with no crossovers in the entire dataset. The
conditional probability of a crossover in one interval based on the presence of a
crossover in another interval, known as interference, was calculated as 1—coincidence for
each measure. This value was then plotted as a function of the interval size. Because a
crossover in an interval can occur in any portion of the region, the distance between the
midpoint of the two regions was used as the interval size. This was calculated using the
following equation:
a/2 + b + c ... + x + y + z/2
where a represents the size of region 1 and z represent the size of region 2 and b
through y represent the size of all the intervals between a and z. Because the previous
study on recombination in D. pseudoobscura did not include interference (Kulathinal et
al. 2008), interference was calculated for this dataset (only chromosome 2 was assayed
in D. pseudoobscura) as a comparison between species and between studies.
26
To compare the recombination rates between D. persimilis and D.
pseudoobscura, I used the recombination map of D. pseudoobscura chromosome 2 from
Kulathinal et al. (2008). For comparison, I reduced my map for D. persimilis to intervals
covering the same genomic region. Because my marker design aimed to produce a map
similar to that from Kulathinal et al. (2008), the borders of the recombination intervals
were very close in genomic location. The recombination rate in D. pseudoobscura was
regressed against the recombination rate in D. persimilis in JMP 8.0 (SAS Institute,
Cary, NC).
For the analysis of diversity and divergence, intron and intergene diversity was
calculated by counting matches and mismatches between the D. persimilis reference
genome and the D. persimilis 454 genome and then dividing the number of mismatches
by the total number of bases. Intron and intergene divergence was calculated similarly,
except the D. persimilis reference genome and the available D. miranda 454 genome
sequences were used. All analyses of introns excluded 10 bp adjacent to each splice site.
More conservative analyses wherein particular bases of small introns were surveyed
(approaches of de Procé et al. 2009; Parsch et al. 2010) were also conducted yielding
nearly identical results, thus I only report the results of the first analysis. For the analysis
of codon bias, a custom Perl script was used to calculate the following for each interval:
total GC content, percent N bases, intergene GC content and GC content for introns
above and below 100 bp (separately). I then summed over each recombination interval:
the %GC3 for all codons, %GC3 for fourfold degenerate codons, average GC3 based on
the number of genes in the interval, and the total number of genes in the interval. Each
calculated parameter was regressed against recombination rate (cM/Mb) for the total
recombination map in StatView (SAS Institute, Cary, NC). To see how these
measurements were affected by averaging over larger recombination intervals, the
27
chromosome 2 dataset was condensed to larger scales (300 kb, 500 kb, 1 Mb, and 2 Mb),
all measures were recalculated, and the regression of each calculation was repeated. The
XR chromosome data were not suitable for scaled analysis because they only spanned 10
Mb (5 Mb once the telomeric end was removed).
2.3 Results
Surveying 1,440 backcross progeny, I observed on average 1.27 crossovers per
individual on chromosome 2 (range 0–5) and 0.51 crossovers per individual on
chromosome XR (range 0–5). Local crossover rates varied along chromosomes from 0.0
cM/Mb to 16.25 cM/Mb. Overall, there was significant heterogeneity of recombination
rates along both chromosome 2 and XR (Figure 2.1, P=0.0001). I also identified a
potential mis-assembly in the D. pseudoobscura genome for Chromosome 2 in the first 3
Mb and on the XR group 6 assembly (see Table 2.1). Comprehensive data associated with
this work can be downloaded from the Dryad repository at doi: 10.5061/dryad.1877.
Table 2.1: Corrected order of potential misassemblies in the D.
pseudoobscura genome based on D. persimilis recombination map.
First 3 Mb of chromosome 2
3' end position
5' start position
3' end position
D. pseudoobscura
D. pseudoobscura
D. persimilis position
D. persimilis position
position
position
Scaffold 7: 4,240,444
Scaffold 7: 2,957,975
Chr 2: 1,440,366*
Chr 2: 172,530
Scaffold 7: 2,734,786
Scaffold 7: 1,745,167
Chr 2: 2,563,917
Chr 2: 1,594,426
Scaffold 7: 1,608,171
Scaffold 7: 58,048
Chr 2: 2,743,727
Chr 2: 4,293,923
Within inversion on XR group 6 that distinguishes D. pseudoobscura & D. persimilis
5' start position
3' end position
5' start position
3' end position
D. pseudoobscura
D. pseudoobscura
D. persimilis position
D. persimilis position
position
position
Scaffold 20: 292,476
Scaffold 35: 799,311
XR group 6: 10,340,780
XR group 6: 8,946,664
Scaffold 27: 941,106
Scaffold 27: 198,713
XR group 6: 7,518,002
XR group 6: 8,475,355
Scaffold 41: 703,394
Scaffold 41: 226,091
XR group 6: 7,347,047
XR group 6: 6,867,474
5' start position
*telomeric end of chromosome arm assembly
2.3.1 Crossover Rates Along the 2nd and XR Chromosomes
The total analysis for XR covered 50 cM with a chromosome average
recombination rate of 5 cM/Mb. The total analysis for chromosome 2 covered 126 cM
28
with a chromosome average recombination rate of 4.1 cM/Mb (frequency of all
crossovers divided by total physical distance assayed). However, without the first and
last 5 Mb of each chromosome’s assembled sequence, the XR covers only 25.6 cM with a
recombination rate of 5.26 cM/Mb, and the 2nd chromosome extends 105 cM with a
chromosome recombination rate of 4.98 cM/Mb. The confidence intervals of these
chromosomal sums overlap extensively, which does not support a difference in the
distribution of recombination rates between the two chromosomes. All analyses
described below were done without these first and last 5 Mb of chromosomes to exclude
possible chromosome- end effects (see “Methods” section) (Gaut et al. 2007; Kliman and
Hey 1993; Kulathinal et al. 2008).
2.3.2 Recombinational Neighborhoods
To test whether recombination in a given interval was independent of
neighboring recombination rates, I regressed recombination rates against the
recombination rate of the adjacent intervals and observed a significant spatial
autocorrelation on chromosome 2 (P = 0.0002; r = 0.415; N = 76) and on the XR (P =
0.0362; r = 0.471; N = 20). The intervals used in the above analysis varied in size from
55 to 547 kb (with one outlier at 620 kb). Because each interval was used twice in the
initial analysis above, once as the focal interval and once as the adjacent interval, I also
removed every other interval to correct for non-independence, and the result remained
significant (P = 0.0002; r = 0.577; N = 46). I further observed a significant Durbin–
Watson autocorrelation of recombination rate (P=0.0001). I attempted to precisely
estimate the size of the ‘‘recombinational neighborhoods’’ by standardizing the intervals
to a size of 1 Mb (using the interference data). At the 1 Mb size-scale, the correlation of
cM/Mb to adjacent cM/Mb was not significant, indicating that the size of the
‘‘recombinational neighborhoods’’ is smaller than 1 Mb (data not shown). Because of
29
variance in initial interval sizes surveyed, I was unable to standardize the dataset to sizescales smaller than 1 Mb (while maintaining comparison of adjacent intervals) to refine
the size of the neighborhoods further.
Since autocorrelation of GC-content has been observed in mammalian genomes
due to isochore structure (Duret et al. 2002), I attempted to rule out any GC-content
structure affected by my recombination autocorrelation. To do this, I took the residuals
of GC-content at a given interval versus recombination rate and regressed these residuals
against the recombination rate in the adjacent window. The results (P = 0.0001; r =
0.389; N = 94) remained highly significant after correcting for GC-content, suggesting
the recombinational neighborhoods are not solely due to GC mediated structuring of
recombination rates.
2.3.3 Motif Analysis
Among all possible 5-bp motifs, AT-rich motifs differed most dramatically in the
three highest recombination regions from the three lowest recombination regions in
abundance. The frequencies of the ten motifs that differentiated the highest and lowest
recombination regions most were counted across all intervals in the total dataset. A
regression with recombination rate was then computed excluding the six regions above. I
found two motifs that correlated significantly, albeit weakly, with recombination rates on
chromosome 2—the 5-bp motifs ATAAA (P = 0.0282; r = 0.262; N = 70) and AATAA (P
= 0.0179; r = 0.282; N = 70). These two motifs were not significantly correlated with
recombination rates on the XR (data not shown). In addition, I analyzed the entire
dataset for the 5-bp motif CACAC identified as significant on the left arm of the X
chromosome in D. pseudoobscura (Cirulli et al. 2007) and found no significant
association on chromosome 2 (P = 0.4293; r = 0.092; N = 76) or XR (P = 0.696; r =
0.084; N = 24).
30
I also examined the 13-bp motif noted as highly correlated to recombination rates
in humans (CCNCCNTNNCCNC: Myers et al. 2008) and found a highly significant
correlation with recombination rates in D. persimilis on the second chromosome
(P<0.0001; r = 0.464; N = 76) and on the XR (P = 0.0231; r = 0.462; N = 24). The
combined results of the 13-bp motif and recombination rates for these two chromosomes
are shown in Figure 2.2a. The association of the 13-bp motif with recombination rates on
both chromosomes was significant via a non-parametric Spearman rank correlation
(P<0.0001; ρ = 0.3814). I corrected for any possible GC-bias due to the 13-bp motif by
computing the residuals of the regression of the percent 13-mer per interval against
percent GC-content per interval. These residuals were regressed against recombination
rate per interval and still shown to be significant (P = 0.0014; r = 0.332; N = 95). I
further examined whether this motif explained recombination rates from D.
pseudoobscura using available chromosome 2 local recombination rates (Kulathinal et
al. 2008) and genome sequence (Richards et al. 2005), and a very strong correlation to
the 13-bp motif was also detected using this dataset (P<0.0001; r = 0.712; N = 40).
I tested for a correlation of repetitive elements with recombination rates using
Tandem Repeats Finder (Benson 1999), and found a highly significant correlation on
chromosome 2 (P = 0.0035; r = 0.331; N = 76) and XR (P = 0.0031; r = 0.578; N = 24).
The combined results of repetitive elements with recombination rates for these two
chromosomes are shown in Figure 2.2b. Because the D. persimilis genome sequence
bears the lowest coverage and quality of the 12 Drosophila genomes, there are many
noncontiguous segments, each containing large numbers of undefined bases. Repetitive
sequences likely contributed greatly to the assembly failures at this low sequence
coverage. Therefore, the percent of undefined bases in an interval is most likely related
to repetitive element abundance in the same interval. Similar to my results for repetitive
31
elements, I found a strong statistical significance between recombination rate and
percent of each interval containing N bases in the sequence on both chromosome 2 (P =
0.018; r = 0.272; N = 76) and XR (P = 0.029; r = 0.456; N = 23).
Finally, I conducted a multiple regression to determine which elements were
most predictive of recombination rate variation—motifs, GC content, repetitive elements,
or %N. The only feature that remained significant in the multiple regression was the 13bp motif (P = 0.0122).
Figure 2.2: Plot of association between crossover rate in D. persimilis and a
13-bp motif abundant in human recombinational hotspots (a), and repetitive
elements (b).
32
2.3.4 Crossover Interference
Crossover interference was very strong in both D. pseudoobscura and D.
persimilis as far as 5–7 Mb, at which point it appeared to vary stochastically around zero
(Figure 2.3a), likely because fewer intervals were surveyed with increasing interval size.
This was true for chromosome 2 in D. pseudoobscura (Figure 2.3b) and D. persimilis as
well as XR in D. persimilis (Figure 2.3c).
Figure 2.3: Interference results on chromosome 2 in D. persimilis (A), in D.
pseudoobscura (B), and on the XR in D. persimilis (C).
2.3.5 Comparison of Crossover Rate Between Species
The recombination dataset for D. persimilis was condensed to intervals
comparable with the dataset available for D. pseudoobscura across chromosome 2. The
recombination rates between these closely related species were highly correlated (Figure
2.4; P<0.0001; r = 0.591; N = 38) across intervals. A non-parametric Spearman rank
correlation also indicated significant covariance of local recombination rates between
species (P = 0.0011; ρ = 0.535; Z = 3.255). In addition, I tested for intervals bearing
recombination rates differing between the species through an interval-by-interval χ2 test
of crossovers versus non- crossovers, using 1 degree of freedom. This analysis revealed
that the recombination rate was significantly different in only 5 of 38 intervals, after
33
Bonferroni correction. One of these 5 intervals (apx. 6 Mb in Figure 2.4) had
recombination markers located on non-contiguous scaffolds, and thus the D.
pseudoobscura physical distance was used (see ‘‘Methods’’ section). This interval may
therefore have a similar measure of recombination if a more precise measure of physical
distance was available. The other notable peaks in D. persimilis recombination rate (apx.
17 and 24 Mb in Figure 2.4) relative to D. pseudoobscura in Figure 2.4 are not
significantly different from one another, suggesting hotspots between these species are
shared at the scale analyzed here.
Figure 2.4: Comparison of chromosome 2 crossover rates between D.
persimilis (generated in this study) and D. pseudoobscura (data from
Kulathinal et al. 2008). Centromeric end is shown on the left.
2.3.6 Nucleotide Diversity Within Species and Divergence Between
Species
Table 2.2 presents regression analyses of recombination rates along the second
chromosome on various measures, including nucleotide diversity within D. persimilis
and divergence between D. persimilis and D. miranda. Along the 20 megabases
surveyed on chromosome 2 (hence, excluding the five megabases at each end),
34
nucleotide diversity of introns was significantly correlated with recombination rate in D.
persimilis, while intergenic sequence diversity was not. Intron and intergene divergence
between species were both not significantly correlated to recombination rates on
chromosome 2.
Table 2.2: Results of regression of recombination rate variation with various
measures on chromosome 2 and XR in D. persimilis.
2nd chromosome
P value
r
0.258
0.133
0.004*
0.333
0.559
0.069
0.128
0.180
0.109
0.191
0.002
0.353
0.002
0.360
0.005
0.329
0.006
0.323
0.011
0.295
Intergene diversity
Intron diversity
Intergene divergence
Intron divergence
Intergene %GC
Big Intron %GC
Small Intron %GC
%GC3
Average GC3 per gene
Total %GC
N
74
73
74
73
72
72
72
72
72
74
XR chromosome
P value
r
0.2329
0.272
0.239
0.269
0.3494
0.215
0.254
0.260
0.052
0.430
0.183
0.300
0.152
0.324
0.670
0.099
0.775
0.067
0.131
0.340
N
21
21
21
21
21
21
21
21
21
21
*Values in bold indicate statistical significance
To determine how variation in size of intervals surveyed influenced
recombination’s correlation to diversity and divergence, I adjusted the dataset to have
standardized interval sizes at varying scales. Intron diversity remained strongly or
significantly correlated to recombination rates across all scales studied (Table 2.3). No
measures were significantly correlated with recombination rate along the XR
chromosome arm (Table 2.2), but only 5 Mb of data were available for analysis after
excluding failed markers and chromosome ends.
Table 2.3: Results of diversity within D. persimilis and divergence
between D. persimilis and D. miranda on chromosome 2 calculated at
increasing scale (300 kb to 2 Mb).
300 kb
500 kb
1 Mb
P value
r
N
P value
r
N
0.034*
0.417
26
0.13
0.305
26
P
value
0.003
*Values in bold indicate statistical significance
35
2 Mb
r
N
0.781
12
P
value
0.161
r
N
0.731
5
2.3.7 Codon Bias and Biased Gene Conversion
GC content was strongly correlated with recombination rates on chromosome 2
and not on XR (Table 2.2). Because small introns (>100 bp) appear to evolve under less
evolutionary constraints than larger introns (de Procé et al. 2009; Haddrill et al. 2005;
Marais et al. 2005; Parsch 2003), I separated the intron GC content analysis into large
(>100 bp) and small (<100 bp) introns (Figure 2.5a). Except for intergene GC, all GC
content calculations were strongly correlated with recombination rates on chromosome
2 (Table 2.2), suggesting the action of BGC.
To examine how codon bias might be influenced by recombination rates, I
computed GC3 across entire intervals and average GC3 per gene within intervals (Akashi
and Schaeffer 1997; Bachtrog 2007). Both measures correlated strongly with
recombination rates (and were nearly identical to each other). Because not all codons can
mutate to an A/T and still create a synonymous change with respect to the amino acid
encoded, I limited my 3rd position codon analysis further to those positions that are
four-fold degenerate and still observed a significant correlation to recombination rate
(Figure 2.5b; P = 0.0146; r = 0.283; N = 74).
36
Figure 2.5: Small intron %GC versus cM/Mb which represents the
effects of GC-biased gene conversion (BGC) only (a), fourfold degenerate
3rd position %GC versus cM/Mb which represents the combined effects of
codon bias and BGC (b), and the subtracted effects of the two which should
reflect the effects of only codon bias (c).
Since a significant association of GC3 at fourfold degenerate sites could be due to
BGC, I sought to separate the effects of BGC and selection-driven codon bias. To do this,
I subtracted small intron %GC content (representing the effects of BGC only) from
fourfold degenerate 3rd codon position %GC (representing the combined effects of BGC
and selection-driven codon bias) within each interval, and I observed a significant
relationship of the independent effect of selection-driven codon bias (Figure 2.5c; P =
0.0183; r = 0.279; N = 71). In addition, following Marais et al. (2001), I computed the
37
residuals from the x versus y regression of small intron %GC content plotted against
fourfold degenerate 3rd codon position %GC and regressed them against recombination
rate. This analysis generated a marginally significant result (P = 0.072; r = 0.215; N =
71), explaining a similar proportion of the variance in recombination rate as the
subtraction analysis above. Another method used by Marais et al. (2001) was to examine
AT preferred codons which should not experience BGC, however, no preferred codons in
D. pseudoobscura have been identified as ending in AT (Akashi and Schaeffer 1997;
Bachtrog 2007). Because neither GC content nor GC3 content was correlated with
recombination rates on the surveyed region of XR, I did not extend analysis of codon
bias or BGC on this chromosome.
2.4 Discussion
Fine-scale linkage maps allow researchers to address a host of genetic and
evolutionary questions (Gerton et al. 2000; Groenen et al. 2009; Ihara et al. 2004;
Maddox et al. 2001; Mancera et al. 2008; Menotti-Raymond et al. 1999; Niehuis et al.
2010; Paigen et al. 2008; Samollow et al. 2004; Shifman et al. 2006; Singer et al. 2006;
van Os et al. 2006; Wong et al. 2010; Zenger et al. 2002). However, because of the
cumbersome nature of their development, direct, empirical estimates of fine-scale
recombination rate are not broadly available. Instead, methods have been developed for
inferring recombination ‘‘hotspots’’ via patterns of nucleotide variation and linkage
disequilibrium (Buard and de Massy 2007; Coop et al. 2008). Some of these methods
rely on particular assumptions such as no population structure, and common
evolutionary forces such as selection can lead to both false positives and false negatives
of hotspots (but see Khil and Camerini-Otero 2010; McVean 2007; Reed and Tishkoff
2006).
38
In this study, I generated a fine-scale map of recombination across two major
chromosomes in Drosophila persimilis using low coverage next-generation Roche-454
resequencing followed by high-throughput genotyping with Illumina BeadXpress
analyzer. Although studies of some other organisms have used far more markers to
develop linkage maps, I focused instead on a large panel of progeny to test, ensuring high
precision of crossover rates estimated. I observed on average 1.27 crossovers per
individual along the 2nd chromosome. I detected both fine-scale and broad-scale
structuring of recombination rate across both chromosomes, crossover rate
autocorrelation in neighboring windows, and crossover interference extending 5–7
megabases from crossover events, consistent with results in D. melanogaster (Foss et al.
1993) and in D. pseudoobscura (Fitzpatrick et al. 2009). Average crossover rate did not
differ between the X-chromosome arm and the second chromosome, but regional
recombination rates were strongly correlated between D. pseudoobscura and D.
persimilis. My primary focus was to evaluate several genetic and evolutionary correlates
with local crossover rate, and I discuss their implications below.
2.4.1 Broad-Scale Patterning of Recombination Rates
I noted a significant correlation of recombination rates among adjacent intervals,
creating what I term ‘‘recombinational neighborhoods.’’ Recombination rate maps of
some other systems appear to also show such spatial autocorrelation, but these studies
fit polynomial curves on observed crossover rates, artificially producing these
relationships (Gaut et al. 2007; Niehuis et al. 2010). I observe these recombinational
neighborhoods without any manipulation of the observed crossover rates, and the
autocorrelation appears to decay in under a megabase. Broad effects on recombination
rates are well known in centromeric and telomeric regions, but such broad structuring of
recombination rates elsewhere in the genome has not been reported previously in
39
Drosophila or other systems to the best of my knowledge, aside from within mutant
constructs (Petes 2001).
This observation opens several additional questions for future study. For
example, do the same mechanisms that specify either fine-scale recombination rate
variation or centromeric/telomeric variation also define the intermediate scale of
recombination rate structuring observed in this study? I observe spatial autocorrelation
of overall G–C content, so this nucleotide compositional patterning may contribute to, or
be a consequence of, the autocorrelation of local crossover rates. Nucleasehypersensitive chromatin may also contribute to the observed broad-scale patterning of
recombination rates (Petes 2001).
2.4.2 DNA Sequence Motifs Associated with Crossover Rate in
Drosophila
I surveyed all possible 5-bp sequence motifs and identified two AT-rich ones
(AATAA and ATAAA) that correlate negatively with recombination rate. This observation
is consistent with my observed positive correlation between recombination rate and GCcontent. Because my motif search was limited to 5-bp motifs, it is possible that these
motifs are part of a larger AT-rich motif that future studies can examine for correlation
with recombination rate variation. I extended my motif analysis to those identified in
previous studies of Drosophila or humans as significantly correlated to recombination
rates. Cirulli et al. (2007) observed significant associations of D. pseudoobscura finescale crossover rate with both simple repeats and the motif CACAC. I confirmed similar
associations in D. persimilis with simple repeats but not CACAC.
More strikingly, I observed a strong positive correlation of both D.
pseudoobscura and D. persimilis crossover rate to a 13-bp motif associated with
crossover hotspots in humans (Myers et al. 2008). The recombinational effect of this
40
motif is suggested to be mediated through the zinc finger DNA binding array of PRDM9
in humans and mice (Baudat et al. 2010). While the association of D. persimilis
crossover rate with this motif suggests similar mechanisms controlling hotspot usage,
the finding of a strong correlation with the exact motif found in humans is surprising
considering the motif is not associated with chimpanzee crossover hotspots (Myers et al.
2010), and there is no known ortholog of PRDM9 in Drosophila. Further investigation of
the variation in recombination rates at different scales may help to tease apart the
mechanistic bases of these associations and the directionality of cause and effect.
2.4.3 Recombination Rates Strongly Correlate Between Species
I observed a strongly significant correlation between D. pseudoobscura and D.
persimilis in fine-scale recombination rates. The conservation of recombination rates in
crosses performed using strains from different species both implies that the crossover
rates I measured are not highly genetically variable within these species and supports the
repeatability of my methods (Kulathinal et al. 2008).
Recombination rate differences between humans and chimps have been found at
very fine scales using different approaches (Ptak et al. 2005; Winckler et al. 2005).
However, the correlation of recombination rates between these species is stronger at
broader scales. These results suggest that correspondence between species for
recombination rates may increase when analyzing larger interval sizes. My study extends
the latter observation in Drosophila by showing a very strong correlation between sister
species for recombination rates at ~500 kb scale. Future studies comparing interspecies
recombination rates should analyze their results at varying scales (see “Correlation to
Diversity and Divergence at Varying Scales” section) to determine if crossover rates are
consistently more strongly correlated between species at broad rather than fine scales.
41
2.4.4 Correlation to Diversity and Divergence at Varying Scales
I note that nucleotide variation within species at introns was significantly
correlated with recombination rates, whereas divergence between species was not
correlated with recombination rate. Because diversity varies significantly across the
genome for reasons other than recombination rate variation (such as functional
constraints), I analyzed intergenic diversity and intron diversity separately (de Procé et
al. 2009; Hellmann et al. 2005). Curiously, I did not observe the same predicted
correlation between diversity and recombination rate on the D. persimilis XR
chromosome arm, but my coverage of this chromosome was low suggesting there may
not have been enough variation in recombination rate or in nucleotide diversity to detect
a correlation. I also did not detect a significant relationship between intergenic diversity
and recombination rate on either chromosome; however, this failure could result from
functional constraints (such as UTRs or ncRNAs) in intergenic regions that limit
nucleotide diversity (de Proce´ et al. 2009). Finally, my measure of divergence based on
a newly obtained D. miranda sequence, is less likely to be confounded with ancestral
polymorphisms than estimates of divergence used previously comparing D.
pseudoobscura and D. persimilis (Kulathinal et al. 2008), which are known to hybridize
at low levels in the wild (Dobzhansky 1973), with detectable gene flow on the nucleotide
level (Machado and Hey 2003; Machado et al. 2002, 2007). Therefore, my measure of
nucleotide divergence using this recently available outgroup sequence is potentially
better than the measure used in our group’s previous publication (Kulathinal et al.
2008).
The lack of correlation of D. persimilis recombination rates and divergence
between D. miranda and D. persimilis is consistent with the selective hypothesis for
explaining the correlation between diversity and recombination rates (Begun and
42
Aquadro 1992; Nachman 2002). However, the lack of a correlation of fine-scale
recombination rate with divergence may also reflect evolutionary changes in the
recombinational landscape (Spencer et al. 2006); although I found that D.
pseudoobscura and D. persimilis share local recombination rates, we do not know that
D. miranda and the common ancestor also share these local rates. This hypothesis is
being tested with a fine-scale linkage map being developed now in D. miranda.
I also detected a continuing association of recombination rate with nucleotide
diversity with interval sizes up to 1 Mb. I tested for scale-dependence of the association
of recombination rate with diversity because nonselective models should predict the
association to be restricted to the finest spatial scale (e.g., Spencer et al. 2006). My
observation that the relationship holds true at much larger scales seems to contradict
this hypothesis, but the observation of recombinational neighborhoods extending out to
similar scales makes strong inference from this pattern uncertain. This increasing
association at broader scales is also consistent with human recombination rate variation
and various correlates (Duret and Arndt 2008).
2.4.5 Codon Bias Versus GC-Biased Gene Conversion (BGC)
I saw a correlation between GC3 content, and intronic GC content to
recombination rates (Table 1). This implies the action of selection-driven codon bias or
BGC (Marais 2003; Marais et al. 2001). These results could be an artifact of higher gene
density in regions of high recombination rates; however, the correlation of gene density
to recombination rate was non-significant (results not shown).
Although calculations of codon bias correlate with recombination rates, in
Drosophila calculations are based on codons which end in G or C. Therefore, the
measure is confounded with non-selective forces casting doubt on the overall conclusion
of these studies (Marais et al. 2001). To further distinguish between selective processes
43
and mutation bias, I examined GC content in small introns (<100 bp) to see if the
mutational bias affected nearby introns which have been shown to be less constrained by
selection (de Procé et al. 2009). I also examined GC content at fourfold degenerate
codons, which have also been shown to be less constrained than non-degenerate codons,
to relate codon bias to recombination rate variation (de Procé et al. 2009). I predicted
that if both BGC and natural selection contribute to stronger codon bias in regions of
high recombination and are acting in concert, then the effects of the two could be divided
by factoring out the effects observed in introns (presumably non-adaptive) from those in
codon 3rd positions (influenced by selection). I found that the difference was still
correlated with fine scale crossover rate (Figure 2.5), and I conclude that both BGC and
codon bias are influencing or influenced by recombination rates. Although, I observe a
signature of selection-driven codon bias, the signature reflects historical processes (see
Zeng and Charlesworth 2010), and selection may not have operated in modern D.
persimilis. By inferring independently, the selective and non-selective forces
contributing to codon third-position GC content, I suggest that both neutral processes
(BGC) and selection processes (codon bias) contribute to the correlation between
recombination rate and GC content in coding regions.
Although BGC is not an adaptive process, the accumulation of codon bias in
regions of high recombination rates is, at least in part, a selective process. Because
recombination acts to increase the local effective population size of the genome, it
reduces the effects of Hill-Robertson interference (HRi). HRi reduces the overall
efficiency of selection because positively selected loci reduce each other’s probability of
fixation (Felsenstein 1974; Hill and Robertson 1966). In regions of high recombination,
HRi is very weak because positively selected loci are more easily combined onto the same
44
haplotype increasing their collective probability of fixation (Felsenstein 1974; Gordo and
Charlesworth 2001; Haddrill et al. 2005).
2.4.6 Conclusion
As high-throughput methods for genome sequencing and assembly as well as
high throughput genotyping technologies increase in availability, recombination maps
are becoming cheaper and less challenging to generate. These types of studies also offer
more empirical evidence as to how recombination rates shape the genome and provide
long-awaited answers to many evolutionary questions regarding recombination rates,
such as what factors shape fine-scale recombination rate variation across genomes and
why there is such a strong relationship between recombination rates and genetic
polymorphism. These studies also generate novel questions such as which factors predict
shifts in recombination landscapes between species and what factors shape broad-scale
recombination patterns across the genome? In the last 30 years, we have come from the
assumption that recombination rates are uniformly distributed across genomes to many
new understandings, but retaining the realization that there is much left to learn about
recombination rates, what makes them so variable, and how they impact patterns of
evolution in natural populations.
45
3. The effect of chromosomal inversions on
recombination in hybrids between Drosophila
pseudoobscura and D. persimilis
3.1 Introduction
In the study of speciation, a crucial question is how species are maintained
despite gene flow. One potential solution is that chromosomal inversions partition the
genome into regions protected from gene flow by reducing recombination over long
stretches. Although reduced recombination in inversion heterozygotes has been well
documented within species (Ishii and Charlesworth 1977; Roberts 1976), there are few
empirical estimates of recombination rates in interspecies hybrids heterozygous for
inversions. Further, by focusing on the importance of inversions in the reduction of
recombination within inverted regions, recent work has largely overlooked the major
global increases in recombination rate in inversion heterozygotes (but see Portin and
Rantanen 1990), known as ‘interchromosomal effect’ (Schultz and Redfield 1951). Rather
than determining the effects of the interchromosomal effect on global nucleotide
variability, this field of research has instead shifted focus towards identifying the
mechanism of this process (Joyce and McKim 2010).
Here, I focus on hybrids between Drosophila pseudoobscura and D. persimilis, a
classical model system for studying chromosomal inversions, hybridization and
speciation (Dobzhansky 1937). D. pseudoobscura is found across Western North
America extending as far north as Canada, as far east as Central Texas and as far south as
parts of Central America. D. persimilis maintains a smaller range restricted mainly to
Western United States, and is contained within the range of D. pseudoobscura
(Dobzhansky and Epling 1944). These species are morphologically identical, however
male hybrids are sterile. They diverged approximately 0.5-0.85 million years ago
46
(Aquadro et al. 1991; Hey and Nielsen 2004; Leman et al. 2005), however parts of the
genome carry a signature of more recent hybridization (Hey and Nielsen 2004;
Kulathinal et al. 2009; Machado et al. 2007; Machado et al. 2002). Despite this ongoing
gene flow, these taxa were diagnosed as different species due to three major paracentric
inversions distinguishing them – two on the X chromosome and one on the 2nd
chromosome (Dobzhansky and Epling 1944). The D. pseudoobscura arrangement of the
right arm of the X (hereafter XR) is also present in populations of D. persimilis
exhibiting meiotic drive with skewed sex ratios, referred to as Sex Ratio D. persimilis
(Sturtevant and Dobzhansky 1936). Additionally, both species segregate multiple
arrangements among individuals within species on chromosome 3 (Powell 1992).
I present here an empirical study examining the extent to which inversions
differentiating these species affect recombination rates (1) within inverted regions, (2) at
inversion boundaries, and (3) throughout the remainder of the genome.
First, I focus on the largest inversion, the XR (not-fixed) inversion, because I
expect higher potential for double-crossover events and gene flow within this inversion
(Navarro et al. 1997). I evaluate the double crossover rate in female hybrids between D.
pseudoobscura and D. persimilis across this ~12.5 Mb inversion through a direct assay. I
also indirectly estimate gene exchange in natural populations along the same inverted
region using a series of markers sequenced in several inbred lines of each species.
I also examined the effects of heterozygosity for inversions on hybrid
recombination rates elsewhere in the genome. To this end, I quantified restricted
recombination at the boundaries of inversions by estimating the amount of
recombination between markers outside, but near, the breakpoints of each inversion.
These regions have higher rates of divergence, similar to markers inside inverted regions
(Kulathinal et al. 2009; Machado et al. 2007; Noor et al. 2007), suggesting they
47
experience less recombination than neighboring regions. Finally, I calculate the effect
inversions have on enhancing recombination rates throughout the rest of the genome by
comparing recombination rates in hybrids to a published recombination map of D.
persimilis (Stevison and Noor 2010).
3.1.1 Expected rates of exchange within inverted regions based on
within species inversion polymorphisms
The major prerequisite of modern chromosomal speciation models (see Faria and
Navarro 2010 for recent review) is that crossover products are rarely recovered between
inverted segments. The rarity of this occurrence is the reason that commonly used
balancer stocks are so effective in their maintenance of homozygous lethal markers (see
section “Unique features of recombination in Drosophila” in Chapter 1). In Drosophila,
single crossovers between heterokaryotypes of a paracentric inversion result in nonviable gametic products, which are subsequently shunted to the polar bodies prior to
oviposition (Carson 1946; Coyne et al. 1993; Navarro et al. 1997; Sturtevant and Beadle
1936). Therefore, exchange between heterokaryotype regions is likely due to either gene
conversion or double-crossover events. The rate of gene conversion is often assumed in
models to be uniform across an inversion, whereas the rate of double-crossover is
expected to be highest near the center, due to the role of crossover interference
(Andolfatto et al. 2001; Navarro et al. 1997). Crossover interference is the impact one
crossover event has on the probability that another crossover event will occur nearby,
determining the optimal distance between crossover events. Therefore, if two crossover
events occur within the inversion loop created at meiosis, they are more likely to span
the center of the inversion than any segments closer to the inversion breakpoints
(Navarro et al. 1997). Due to the strong effect of interference acting over distances as
long as 8-10 Mb in Drosophila (Fitzpatrick et al. 2009; Foss et al. 1993; Stevison and
48
Noor 2010; Weinstein 1918), only inversions with large recombinational distances are
expected to achieve an observable rate of double-crossover (>20 cM, Navarro et al.
1997).
Empirical analysis on exchange in inversion heterozygotes has followed two
approaches: controlled crosses to measure the rate of double-crossover and sequence
analysis of markers within an inversion to infer historical exchange. Empirical estimates
between segregating inversions in Drosophila have observed between 10-3-10-5 double
crossovers in a single generation with an average of 10-4 for a simple inversion (Ishii and
Charlesworth 1977; Levine 1956). These studies used phenotypic mutants located near
the center and ends of inversions spanning 30-80 cM (Ishii and Charlesworth 1977). In
comparison, the double-crossover rate across similar regions in homokaryotypes is
greater than 10% (Gruneberg 1935; Novitski and Braver 1954; Robbins 1974; Spurway
and Philip 1952), showing that double-recombination products are not produced as
frequently in inversion heterozygotes. Further, these observed rates of double-crossover
in inversion heterokaryotypes approach the expected genomic rate of gene conversion
(~10-5-10-6), estimated using the rosy locus in Drosophila (Chovnick 1973; Smith et al.
1970). Sequencing approaches have enhanced our understanding of differentiation
within inverted regions by being able to detect historical gene flux (Hasson and Eanes
1996; Laayouni et al. 2003; Nobrega et al. 2008; Pegueroles et al. 2010; Schaeffer and
Anderson 2005; Wesley and Eanes 1994; White et al. 2009) due to conversion or doublecrossover events within an inverted region. These studies confirm predictions by
theoretical models for flux rates to be highest near the center of inverted regions (Feder
and Nosil 2009; Kirkpatrick and Barton 2006; Navarro et al. 1997).
Despite knowing the rates of double-crossovers in inversions differentiating
populations of Drosophila, there are few, if any studies examining how inversions that
49
differentiate species might differ in rates of exchange from inversions segregating within
species. There are three reasons to expect exchange rates might differ in interspecies
hybrids: (1) inversion heterozygotes segregating within populations are more abundant
than those between species, which are dependent on the rate of migration between
species and premating isolation, (2) mechanistically, levels of sequence divergence
influence rates of crossover, predicting lower levels of exchange when sequence
divergence is higher between homologous chromosomes (Modrich and Lahue 1996), and
(3) inversion heterozygotes between species will likely suffer some fitness consequences
because of alleles conferring differential adaptation or incompatibilities within them,
allowing selection to influence the detectable rate of historical exchange (Kirkpatrick and
Barton 2006; Noor et al. 2001). Although this latter difference may not influence the rate
observed in a single generation cross, it may impact the observed exchange rate in a
sequence-based analysis. For this reason, I used both direct cross analysis and sequencebased approaches to detect exchange within the inverted region of XR.
3.1.2 Recombination suppression of inversions extends beyond
breakpoints
Markers outside inverted regions, but near the breakpoints, tend to have
heightened divergence similar to markers found within inversions (Kulathinal et al.
2009; Machado et al. 2007; Noor et al. 2007). This effect is consistent with the
observation of persistent recombination suppression in inversion heterozygotes outside
the inversion, near boundary regions (hereafter inversion boundaries) (Dobzhansky and
Epling 1948; Kulathinal et al. 2009; Maynard Smith and Maynard Smith 1954; OrtizBarrientos et al. 2006; Pegueroles et al. 2010; Spurway and Philip 1952). This pattern is
not unique to Drosophila (Roberts 1976; Strasburg et al. 2009; Stump et al. 2007; White
50
and Morley 1955), and similar effects of extended divergence at centromeric boundaries
suggest this pattern is not unique to inversion boundaries (Carneiro et al. 2010).
Using hybrids between D. pseudoobscura and D. persimilis, recombination
suppression at inversion boundaries was empirically determined to extend between 2-4
Mb (Kulathinal et al. 2009). However, the pattern of heightened divergence between D.
pseudoobscura and D. persimilis was not consistent at all inversion boundaries,
suggesting recombination suppression may be variable. In the D. pseudoobscura – D.
persimilis system, some inversion breakpoints are near chromosome ends where
interchromosomal effect should be strong (see “Inversions as global recombination
modifiers” section below), perhaps reducing the extent of recombination suppression at
these inversion boundaries. My systematic analysis of recombination rate at inversion
boundaries aims to refine the previous estimate from Kulathinal et al. (2009) for the
extent of recombination suppression at inversion boundaries.
3.1.3 Inversions as global recombination modifiers
In addition to their unequivocal effects on species maintenance, chromosomal
inversions have been shown to increase recombination rates significantly throughout the
rest of the genome (Schultz and Redfield 1951; Sturtevant 1919). Inversions have been
shown to alter crossover rates on the same chromosome (intrachromosomal effect)
and/or on other chromosomes (interchromosomal effect). Early studies made two
important observations with regard to inter/intrachromosomal effect (collectively
hereafter ICE): (1) ICE has a stronger impact on proximal (centromeric) and distal
(telomeric) portions of chromosomes, and (2) due to its impact on interference, ICE
seems to impact the preconditions necessary for exchange, rather than influencing
whether the process of recombination will result in a crossover or not (Lucchesi and
Suzuki 1968).
51
The first observation comes as the result of several studies on inversions
segregating within Drosophila melanogaster (Krimbas and Powell 1992; Lucchesi and
Suzuki 1968; Portin and Rantanen 1990; Schultz and Redfield 1951), and thus is
surprising because these regions (chromosome ends and centromeres) tend to have more
restricted recombination within homokaryotypes. The observation of heightened effect
on chromosome ends/centromeric regions has also been observed in maize and
grasshoppers (Lucchesi and Suzuki 1968); however, restricted recombination in these
regions is not universal. For example, while in Drosophila both centromeric and
telomeric regions tend to have reduced recombination (Begun et al. 2007), in mammals,
telomeric regions tend to have higher than average recombination (Kong et al. 2002).
The second observation is based on data showing coupled changes between ICE and
levels of interference, however recent studies have identified pch2 (a pachytene
checkpoint protein) as responsible for ICE by stalling the pachytene checkpoint in
inversion heterozygotes likely allowing more double-strand breaks to be resolved as
crossovers instead of non-crossovers (Joyce and McKim 2010).
Despite renewed interest in inversions and their importance in speciation (see
recent reviews: Brown and O'Neill 2010; Faria and Navarro 2010; Hoffmann and
Rieseberg 2008), advances in fine-scale recombination mapping technology, and the
observation of extensive variation in recombination rate within Drosophila (Cirulli et al.
2007; Kulathinal et al. 2008; Stevison and Noor 2010), there has been relatively little
research done on the role of ICE in contributing to nucleotide variation. Here, I map
recombination rates on the 2nd chromosome in single-generation hybrids (heterozygous
for 3 inversions) and compare these measures to the fine-scale homokaryotype
recombination map on the same chromosome recently published in D. persimilis
(Stevison and Noor 2010, see Chapter 2).
52
3.2 Methods
3.2.1 Single generation estimate of inversion crossover rate in
hybrids
Genome lines of Drosophila pseudoobscura, Mesa-Verde, CO 2-25 (MV2-25, San
Diego stock number #14011-0121.94, Richards et al. 2005) and D. persimilis, Mount
Saint Helena 3 (MSH3, San Diego stock number #14011-0111.49, Clark et al. 2007), were
crossed to generate heterozygous F1 hybrid females. These females were backcrossed to
MV2-25 males to generate approximately 10,000 progeny to screen for crossovers along
the XR inversion.
Using microsatellite or indel markers at the breakpoints and center of the XR
inversion (spanning 12.5 Mb), I tested for double crossovers in F1 hybrids between D.
pseudoobscura and D. persimilis. Chromosome arm XR differs by a single inversion
between D. pseudoobscura and D. persimilis, and the breakpoints of this inversion have
been mapped (Bhutkar et al. 2008; Noor et al. 2007). The following markers were used
to assay interspecies double crossover rate inside the inversion –DPS X063 (XR_group6:
12,588,339, center of inversion), Centromeric breakpoint, CBP (XR_group6: 6,219,093)
and Telomeric breakpoint, TBP (XR_group8: 7,199,440). Primer sequences are available
upon request. All individuals were genotyped at the center of the inversion and at least
one of the breakpoint markers. Double crossovers were defined as individuals with an
allele in the center of the inversion that was different from the alleles at the two
breakpoints, indicating exchange within the inverted segment.
3.2.2 Population genetics analysis of interspecies migration rates
within inversion
Samples: Twelve stocks from D. pseudoobscura, eight stocks of D. persimilis,
and three stocks of D. miranda were assayed (Stock names: D. pseudoobscura: Mesa
Verde, CO 2-25 [genome line], Mesa Verde, CO 17 [collected 2001], Flagstaff, AZ 16
53
[inbred 15 generations (collected 1997)], Flagstaff 1993 [cross of four lines collected
1993]; Mather, CA lines 17, 32, 48, and 52 [all collected 1997], Baja, Mexico 1 [collected
2001], Sonora, Mexico 3 [collected 2001]; Zapotitlan, Puebla, Mexico [collected 2000],
American Fork Canyon 12; D. persimilis: Mt. St. Helena 3 [genome line, collected 1997],
Mt. St. Helena 1993 [collected 1993], Mather, CA lines 37, 40, and G; Mt. St. Helena lines
1, 7, and 42; D. miranda: Mt. St. Helena lines 22 and 38 and Mather, CA 28).
Sequencing: Seven markers were designed and sequenced in individuals from
stocks listed above (Marker names: MidPoint, MP (XR_group6: 12,536,933), Telomeric
Breakpoint, TBP (XR_group8: 7,199,934), Quarter 1, Q1 (XR_group6: 9,546,799),
Quarter 2, Q2 (XR_group8: 2,402,917), Near-TBP (XR_group8: 6,429,839), XR6_2.5
(XR_group6: 2,493,745) and Far Out (XL_group1a: 4,181,765)). Markers were chosen
based on the published D. pseudoobscura genome sequence assembly (Richards et al.
2005) to be ~750bp long within intergenic regions at varying distances along the XR
chromosome arm. Each locus was amplified from genomic DNA template in 25 µL
reaction volumes using a touch-down thermal cycling profile consisting of an initial
denaturation (95°C for 2 min) followed by 12 cycles of a 95°C denaturation for 30 s, a
62°C annealing for 30 s reduced by 1°C/cycle, and a 72°C elongation for 1 min, followed
by 25 cycles at an annealing temperature of 50°C, and a final elongation at 72°C for
2min. PCR products were cleaned using ExoSAP-IT (USB Corporation). Sequencing was
performed at Duke University's IGSP core sequencing facility on Applied Biosystems
3730xl DNA Analyzer. DNA sequence chromatograms were aligned and edited in
Lasergene SeqMan 7.0 (DNASTAR, Inc.). Custom Perl scripts were then used to export
sequences and create consensus alignments for importing into other software platforms.
54
Migration estimates and other sequence analysis: In addition to the seven
markers designed for sequencing, four markers were selected from GenBank to
supplement additional analysis. Statistical analysis of each marker including FST,
diversity within each species, and DXY was performed in DnaSp 5.0 (Rozas et al. 2003).
Additionally, a coalescent-based migration rate for each marker was estimated using IM
(Hey and Nielsen 2004). Sequences were processed to meet restrictions/ assumptions of
IM by removing any gaps in the alignments and accounting for intragenic recombination
for each marker as in Stevison and Kohn (2009). Input files were executed in IMa using
10 Metropolis-coupled chains with a 300 000-step burn-in followed by 50 million
iterations of the Markov chain. Autocorrelation values, effective sample sizes (ESS), and
inspection of parameter trend plots by eye indicated adequate convergence of the
Markov chain (not shown).
3.2.3 High-throughput genotyping to analyze recombination rate
changes throughout the genome outside inversions
A subset of 480 individuals from the cross described in “Single generation
estimate of inversion crossover rate in hybrids” were genotyped to assay recombination
rate changes relative to the within species rate outside of the inversion due to the
interchromosomal effect and at inversion boundaries. This cross design did not account
for direction of the cross because a previous study in this system showed no difference in
ICE between F1 hybrids backcrossed to either D. persimilis or D. pseudoobscura (OrtizBarrientos et al. 2006).
DNA was isolated from the 480 backcross offspring individually at the Genomic
Sciences Lab at North Carolina State University for subsequent genotyping with 96
single nucleotide polymorphisms (SNPs) (see methods and scripts described in Chapter
2: Stevison and Noor (2010). A subset of 42 markers corresponds to positions on the
55
second chromosome of the within-species recombination map generated previously for
D. persimilis (Stevison and Noor 2010). These markers were designed to assess the
pervasiveness of the interchromosomal effect on this chromosome due to inversion
heterozygosity in the F1 females of this cross. This chromosome was targeted due to the
fine-scale nature of the existing within-species map (markers apx. 150-200kb apart)
relative to other chromosomes.
Another subset of 48 markers was designed to further fine-map recombination
reduction at inversion boundaries as in Kulathinal et al. (2009). These markers
correspond to 2-3 Mb outside the breakpoints of each of the three major chromosomal
inversions differentiating D. pseudoobscura and D. persimilis (six boundary regions
total). Of these 48 markers designed for inversion boundaries, there were 7 on each XR
boundary, 6 on each XL boundary, 6 on each Chromosome 2 (C2) boundary, 2 within the
XR inversion (plus 2 additional markers designed as genotyping replicates at the center
of the inversion, see below), 4 within the XL inversion and 4 within the C2 inversion.
Finally, six markers were designed to duplicate the genotyping markers used in
the inversion recombination survey of XR, so as to include this subset of individuals in
the data for the original cross (See: “Single generation estimate of inversion crossover
rate in hybrids” section above).
SNP markers were screened in all offspring using the Illumina BeadXpress
platform (Fan et al. 2003) (Illumina, Inc. San Diego, CA) at the Genomic Analysis
Facility within the Duke University Center for Human Genome Variation. The output
consisted of raw genotypes at all markers for all individuals. Eleven total markers were
not useful for analysis for various reasons such as polymorphism within strains or
monomorphism between strains. Further, eight individuals did not amplify at any of the
markers and two individuals were removed from two different subsets due to greater
56
than 10% missing data in that subset of markers. One individual was removed from the
C2 assembled map and another individual was removed from the XL boundary map
prior to assembly.
The raw data was processed to assess crossovers at each interval via scripts from
Stevison and Noor (2010). For C2, a Kosambi cM/Mb value was calculated and
compared to the published rate in D. persimilis (Stevison and Noor 2010) to calculate a
fold change difference between the hybrid map and the pure species map (Figure 3.1).
Previous studies comparing recombination rates between D. pseudoobscura and D.
persimilis have noted very tight correspondence (Stevison and Noor 2010), therefore I
limited my pure species comparison to the D. persimilis recombination rates.
3.3 Results
3.3.1 Single generation estimate of inversion crossover rate in
hybrids
Out of 9,739 individuals screened, 1 male sample was confirmed as a double
recombinant across the XR inversion, with the D. pseudoobscura allele at the center of
the inversion and the D. persimilis allele at both breakpoints. Confirmation consisted of
first repeating the initial genotyping at all three markers spanning the XR inversion.
Next, a gene conversion event was ruled out by genotyping the sample at a marker
~360kb from DPSX063, DPSX051 (chromosome scaffold XR_group6: position
12,953,124), much larger than gene conversion tract lengths in Drosophila (Hilliker et al.
1994), and confirming the D. pseudoobscura alleles extended to the adjacent marker.
Finally, the sample was confirmed via Sanger sequencing at markers nearby the three
genotyping markers, again showing a mismatch between allele at the center relative to
the breakpoints. This finding corresponds to a rate of 0.01% or ~10-4 double crossingover across this 12.Mb inversion in a single generation, though this figure may
57
underestimate the total double crossover rate since only a single position within the
inversion was surveyed.
3.3.2 Population genetics analysis of interspecies migration rates
within inversion
Sequencing seven loci together with four markers from GenBank yielded apx. 8
kb of sequence, 416 SNPs, 52 sites fixed between species, and 16 sites shared between
species. Earlier studies have noted lower diversity within the inverted segments relative
to collinear regions for D. persimilis, consistent with a loss of diversity with the fixation
of the novel chromosome arrangement (Hey and Nielsen 2004; Machado et al. 2007).
With the limited number of regions surveyed in the collinear region of XR (N=3, Table
3.1), I was unable to detect any significant difference in diversity between these regions
and those within the inverted segment (N=5, Table 3.1) (p=0.83).
Further, I followed up on previous work in this system, which has hinted at some
increased gene flow near the center of the smaller 2nd chromosome inversion (not
significant, Machado et al. 2007). Given this observation, I expect the relationship
between gene flow and distance from breakpoint to be stronger inside the larger XR
inversion. I tested for a decline in population differentiation with increasing distance
from the inversion breakpoints, restricting the analysis to markers inside the inverted
region (p=0.3; N=5; r2=0.341). I also looked for a decrease in divergence and did not
observe a significant trend (p=0.38; N=5; r2=0.26). Conversely, I looked for an increase
in estimated gene flow (Nm) and still did not observe a significant trend (p=0.2; N=5;
r2=0.47).
To further test for a positive relationship between gene flow and distance from
the inversion breakpoints, I estimated a locus-specific migration parameter
implemented in the software IM. Except one locus (XR6:2.5, outside inversion), all
58
markers had migration estimates not significantly different from zero. This result
precluded the option of examining how variation in migration correlates to distance from
inversion breakpoints. The single marker to show non-zero migration also has the
highest level of detectable diversity within D. persimilis (see Table 3.1). Further, the
locus-specific migration estimates of X009 and X010 corresponded with published
migration estimates for these markers (Hey and Nielsen 2004).
59
60
In
Near
Far
TBP
(XR8:7.2)
Near TBP
(XR8:6.43)
X0102
824
606
592
680
621
669
297
677
1798
612
638
Length
20
11
12
12
12
12
9
18
11
11
11
N
15
10
19
3
16
22
92
38
32
32
7
S
0.0024
0.0046
0.0071
0.0012
0.0077
0.0107
0.0685
0.0146
0.0044
0.0153
0.0047
π
D. pseudoobscura
14
7
7
8
7
7
5
14
11
8
8
N
9
11
5
2
11
6
17
36
10
19
4
S
0.0016
0.0061
0.0032
0.0007
0.0064
0.0030
0.0229
0.0158
0.0013
0.0104
0.0024
π
D. persimilis
0.010
0.017
0.009
0.006
0.017
0.018
0.097
0.027
0.007
0.025
0.014
DXY
0.50
0.45
0.39
0.48
0.18
0.33
0.43
0.38
0.15
0.32
0.52
FST
0.60%
1.49%
0.00%
0.44%
0.81%
0.75%
2.69%
0.15%
0.39%
0.82%
0.63%
Fixed
0.00%
0.00%
0.00%
0.00%
0.48%
0.00%
0.67%
0.89%
0.06%
0.65%
0.00%
Shared
0.25
0.31
0.39
0.27
1.15
0.5
0.33
0.42
1.43
0.53
0.23
Nm
0.0076
0.0076
0.0076
0.0076
0.0076
0.0076
0.0076
0.0076
0.0076
9.3068
0.0076
m1
0.0056
0.0056
0.0169
0.0056
0.0056
0.0056
0.0056
0.0056
0.0056
0.0395
0.0056
m2
IM Results
2MACHADO,
1WANG,
R. L., J. WAKELEY and J. HEY, 1997 Gene flow and natural selection in the origin of Drosophila pseudoobscura and close relatives. Genetics 147: 1091-1106.
C. A., R. M. KLIMAN, J. A. MARKERT and J. HEY, 2002 Inferring the history of speciation from multilocus DNA sequence data: The case of Drosophila
pseudoobscura and close relatives. Molecular Biology and Evolution 19: 472-488. 3MACHADO, C. A., T. S. HASELKORN and M. A. F. NOOR, 2007 Evaluation of the genomic
extent of effects of fixed inversion differences on intraspecific variation and interspecific gene flow in Drosophila pseudoobscura and D. persimilis. Genetics 175: 1289-1306.
3.5
0.77
0
4.42
In
Q3 (XR8:2.4)
3.97
0.78
6.4
In
X_21023
1.22
In
Near
X0092
2.1
MP
(XR6:12.54)
Near
Hsp821
3.67
In
Far
XR6:2.5
7.69
Q1
(XR6:9.55)
Far
Far Out
(XL1a:4.2)
Location
Distance
to break
point
(Mb)
Table 3.1. Summary of diversity and divergence calculations for each of the 11 markers along the XR chromosome
arm.
3.3.3 Analysis of recombination rate reduction in single-generation
hybrids at inversion boundaries
Genotypes of markers at inversion boundaries and the markers within each
inversion were compared for each individual to determine how far complete suppression
of recombination extends outside the inversion breakpoints. At the C2 centromeric
boundary, no recombinants were recovered as far as 2.4 Mb, and as far as 2.56 Mb on the
telomeric boundary. At the XR centromeric boundary, no recombinants were recovered
as far as 2.44 Mb and on the telomeric boundary, no recombinants were recovered as far
as 2.79 Mb. At the XL boundaries, there was an absence of observed recombination as far
3 Mb on the centromeric side of the inversion and no recombinants were recovered on
the telomeric side of the inversion as far as 2.73 Mb.
The range between minimum recombination suppression and the maximums
presented above are summarized in Table 3.2 and compared to previous results from
Kulathinal et al. (2009). The results from the current study used more markers to fine
map both the lower and upper limit of this range. The third column represents the
narrowest range as summarized from both studies.
Table 3.2. Summary of recombination suppression at inversion boundaries
as compared to Kulathinal et al. 2009. The refined range combines the
results of both studies.
Kulathinal et al. 2009
Current study
Refined range
Centromeric side XL
0 - 2.84 Mb
2.75 - 3 Mb
2.75 – 2.84 Mb
Telomeric side XL
0.4 - 2.8 Mb
2.73 Mb +
2.73 – 2.8 Mb
Centromeric side XR
0 - 3.35 Mb
2.25 - 2.51 Mb
2.25 - 2.51 Mb
Telomeric side XR
2.1 - 2.8 Mb
2.79 Mb +
2.79 – 2.8 Mb
Centromeric side C2
0 - 4.55 Mb
2.33 - 2.45 Mb
2.33 - 2.45 Mb
Telomeric side C2
0 - 2.77 Mb
2.27 - 2.51 Mb
2.27 - 2.51 Mb
61
3.3.4 Analysis of interchromosomal effect using high-throughput
genotyping
The recombination rate of hybrids along C2 ranged from 0 (within inverted
regions) to 18.92 cM/Mb with an average interval size of 500kb. The comparable
recombination rate in the same intervals for pure species ranged from 0.15 - 20.54
cM/Mb. Although the range of recombination rate values was similar, the distribution of
these events was significantly different in the hybrids in 31 out 47 intervals. The ratio of
cM/Mb in hybrids to cM/Mb in pure species (D. persimilis) was the calculated fold
change, which ranged from 0 to 9.14. The fold change is highest (>2 fold) in the first 4
Mb and the last 5.5 Mb of the chromosome. Figure 3.1 shows a plot of the log (base 10)normalized fold-change of recombination rate in hybrids relative to D. persimilis.
Because previous studies on ICE have also observed large effects in distal and
proximal regions (which normally have low recombination in Drosophila), I tested for an
association between recombination rate with D. persimilis and the fold change observed
in hybrids. I found that regions with low recombination rate in D. persimilis were among
those with the highest log fold change in hybrids (p<0.0001; r2=0.373; N=36),
suggesting these regions are most susceptible to changes in recombination. I excluded
regions with no observed recombination in hybrids.
Next, I tested for an association of nucleotide divergence between species and
hybrid recombination rate. I calculated divergence between D. pseudoobscura and D.
persimilis at 4-fold degenerate 3rd position codons (C4). I further calculated a
corresponding diversity measure along the same regions between two sequences of D.
pseudoobscura at C4 sites, to correct for variation in levels of diversity contributing to
variation in divergence. In a multiple regression analysis with log-fold change as the
response variable and the diversity and divergence calculations as predictor variables, I
62
showed that divergence was a significant predictor of variation in log fold change
(p=0.0088; N=31), but that intraspecies diversity did not explain a significant portion of
the variance in log fold change (p=0.0831). This result suggests that the correlation
observed between divergence and log fold change was not driven by lower segregating
ancestral diversity in chromosomal ends (Noor and Bennett 2009). In the above
analysis, I excluded regions with no recombination in hybrids, one interval with no C4
bases for the diversity measure, and the last 4 intervals of the chromosome which are
misassembled in the D. pseudoobscura genome making it difficult to obtain reliable
divergence estimates. When I further reduced the analysis to exclude chromosome ends,
the results remained the same with a significant result for divergence (p=0.0049; N=18),
but not for diversity (p=0.0569), despite the loss of variation in the log-fold change
variation as ICE affects the ends most strongly.
Figure 3.1. Inter-chromosomal Effect. A plot of the log (base 10) fold
change difference in recombination rates along the 2nd chromosome
between the published D. persimilis map and the map of recombination rate
in between species hybrids.
63
3.4 Discussion
3.4.1 Exchange across XR inversion detected in one generation,
however prolonged evidence of exchange not observed
The observed rate of exchange via double crossovers along the largest inversion
(XR inversion, 12.5 Mb) differentiating D. pseudoobscura and D. persimilis was
empirically determined to be of the same order of magnitude as rates of exchange across
inversions segregating within species (Ishii and Charlesworth 1977; Levine 1956), but
much lower than recombination in collinear regions greater than ~2.5Mb outside of the
inversion or within the same region in homokaryotypes (Kulathinal et al. 2008; Stevison
and Noor 2010). My estimate of 10-4 is much higher than estimates previously used in
models examining the role of inversions in persistence of species with gene flow (Feder
and Nosil 2009; Kirkpatrick and Barton 2006; Navarro et al. 1997).
Here, I estimate much higher rates of exchange between inversions empirically,
and yet I observe no evidence of prolonged interspecies gene flow within the inversion
using population level sequencing at evenly spaced markers within the inverted segment.
My results suggest that models determining the role of inversions in maintaining species
should consider independently the expected frequency of heterokaryotes as a function of
the expected level of migration between species and any potential fitness consequences
that prevent the success of progeny that carry exchange products from propagating in
the next generation (see also Faria and Navarro 2010; Jackson in press). Further,
because the accumulation of genetic incompatibilities occurs gradually over time,
selection does not need to be very strong in early generations when migration is nearly
zero. However, upon secondary contact, higher rates of migration, and thus selection are
likely, indicating that these parameters should be considered non-independent. These
three factors – the frequency of heterokaryotypes/migrants, the fitness consequences of
64
exchange in heterokaryotypes, and the rate at which genetic incompatibilities
accumulate – are likely the most important factors contributing to the absence of longterm exchange detected between D. pseudoobscura and D. persimilis along the inverted
region on the XR chromosome arm.
3.4.2 Recombination suppression extends 2.5-3 Mb at inversion
boundaries
One of the more puzzling phenomena associated with inversion heterozygosity is
the extension of recombination suppression beyond inversions, outside the breakpoints.
Here, I refined the known ranges of recombination suppression at inversion boundaries
for the three major inversions differentiating D. pseudoobscura and D. persimilis from
an average of 2.76 Mb between markers observing no recombination and the first marker
to detect any recombination (Kulathinal et al. 2009) to an average of 132 kb between
markers (Table 3.2).
The results in Table 3.2 clearly show that the level of recombination suppression
due to inversion heterozygosity cumulatively extends 2.5-3 Mb beyond the breakpoints
of each inversion, adding 5-6 Mb to the total expected size of the inversion itself. Because
recombination is not suppressed in these boundary regions in the absence of an
inversion (Stevison and Noor 2010), it is not immediately obvious how recombination
suppression would extend this far outside of the inversion loop. Previous studies have
accredited this result to the difficulty of the synaptonemal complex from forming at
inversion boundaries (Roberts 1976). Another possibility is that because inversion
boundaries tend to accumulate in unstable/repetitive regions of chromosomes
(Andolfatto et al. 1999; Caceres et al. 1999; Ranz et al. 2007), inversion boundaries serve
to recruit recombination suppression over the length of the inversion, triggered perhaps
by heterozygosity immediately outside the inversion boundary. If inversion boundaries
65
were indeed the molecular trigger for reduced recombination inside inverted regions, it
would follow that these regions are also susceptible to recombination suppression.
3.4.3 Interchromosomal effect highest in regions of low
recombination along chromosome 2
My study investigates both ‘intra’- and ‘interchromosomal effect’ using ~50
markers along chromosome 2. Because I analyze hybrids between D. pseudoobscura
and D. persimilis, the ICE I observe is due to heterozygosity at three inversions – two on
the X and one on the 2nd (focal chromosome). Previous studies have observed ICE
yielding differences in recombination rates as high as 250% higher than standard map
distance (Schultz and Redfield 1951), whereas I observed greater than 800% higher
recombination rates with an average of 224%.
The higher proportional increase in recombination rate in this study relative to
earlier studies is likely due to heterozygosity for more inversions (other studies observe
ICE with only 1-2 heterozygous inversions) and/or the scale at which I analyzed
recombination in hybrids. Averaging over larger intervals may have masked the effects of
smaller regions with very strong ICE in previous studies. However, the average percent
change in the first and last 5 Mb is ~250% higher recombination in inversion
heterozygotes, suggesting that the size of intervals assayed in the current study is most
likely responsible for the higher observed change in recombination rates.
Similar to previously published results, I found that interchromosomal effect
(ICE) was strongest in regions distal and proximal on chromosome 2. Because these
regions are known to have lower recombination rates overall, I showed that, irrespective
of chromosome position, regions of low recombination were most susceptible to ICE,
supporting that these regions may be less resilient to disruptions in recombination rate.
66
Finally, I examined how changes in recombination rates in hybrids correspond to
patterns of gene flow and differentiation, as an extension of how inverted regions (which
have low recombination in hybrids) often bear high divergence between species. To test
for this, I calculated divergence between D. pseudoobscura and D. persimilis and found
a strong correlation with log-fold change of recombination in hybrids relative to pure
species. When I corrected for regions of low within species diversity (e.g., chromosome
ends), I still observed a significant association, showing that low diversity at
chromosomal ends (where ICE is strongest) does not influence the relationship between
divergence and recombination rate changes in hybrids. Hence, it appears that the
interchromosomal effect may actually increase interspecies gene flow outside of the
inverted regions- a factor not considered previously in the effects of inversions on species
persistence.
In the past ten years, interest in the role of chromosomal inversions in speciation
has been rekindled based on the inherent properties of inversions to restrict
recombination in heterokaryotypes. This reduced recombination protects existing
adaptive complexes and genetic incompatibilities and allows for the accumulation of
additional incompatibilities between species. The research presented here confirms that
some features found in inversions segregating within species apply to interspecies
inversion differences, but also identifies potential differences in long-term effects of
inversions differentiating species and hypothesizes their causes. Further research should
explore some of the patterns suggested here, particularly considering 1) what factors may
maintain divergence in inverted regions between species despite detectable exchange
due to double crossover, 2) the relationship between size of recombination intervals
67
assayed on the intensity of inter- and intrachromosomal effects (ICE) observed, and 3)
how much ICE increases interspecies gene exchange outside inverted regions.
68
4. Male-mediated effects on female recombination
4.1 Introduction
One of the major evolutionary advantages to meiotic recombination, or ‘crossing
over’, is the combinatorial effect of shuffling beneficial alleles onto a common genetic
background (Crow and Kimura 1970; Fisher 1930), favoring their fixation, and
deleterious alleles onto separate genetic backgrounds, facilitating their elimination
(Felsenstein 1974; Muller 1964). Mechanistically, meiosis requires crossing over, which
is proposed to stabilize chromosomes in metaphase (Kucherlapati and Smith 1988).
However, there is an upper limit to the level of crossing over that is favorable. For
example, too much recombination can lead to non-disjunction in meiosis and therefore
be unfavorable due to offspring with too many or fewer than necessary chromosomes
(Roeder 1997). There is also an evolutionary upper limit to how much recombination is
selectively advantageous. Although recombination can combine the advantages of two
beneficial mutations by putting them on a common genetic background, it can also break
them apart. This separation of selectively advantageous alleles can cause ‘interference’
between these alleles, reducing the selective advantage of the allelic combinations and
preventing the population from reaching the optimal genotypic combinations (Comeron
et al. 2008; Hill and Robertson 1966) and see section “Evolutionary Effects of
Recombination on Structure of the Genome” of Chapter 1.
Despite the evolutionary importance of recombination, recombination rates vary
drastically between species, ranging from no recombination in asexual systems to very
high levels of allelic exchange across the genomes of some sexual systems. Further,
within any specific genome, meiotic recombination rates are also variable, such that
some regions have higher rates of crossing over than others. Additionally, recombination
69
within the genome can vary based on genotype. There are also condition-dependent
features that lead to changes in recombination within an individual’s lifetime, such as
variation with maternal age whereby eggs laid later in life exhibit higher numbers of
crossovers than eggs laid when females are younger (Bridges 1927; Bridges 1929).
Attempts to examine plasticity in recombination rates have demonstrated that factors
such as temperature (Plough 1917; Plough 1921), nutrition (Neel 1941), age of mating
(Redfield 1966), and the number of matings (Priest et al. 2007) also affect recombination
rates (for full summary, see section “Variation in Crossing Over” of Chapter 1).
One of the more surprising patterns of plasticity in recombination rates is the
change elicited in response to mating frequency (Priest et al. 2007; Redfield 1966). An
early study on the effects of age on recombination rates focused on the effects of age at
the time of mating, comparing recombination rates of females mated at different time
points after emergence (Redfield 1966). These results indicated that recombination rates
shortly after mating are generally high for up to 2 days (Redfield 1966). Next, the results
showed a rapid decline 3-4 days post mating, followed by a recovery period of at least 6-8
days (Figure 4.1). Female age at mating significantly influenced both the time for
recombination rates to decline and the recovery period, such that both were longer for
females mated on emergence than for older females. These results were the first
demonstration that the act of mating may lead to variation in recombination rates.
70
Figure 4.1 Results of Redfield experiment showing changes in postmating recombination rates between markers Stubble (Sb–3R:11.9Mb) and
scarlet (st–3L:16.5Mb) on the 3rd chromosome (Modified from Redfield
1966).
Priest et al. (2007) followed Redfield’s work with a series of experiments to test
the effects of maternal age and number of matings on recombination rate of offspring
collected. They confirmed that maternal age impacts recombination rate. Furthermore,
they demonstrated that females mated twice (once early and once late) showed no
difference in recombination rate from females mated late only, with eggs for both
females collected at the same age. This observation suggests that a twice-mated female
would exhibit patterns of gametic recombination resembling the patterns in Figure 4.1.
Though this result suggested that mating history does not affect recombination rate, they
then manipulated the number of matings, where females experienced low, medium or
high levels of mating, and compared recombination rate based on levels of mating. The
offspring of females in the high and medium treatments did not exhibit significantly
different levels of recombination. However, recombination rates observed in both of
these treatments differed significantly from the recombination rate of offspring from the
71
females in the low mating treatment. The combined conclusion from these experiments
is that mating itself alters the fraction of recombinant progeny produced by the female.
This effect of mating on recombination rates could result from the stress of
mating on the female in a manner analogous to temperature- or nutrition-based stress,
and/or it could indicate that the male is inducing a physiological change in the female,
resulting in a short-term change in the proportion of recombinant offspring. To
distinguish between the possible causes for differences in recombination rates in
response to mating, I present here a test for male-mediated effects on recombination.
Given that female meiosis is incomplete at mating, and males induce a suite of known
phenotypic effects in females with whom they mate, it is biologically plausible that males
could alter female recombination. My experimental test assayed the putative variation in
this novel trait among males in two ways – (1) by altering male rearing environments to
induce potential changes that vary the degree of male-mediated effects and (2) by
examining natural standing genetic variation among males in their ability to alter female
recombination rates.
Induced differences among males were imposed by altering male rearing
temperature and via treatment with Juvenile Hormone. Temperature has been
demonstrated to play a major role in physiology (Angilletta et al. 2002; Foguera et al.
2007), circadian rhythms, (Busza et al. 2007), behavior (Herrel et al. 2007), and
recombination rates, with changes in recombination rates observed at both high and low
temperature in several systems, including Drosophila (Plough 1917; Plough 1921).
Juvenile Hormone (JH) affects many aspects of insect development and physiology.
Levels of JH trigger transitions between larval stages in insects along with multiple
aspects of reproduction, stress, and immunity affected in the adult stage. Specifically, JH
72
regulates protein synthesis in the male accessory gland organ, and oocyte maturation
and behavior in the female (Flatt and Kawecki 2007). Both sources of environmental
variation could be of evolutionary importance because, unlike other factors known to
influence recombination rates, the environmental factor in this scenario would not
directly affect the individual (i.e. the female) that exhibits the variation. Thus, if
environmental manipulation of the male leads to observable differences in female
recombination rates, I can isolate the possible cause of this variation.
Further, to examine variation in standing genetic variation, stocks collected from
Zimbabwe were used due to their greater genetic variability as compared to stocks
collected from the rest of the world (Andolfatto and Przeworski 2001; Begun and
Aquadro 1993). Genetic variation in male ability to alter the proportion of recombinant
progeny could suggest a link between this trait and male reproductive success, raising
the possibility that this trait could be influenced by sexual selection. Lower
recombination rates in most Drosophila males than females is consistent with the
Haldane-Huxley rule, where, when one sex has no crossing over, that sex is often the
heterogametic sex (Haldane 1922; Huxley 1928), see section “Unique features of
recombination in Drosophila” of Chapter 1. Because natural selection requires heritable
phenotypic variation and females are the only sex undergoing recombination,
recombination rates would evolve only in response to associated fitness variation in
females. Selection could, however, favor the spread of a trait whereby males could
indirectly influence their proportion of recombinant progeny.
Due to the evolutionary significance of a male-mediated affect on female
recombination rates, it is worth noting that, although crossing over has already occurred
at the time of mating, meiosis in females is incomplete (Page and Hawley 2003). The
73
haploid gametes have not yet formed, and none of the daughter cells have been lost due
to absorption into the polar bodies (Roeder 1997). At the end of meiosis, 15 of 16
daughter cells become polar bodies (King 1970), which may lead to competition among
developing ootids (perhaps based on chiasma frequency) to become the nucleus of the
developing egg, altering the outcome of female meiosis (Ashburner and Wright 1980;
Buckler et al. 1999; Zwick et al. 1999). Chiasma are retained throughout meiosis I into
anaphase I as physical links between chromatids which underwent crossing over (Page
and Hawley 2003). This Ootid Competition (OC) model is a unique form of female
meiotic drive on the basis of chromosomal features such as chiasma frequency (Zwick et
al. 1999), and would lead to with the cumulative affect over many rounds of meiosis of
altering the frequency of recombination in the offspring. Therefore one of the potential
mechanisms whereby males could elicit the effect proposed here is by manipulation of
competition among individual ootids in developing oocytes. Further, experimental
evidence of male-mediated influences on rates of non-disjunction in females,
demonstrates that males can influence the outcome of meiosis in females, perhaps via
influencing this meiotic drive process in females (Gilliland 2003).
In many insect species, several male mediated factors have been shown to have
strong antagonistic effects on both female behavior and physiology owing to sexual
selection. Male seminal fluid proteins mediate some of these traits. These proteins are
transferred to the female during mating to elicit a variety of responses that increase male
reproductive success, often at the cost of female fitness. For example, some seminal
proteins yield drastic effects in females including decreased lifespan, altered receptivity
to mating, and changes in oogenesis as well as egg-laying rates (Ram and Wolfner 2007;
Wolfner 2002). To date, hundreds of male seminal proteins have been identified, but
74
only some have known functions. Regardless, the possibility exists that male-derived
substances act to influence female reproductive mechanisms, potentially including the
rate of recombination among the progeny.
4.2 Methods
Cross design for both experiments are outlined in Figure 4.2. For experiment 1,
stocks of Drosophila melanogaster with homozygous lethal dominant phenotypic
mutations Stubble (Sb–3R:11.9Mb) and Glued (Gl–3L:13.9Mb) were maintained in the
heterozygous condition using the 3rd chromosome balancer TM6B (developed via stocks
obtained from the Model Systems Genomic group at Duke University). Virgin females of
this stock were crossed to males from the stock OregonR (wild type) to yield genetically
identical female replicates heterozygous for the visible markers (Sb-Gl). These females
were used for subsequent mating to either control or treated males from the wild type
stock.
To vary male rearing temperature, I maintained control and treatment stocks of
wild type males at 21°C and 25°C, respectively. Virgin males collected from the treatment
group were stored at 25°C prior to mating, while control males were reared and stored at
the control temperature. To treat males with JH, the synthetic analog, methoprene
(JHa), was suspended in 95% ethanol at a concentration of 2 µg/ml. The JHa solution
was then added to standard fly media in a concentration of 1.04 µl/ml as described by
previous experimental manipulations to fruit flies by application of JHa to food (Flatt
and Kawecki 2007; Flatt et al. 2005). At this concentration, changes to males included
decreased courtship display and mating success. Further, the stocks treated with JHa
showed highly reduced rates of eclosion relative to control stocks. Care was taken to limit
variation in rearing density, which may affect rates of JHa uptake, by only allowing ~575
15 females to lay eggs for no longer than 1 week. Control males for this treatment group
were reared on food prepared the same way with 95% ethanol substituted for the JHa
solution. To remove residual hormone on the exterior of male flies, virgin males collected
from this treatment group were stored on normal media prior to mating. This precaution
ensured that any effect in females was due to male treatment and not residual hormone
that rubbed off of the male during mating. All other conditions between control and
treated males were identical.
For experiment 2, strains of D. melanogaster from Zimbabwe, Africa used were
Zim30, Zim11, Zim152, Zim6, Zim42, and Zim44 (the latter five provided courtesy of C.
Aquadro). Additionally, the strain OregonR was used to increase the variance in genetic
background among males. Virgin females from the stock Zim11 were crossed to males
from the stock Zim152 to generate genetically identical female replicates heterozygous
for zim11 and zim152 alleles. These females were used for subsequent mating to the
remaining strains. Stocks used to collect males and females were stored in identical
environmental conditions.
For both experiments, males and females were collected within eight hours of
eclosion and held virgin for four days. Males were isolated 24 hours prior to mating to
avoid crowding, which has been previously observed to influence mating rates (Noor
1997). Males and females were singly paired in vials and observed for a period of 60-90
minutes. To control for any effect of mating frequency on recombination, males were
removed once mating was complete to prevent remating (Priest et al. 2007). To control
for age effects, mated females were transferred every two days up to eight days postmating to ensure offspring in each vial were the result of eggs laid within a specific
period of time after mating (Redfield 1966). Based on a pilot experiment (see Appendix
76
A), I aimed for a target sample size of 3,000 offspring per male parent in order to have
enough power to detect a statistically significant difference in recombination rates.
Figure 4.2. Experimental Design. (A) Experiment 1: Females with
homozygous lethal dominant mutants were crossed to wild type males to
produce heterozygous females. Genetic replicate heterozygous females were
then crossed to either control or treated males from two treatment groups:
temperature (25°°) or Juvenile Hormone (JHa). (B) Experiment 2: Inbred
lines were crossed to produce heterozygous females. Genetic replicate
heterozygous females were then crossed to 1 of 5 males from different
genetic backgrounds. For both experiments, proportion of recombinant
offspring were compared based on the identity of the male parent.
4.2.1 Experiment 1 – Effect of male treatment on female
recombination rate
To assay recombination rate, progeny were visually assayed for the presence/
absence of each visible marker. Individuals that were either wild type or mutant for both
alleles were counted as parental, and individuals mutant for one allele and wild type for
the other were counted as recombinant. The final dataset was analyzed using JMP v8.0.2
statistical analysis software. I conducted a logistic regression to test the effects of day
collected relative to mating and paternal treatment, and their interaction term on the
binary response variable of whether an individual was recombinant or non-recombinant.
77
4.2.2 Experiment 2 – Effect of male genetic background on female
recombination rate
Rather than use phenotypic markers, which can have segregation distortion due
to viability differences in different genetic backgrounds (Mark and Zimmering 1977), I
used molecular markers to assay recombination rate. The microsatellite markers
X3821715gtA (X:3,856,931), DMC30B8a (X:2,111,512), dmu14395 (3L:6,691,165), and
AC004658 (3L:3,484,107) were used to distinguish alleles inherited from Zim11 versus
Zim152. Primers for AC004658 and dmu14395 were previously published, except a
modified reverse for dmu14395: 5'-CGTACTTGTATTAAGTCGAGCG-3' was used (Colson
et al. 1999). DNA was isolated according to Gloor and Engels (1992) and amplified via
touchdown cycle according to Palumbi (1996). To minimize reagent use and time,
markers were multiplexed into single PCRs with 2 primer pairs per reaction. PCR
products were separated by size on LiCor 4300 genotypers on 6% acrylamide gels.
Individuals were scored as recombinant if one allele on a chromosome was derived from
Zim11 and the other from Zim152.
Results were fitted to a logistic regression model with the following independent
variables: days post-mating (D), paternal genetic background (P), and genomic position
(G).
Rijkl = µ + Pi + Gj + Dk + PGij + PDik + GDjk + PGDijk+ εl,
for i = 1..5; j = 1..2; k = 1..4; l = 1..30,553
Genomic position was treated as a repeated measure effect by incorporating a
repeated statement in SAS (R-sided random statement in Proc Glimmix) for individual
because each individual was assayed for two pairs of markers. All other data were
analyzed using JMP v8.0.2.
78
Late into the experiment, when I decided to assay recombination rate in a third
region of the genome to determine if recombination rate differences were a global
phenomenon, I encountered difficulty in finding markers that distinguished zim11 and
zim152, particularly due to higher levels of heterozygosity of the 2nd chromosome within
each stock. On the 3rd and X chromosomes, both of these parental strains were
homozygous; however, I found that one or both strains were often heterozygous on the
2nd chromosome. Heterozygosity within the parental strains introduced potential
variation among the F1 females that must be taken into account. Therefore, I split the F1
females from this cross (see Figure 4.2B) into one of the four possible allelic
combinations based on their genotype at the marker CAD (2L: 20,759,256),
heterozygous in both zim11[A/A′] and zim152[a/a′]. This marker was chosen based on its
putative linkage to segregating inversions within the two stocks. I then entered female
genotype as a categorical parameter with four levels (F1 genotypes: Aa, Aa′, A′a, A′a′) into
the final logistic regression model (see “Results” section for Experiment 2).
4.3 Results
For experiment 1, neither male temperature treatment nor male hormone
treatment had a significant effect on female recombination rate (p=0.674 [temperature],
Figure 4.3A; p=0.355 [JHa], Figure 4.3B). However, when only considering
recombination rate of offspring collected from days four through eight post-eclosion for
the male hormone treatment, a weakly significant effect was observed (p=0.0401),
suggesting a delayed effect of this hormone on altering recombination rate in females.
Though this latter result is not significant after correction for multiple tests using
sequential Bonferroni methods.
79
I also noted that males treated with Juvenile Hormone courted less (qualitative
observation only) and had marginally significantly lower mating success than control
males (p=0.0687; Z=-1.820, Wilcoxon signed rank test). This same trend was observed
when comparing the mating success of males reared at higher temperatures relative to
control males (p=0.0759; Z=-1.775, Wilcoxon signed rank test). Despite the lack of
consistent male-mediated effects on female recombination rate between the two paternal
treatments, both experiments revealed compelling variation in recombination rates
based on maternal age via days post-mating (p=0.0019 [temperature], p=0.0036 [JHa]).
This variation is consistent with the experimental results of Redfield (1966), who assayed
nearly the same section of chromosome (Figure 4.1). Despite the sample size difference
between the hormone treated and control group shown in Figure 4.3B, these numbers
reflect variation in the total number of crosses set up, not differences in fecundity
between treatment and control vials. Indeed, there was no difference between the
numbers of progeny per vial between the control versus hormone treated crosses
(p=0.81; F=0.0596). These lower sample sizes came from unanticipated limitations in
percentage of virgin males eclosing from treated vials and percentage of successful
matings in crosses involving hormone treated males.
80
Figure 4.3. Results for experiment 1. Results of (A) male temperature
treatment and (B) male hormone treatment on female recombination rate
between markers Stubble and Glued on the 3rd chromosome (~17cM apart).
The numbers next to each data point indicate the sample size for each male
treatment group per two-day collection period and the bars represent
standard error values.
For experiment 2, I detected a significant difference in the proportion of
recombinant offspring based on paternal genetic background (p=0.0192; Figure 4.4C),
which I initially interpreted as suggesting that males vary in their ability to influence
recombination rate in their mates. Females showed up to 18% higher recombination
rates based on their mated partner (zim6 vs. zim30 in Figure 4.4C). Model effects of days
81
post-mating and genomic position were not significant (p=0.579, p=0.992), nor were
any interaction terms. The lack of an effect of days post-mating on recombination rates
was surprising given previous literature on the topic (Redfield 1966); however, the Xchromosome data analyzed by itself does exhibit a significant day effect (Figure 4.4A).
The full model most likely did not exhibit a significant day effect because the autosomal
region analyzed did not show any variation in days post-mating (Figure 4.4B). I observed
no difference in the number of progeny per vial based on paternal genetic background
(ANOVA p=0.173; N=300), suggesting the effect observed is not due to differences in
offspring viability. There was also no association between recombination rate and mating
success among the crosses (p=0.82; r2=0.021; N=5).
82
Figure 4.4. Results from Experiment 2 prior to accounting for female
variation. Recombination rate variation as a function of male parent for sexlinked markers (A) and autosomal markers (B) and a summary (C) showing
the total results of the experiment. The numbers next to each data point
refers to the number of offspring genotyped to obtain the mean and
standard error values plotted.
83
Upon identifying genetic variation among the female parents in this cross, I
genotyped the mothers to add their genotype as a parameter in the logistic regression
model. Based on genotyping within each parent strain (zim152 and zim11), I determined
that the 2nd chromosome was being inherited nearly as a single unit (data not shown).
This indicated that variation among the females is most likely due to the presence of
inversions, which are common in African lines of D. melanogaster. Therefore, I account
for the known genetic variation using a single marker located centrally along
chromosome 2. Incidentally, none of the females with genotype Aa was mated with
OregonR males (Figure 4.5); therefore, to test more accurately for an interaction effect
between male and female genotype, I excluded the ~4,000 offspring from this maternal
genotype.
The final logistic regression model, therefore, followed the equation below:
Rijklmnp = µ + Mi + Pj + V[M,P]k + Sl + Gm + Dn + MPij + εp,
for i = 1..4; j = 1..5; k = 1..200; l = 1..2; m = 1..2; n = 1..4; p = 1..22,364
where R represents the binary response variable of whether an individual was
recombinant at the markers assayed or not, M represents maternal genotype, P
represents paternal genotype, V represents F2 vial as a nested parameter within both
maternal and paternal genotype, S represents sex of each individual, G represents
genomic position assayed (X-linked markers or autosomal markers), and D represents
days post-mating. Once this model was fitted to the data, only M, maternal genotype
(p=0.0003) and V[M,P] (p=0.0253) were found to significantly influence offspring
recombination rate Figure 4.5.
84
The model parameters sex, genomic position, days post-mating, and paternal
genotype, as well as the interaction between female and male genotype did not
significantly contribute to variation in offspring recombination rate.
Figure 4.5. Results for Experiment 2 accounting for variation in female
genotype plotting recombination rate of offspring based on the genotypes of
both the male and female parents. Numbers indicate the sample size of
offspring genotyped at each category on both the 3rd and X chromosomes.
Female genotype was split based on the 4 possible allelic combinations at
the marker CAD (heterozygous in both parental stocks) on chromosome 2L
(20.76Mb).
One notable difference between Figure 4.4 and Figure 4.5 is the differences in
samples size among male genotypes in the latter figure. The uneven sample sizes of male
genotypes when split by female genotypes were the result of females not being genotyped
prior to mating. Therefore, the split of females by genotype into male treatment was
random, and likely reflects the frequency of maternal alleles segregating within the
parent stocks.
85
4.4 Discussion
The experiments described here were designed to test for male mediated effects
on recombination by examining both environmental and genetic variation among males
in their ability to influence recombination rates in females. The results of Experiment 1
showed that while male temperature treatment does not influence offspring
recombination rates, male hormone treatment may alter recombination rate of offspring
from eggs laid 5-8 days post-mating (Figure 4.3). However, differences in mating success
between control males and hormone treated males suggest differences in levels of
promiscuity among the females who mated with males treated with JHa. If there is a
physiological association between female promiscuity and recombination rate, then the
variation in recombination rates seen in the treatment group relative to the control may
not be directly attributable to the male treatment.
Given that at the time of mating, the physical act of crossing over is completed,
the only potential influence males can have on crossing over involves altering the
competition between individual chromatids. The previously proposed OC model
describes how competition might take place on the basis of chiasma frequency to become
the nuclear component of the developing egg, or the pronucleus, via female meiotic drive
(Zwick et al. 1999). Further, experimental evidence suggests that males may influence
the outcome of this competition altering the rate of non-disjunction in females (Gilliland
2003). The experimental results here outline how males potentially affect the outcome of
this competitive process, leading to variation in the frequency of chiasmate tetrads that
ultimately become the pronucleus.
The results of Experiment 2 showed that, once female genotype was considered,
male genotype did not significantly influence differences in offspring recombination
86
rates (Figure 4.5). Complications with experiment 2 resulted in unanticipated genetic
variation among the maternal parents in the cross. Although my attempt at accounting
for this variation indicates that female genotype is a strong predictor of recombination
rate, there is no guarantee that I have accounted for all possible genotypic variation
among females. A lack of male genotype on female recombination rates suggests that,
although there may be environmental variation among males in their ability to alter
female recombination rate, there is no detectable genetic variation for this trait among
the males assayed in this experiment.
4.4.1 Experiment 1 – Male hormone treatment alters female
recombination rate
The decreased recombination rate in the offspring of hormone treated males is
delayed until four days after mating. This delayed effect is compatible with the length of
time it takes for a oocyte to progress through the ovariole to oviposition (King 1970; pg.
50-54), and corresponds then to the event which happened four days prior (i.e. mating).
Although there were differences in mating success among males treated with
Juvenile Hormone and control males, this difference may not influence the difference
observed in recombination rates between the two groups. First, if the effect were
associated with promiscuity differences among the females, the conclusion from my
experiment then, would be that the more promiscuous females who mated with the
hormone treated males (males in poorer condition with lower mating success) have
lower recombination rates. This interpretation is perhaps inconsistent with previous
research showing higher rates of recombination in females experiencing higher levels of
mating (Priest et al. 2007). Second, a similar difference in mating success was observed
between temperature treated and control males, but without a corresponding change in
87
recombination rate, suggesting this difference does not contribute to differences in
recombination rates.
An influence of male condition on the outcome of crossing over is evolutionarily
significant in the context of sexual selection, providing direct evidence of how
environmental factors may influence sexually selected traits. Further, this result suggests
that male condition, such as hormone treatment, may have novel repercussions for a
female’s reproductive fitness by altering the recombination rate of her offspring.
4.4.2 Experiment 2 – Variation in male genotype does not contribute
to variation in female recombination rate
As seen in Figure 4.4, the initial results of experiment 2 (before accounting for
variation of female genotype) highlighted strains zim6 and zim30 as most different in
their effect on recombination rate in their mated partner. As an attempt to identify
candidate proteins involved in eliciting the response, I performed comparative
proteomic analysis on the male reproductive tissue of these two strains which identified
three candidate proteins which are differentially expressed and transferred to females
during mating (detailed results presented in Appendix B). However, the results
accounting for female variation do not identify these two strains as having any difference
in their effect on recombination (though recombination rates of females mated to zim6
vs. zim30 are significantly different within female genotype A′a′).
The results of the experiments presented here show that environmentally induced
changes in male condition influence their ability to alter the level of genetic diversity in
their offspring. This result is the first demonstration of males exerting a direct effect on
recombination rate in females with whom they mate. Because females often select mates
based on their condition, females could select mates that generate fitness-enhancing
recombination rates in their offspring, especially in populations experiencing parasitism
88
or fluctuating environments where novel genetic combinations are favored (Keightley
and Otto 2006; Parsons 1988). Although a verbal mechanism has been previously
described for such a male-driven meiotic drive, future work is required to explore the
physiological means through which males mediate recombination rate differences.
Further, future research should evaluate the role of this effect, if any, in sexual selection
or sexual conflict over recombination rates within populations.
89
5. Major Conclusions
A classic struggle in evolutionary biology has been coming to a consensus on what
conditions are necessary for new species to arise and persist. In this struggle, some
scientists have invoked the necessity for groups of organisms to be in isolation to grow
apart (Mayr 1942; Mayr 1963). Darwin himself struggled with the idea that species could
be kept distinct while being allowed to interbreed, as expressed below in the excerpt
from the Origin of Species:
“The view commonly entertained by naturalists is that species, when
intercrossed, have been specially endowed with sterility, in order to
prevent their confusion. This view certainly seems at first highly
probable, for species living together could hardly have been kept distinct
had they been capable of freely crossing.”
Charles Darwin (1859), Ch. 9
Indeed, the widespread adherence to the biological species concept (BSC) aided
the persistence of isolation as key to species maintenance through the modern synthesis
(Hey 2006). However, modern genetics revealed the importance of migration between
populations to best explore the adaptive landscape. This discordance was presented
clearly by Dobzhansky in this excerpt from Genetics and the Origin of Species:
“We are confronted with an apparent anitomy. Isolation prevents the
breakdown of the existing gene systems, and hence precludes the
formation of many worthless gene combinations that are doomed to
destruction. Its role is therefore positive. But on the other hand, isolation
debars the organism from exploring greater and greater portions of the
field of gene combinations, and hence decreases the chance of the
discovery of new and higher adaptive peaks. Isolation is a conservative
factor that slows down the evolutionary process. The anitomy is
removed if one realizes that an agent that is useful at one stage of the
evolutionary process may be harmful at another stage…isolation is
necessary but it must not come too early.”
Theodosius Dobzhansky (1937), pg. 228-9
Although isolation allows locally adapted groups to form co-adapted gene
complexes that function optimally in a given environment, novel alleles from
90
neighboring groups may provide new combinations that may have even higher adaptive
values. This ‘anitomy’ was influential to Wright’s shifting balance theory (Wright 1982),
presented around the same time period, which outlined a key role for migration between
populations to the process of species formation. Although this particular idea has since
come under scrutiny (Coyne et al. 2000), scientists held tightly to the idea that gene flow
could play a role in speciation.
Modern texts on the topic of speciation, although continuing to adhere strongly
to the BSC, note the importance of understanding isolation barriers that specifically act
at the forefront of species formation – those existing while species are in sympatry – as is
presented in the following excerpt from Coyne and Orr’s Speciation book:
“If one accepts some version of the biological species concept, then the
central problem of speciation is understanding the origin of those
isolating barriers that actually or potentially prevent gene flow in
sympatry… First, one must determine which reproductive barriers were
involved in the initial reduction of gene flow between
populations…Second, one must understand which evolutionary forces
produced these barriers.”
Jerry Coyne and Allen Orr (2004), p. 57
More recently, through genetic research of hybrid zones analyzing many hybrids
at multiple loci, we have learned that gene flow varies considerably across the genome
(Arnold 1997; Arnold 2006; Stevison 2008). This variability suggests that there are
many battles being fought between the build up of barriers to reproduction and opposing
gene flow, with regional variation in the outcome of this battle across the genetic
landscape. Further, this unexpected variability shifted thought about speciation as a
simple battle between isolation and gene flow to realizing that the relative strength of
these two forces, rather than the complete absence of gene flow, was critical to the
outcome of the process of species formation.
91
Much work has been focused on the 2nd task suggested by Coyne and Orr (see
excerpt above), to understanding those factors that produce barriers to reproduction, or
more broadly which shape the variation in gene flow across the genome. Focusing
particularly on regions where ‘speciation is winning’ (or regions of low gene flow
between species), we expect to find genes responsible for speciation. This approach has
been useful in identifying several loci important in the process of speciation between
various species. However, much of the variation in estimates of gene flow can be
attributed to the sensitivity of migration estimates to within species nucleotide diversity.
With a tight association between recombination rates and genetic diversity (see below),
regions of low within species recombination tend to have very low genetic diversity,
leading to artificially low estimates of between species gene flow (such as near
centromeres: Strasburg et al. 2009; Stump et al. 2005). However, this underestimate of
gene flow is simply an artifact of low diversity and does not necessarily point to any
specific sets of genes in these regions contributing to speciation (Noor and Bennett
2009).
Genetic mapping studies attempting to dissect regions of low gene flow between
species have often been unable to pinpoint genes conferring differences between species
because these genes are frequently located within chromosomal inversions. Regions
within chromosomal inversions show reduced gene flow/higher nucleotide divergence
due to lower between species recombination (Kulathinal et al. 2009; Machado et al.
2007). Although these regions may indeed contain (and often do contain) isolation loci
(Noor et al. 2001), mapping these loci has proven difficult due to the inherent low
recombination along inverted segments. Collectively, this research has highlighted the
92
impact recombination rate differences both within and between species have on
influencing genomic structure.
My doctoral work, then, has been focused on understanding what shapes
variation of recombination across the genome by focusing on: (1) what shapes variation
in recombination rates within species, specifically understanding the ubiquitous
correlations between recombination rates and both genetic diversity and GC content,
and which evolutionary forces are more likely to contribute to these patterns (see
Chapter 2: “Genetic and Evolution Correlates of Fine-Scale Recombination Rate
Variation in Drosophila persimilis”), and (2) what shapes variation in recombination
rates between species, particularly focusing on how chromosomal inversions influence
changes in recombination rates in hybrids (see Chapter 3: “The effect of chromosomal
inversions on recombination in hybrids between Drosophila pseudoobscura and D.
persimilis”). Finally, I examined plasticity of recombination rates by examining potential
novel forces that disrupt recombination patterns within individuals, mainly malemediated effects and interchromosomal effects (see Chapter 4: “ Male-mediated effects
on female recombination”).
Perhaps owing to the elusive properties of the mechanism of recombination,
classical theory in population genetics began by assuming recombination rates were
uniform throughout the genome, as is revealed in the classic paper by Hudson and
Kaplan on the derivation of the four-gamete test, “Since recombination is assumed to
occur uniformly across the genome…” (Hudson and Kaplan 1985). In 1992, a critical
association was observed between recombination rate and genetic diversity, that would
open a new line of research into understanding how selection versus neutral forces could
predict changes in recombination rates (Begun and Aquadro 1992).
93
My research attributes the association between recombination and diversity
mainly to selective processes, whereas both selective and neutral processes explain the
association with GC content and recombination rates. Further, I was able to show that
recombination rates tend to be conserved between species, by comparing my fine-scale
recombination map in D. persimilis to a previously published recombination map in D.
pseudoobscura (Kulathinal et al. 2008). Finally, I identified fine-scale structuring of
recombination rates where recombination rates were highly correlated with one another
on a 1 Mb scale. Still, the struggle to understand by such different conclusions come out
of various studies on recombination rates continues, and while I provide additional
empirical evidence on the topic, I think we have much to learn of what dictates changes
in recombination rates.
While chromosomal inversions were initially discovered and intended for use of
their supposed neutral features, these structural variants were later discovered to be far
from neutral. Indeed we now know that chromosomal inversions preserve co-adapted
gene complexes by restricting recombination within inverted segments along genomes.
Previous research on inversions segregating in Drosophila has focused on inversions
segregating within populations, while overlooking those that segregate between species,
which are key to our understanding of the process speciation. While recombination
suppression at inversion boundaries and inside inversions tend to be similar for
inversions segregating within and between species, my results suggest that the frequency
of inversion heterozygotes dictates the strength of each of these patterns. Because the
frequency of inversion heterozygotes is almost always lower between species, the impact
of double crossovers within the inverted segment is much lower over long-term timescales between species.
94
Given that recombination rates can be significantly altered within an individual
over the course of its lifetime was unexpected, it is equally surprising then that
recombination rates are considerably variable across the genome and among individuals.
It is not clear why scientists were open to drastic changes in recombination rates with
simple changes in temperature and not with the idea that one chromosomal region could
have differences in the relationship between map distance and physical distance. Even
recently developed population genetics software testing for gene flow does not require
any information on the recombination rate of the surrounding genomic region (but see:
Becquet and Przeworski 2007; Hey and Nielsen 2004; Hey and Nielsen 2007).
Finally, in an experiment to test for environmental variation in male ability to
alter female recombination rates, I have shown that males influence the rate of
recombination in the offspring of their mates perhaps through some sort of segregation
distortion based on chiasma frequency. This kind of segregation distortion, while not a
new concept, is particularly confusing when one considers the selective implications of
males altering the recombination frequency of their progeny.
Finally, I have shown that inversions further act to disrupt recombination rates
elsewhere in the genome, affecting those regions with lowest recombination most,
suggesting these regions to be more susceptible to changes in recombination rates.
Further, I have shown that these disruptions lend themselves to correspondence with
levels of interspecies divergence, suggesting that inversions not only shape patterns of
gene flow and divergence within inverted segments, but that their effects on
recombination disruptions in hybrids also contribute to variation in gene flow between
species across the genome.
95
The research that I have done during my graduate work has shaped my own ideas
about what are the central questions in speciation and the best approaches to answer
these questions using available experimental and genetic data. For my post-doctoral
research, I would like to investigate how changes in recombination profiles in
interspecies hybrids (inferred using genomic sequences of genetically admixed
individuals) relative to pure species contributes to nucleotide variability between species
in a context that is independent of the presence of chromosome inversions. Furthermore,
I plan to survey existing evidence from available research systems with genetic data
where recombination rates and gene flow have been extensively documented (e.g. mice,
sunflowers, tomatoes) to develop a generalized understanding of the role of between
species recombination and hybridization.
96
Appendix A
Pilot experiment for Chapter 4: Experiment 2 – Test if standing genetic
variation of males affects the proportion of recombinant offspring
The purpose of the pilot experiment was to determine the degree of natural
variation in male ability to influence recombination rates in offspring. The use of genetic
markers allowed the measurement of recombinational distances in offspring of different
males. For this experiment, I used 5 lines of D. melanogaster stocks from Zimbabwe (4
additional lines courtesy of C. Aquadro lab). These stocks, though inbred, have
heterozygosity rates higher than many other D. melanogaster stocks (Begun and
Aquadro 1993). By screening microsatellites, I identified two markers that differentiate
the two Zimbabwe stocks zim11 and zim152 (see below). These markers are the same X
chromosome markers assayed in the full-scale experiment (see Chapter 4).
Similar to the cross design for the full-scale experiment described within Chapter
4 (see Figure 4.3), a heterozygous female was generated by crossing 2 of the Zimbabwe
lines: Zim 11 and Zim 152. Each heterozygous female was then mated to a single male
from 1 of 3 Zimbabwe lines to generate an outbred generation with 3 categorical
response variables that were compared. Original parents were not used in the
subsequent cross to avoid inbreeding. In this pilot study, I genotyped 2521 individuals at
both markers for the 3 Zimbabwe crosses: Zim 6, Zim 30, and Zim 42.
97
Figure A.1. Preliminary results for experiment 2. Contingency tables of
the number of recombinant (R) and non-recombinant (NR) offspring of males from
three different Zimbabwe stocks split into offspring from (A) first 1-4 days post-mating
and (B) next 5-8 days post-mating. These were analyzed separately because maternal age
affects recombination rate.
I obtained a G-squared p-value for each comparison using a 2x3 contingency
table of the counts of recombinant and non-recombinant offspring of males from each of
the lines. The two comparisons were between recombination rates of the offspring
generated 1-4 days post-mating and 5-8 days post-mating respectively. This separation
was done to account for changes in recombination rate with maternal age (Bridges 1929).
The p-value for the comparison for the first four days after mating was 0.07 (Figure A.1),
whereas the p-value for the second four days after mating was 0.36 (Figure A.2). The
marginally significant effect in the first four days after mating was evidence that there
may be heterogeneity in recombination rates among crosses due to the genetic
background of the father. Evaluating the 4-day age pools from these crosses in light of
the data from the Redfield experiment suggested that 2 day pools of progeny would be
more informative than four day pools in this pilot experiment, especially in the initial 4
day period post-mating (Redfield 1966).
98
A post-hoc power analysis indicated that based on the sample sizes in this
preliminary data, I had as much as 60% power to detect a 2% recombination rate
difference in the first set of day 1 – 4 offspring (Figure A.1A) (Faul et al. 2007). The
difference in sample size and the smaller effect size (minimum effect size 1.2%) may
explain the reduced power (only 26%) in the next four days (Figure A.1B). Increased egglaying rate soon after mating explained the difference in sample size between the 1-4
days post-mating and 5-8 days post-mating offspring pools. Published results from the
Priest et al (2007) study measured around 4% difference in recombination rate between
categories, whereas my analysis suggested effect sizes as small as 2%. To detect this
small of an effect size, I would need approximately 3000 individuals for each cross (Faul
et al. 2007). In addition to increasing the sample size of each cross, I decided to add an
additional non-Zimbabwe D. melanogaster line into the experiment. Due to the larger
expected divergence between lines, higher levels of genetic variation among males will
potentially increase the effect size difference in recombination rates due to malemediated influences. Therefore, the pilot experiment aided the design of the full-scale
experiment, where I transferred females every 2 days rather than every 4 days, and
increased the divergence among the males by adding the OregonR males to the
experiment, with a target sample size of ~3,000 offspring per male treatment group.
99
Appendix B
Proteomic analysis to follow-up on Experiment 2 from Chapter 4: Test for
quantitative differences between zim6 and zim30 in male reproductive
proteins transferred to females during mating
Based on the initial results of experiment 2 (Figure 4.4), a follow-up proteomic
analysis was designed to test for differences between males with the largest effect size
difference (zim6 vs. zim30). Male reproductive tissue, including both testes and
accessory glands, was dissected from 4-day old virgin males in Ringer’s solution. Tissue
was then transferred to ~50 uL of 50 mM Ammonium Bicarbonate with 0.1% Rapigest
solution to preserve proteins and stored at -80° prior to protein analysis. Approximately
20-30 males (all collected on the same day) were pooled to generate enough protein for
each sample. A total of three samples per Zimbabwe stock (six total) were generated for
quantitative comparison of proteins with each stock and between the two stocks.
The proteomic analysis took place at the Duke IGSP Proteomics Facility. Organs
were lysed by sonication with a probe sonicator, followed by a Bradford assay to any
remove debris. Approximately 10 µg of each of six samples was used for quantitation via
Mass Spectrometry. The software Elucidator was then used to compare the three zim6
samples to the three zim30 samples, examining differences within the pooled samples to
the differences between zim6 and zim30.
Once a report of proteins differentially expressed between the two samples was
received, I cross-referenced this list with a list of proteins known to be transferred to
Drosophila melanogaster females during mating (Findlay et al. 2008). The results of
this cross-referencing (Table B.1) revealed three significant candidate proteins that could
contribute to the effect difference between these males on female recombination rate.
100
Figure B.1 shows graphically the variation within and between each sample for the
candidate proteins identified in Table B.1.
As discussed within Chapter 4, there is no difference in offspring recombination
rate between zim6 and zim30, once female genotype is considered. Therefore, these
proteomic results do not reflect differences in the phenotype originally intended.
However, these results may be useful in highlighting male reproductive proteins that are
commonly differentially expressed between closely related Drosophila males.
101
102
# of
Peptides
1
4
2
1
1
11
7
1
3
3
1
1
1
10
9
2
1
1
12
1
13
1
2
2
9
2
2
1
Fold
Change
14.473
5.291
2.129
1.856
-1.361
-1.507
-1.675
-2.137
-2.18
-3.122
-3.636
-119.514
2.789
-1.312
-1.395
-1.422
-1.516
4.272
-1.292
-1.451
-1.483
10.818
2.51
1.639
1.522
-1.274
-1.406
-1.692
Log
(Intensity)
4.353***
4.536***
4.215***
4.37***
4.246***
5.414***
5.113***
3.856***
4.91***
4.865***
3.784***
2.437***
4.124***
5.575***
5.431***
4.831***
4.319***
4.285**
5.506**
4.21**
5.639**
3.709*
4.474*
4.521*
5.49*
4.44*
4.869*
4.151*
Intensity
Zim30
5932.124
14922.787
11250.851
17201.17
20544.299
318473.375
167892.531
10481.583
120158.273
129338.656
11605.611
2990.176
7965.793
430387.188
318651.219
80778.953
25656.068
9318.477
364257.031
19536.213
530716.625
1555.772
18820.738
25918.08
250486.297
31115.043
87709.117
18435.418
Intensity
Zim6
85855.281
78955.406
23947.736
31918.424
15100.542
211277.484
100208.977
4905.57
55106.445
41427.551
3192.271
25.019
22219.812
328065.031
228490.844
56825.219
16924.465
39807.637
281918.438
13466.661
357863.125
16830.027
47235.199
42470.418
381213.188
24418.756
62378.391
10895.5
2
1
2
2
2
2
2
2
2
2
2
1
2
2
Ref
P<0.0001***; P<0.01**; P<0.05*; Bolded proteins identified as being transferred to females during mating (references: 1-Wolfner et al. 2006; 2-Findlay et al. 2008). Only
proteins cross-referenced are listed for P>0.0001 (Bonferroni-corrected cutoff).
CG34034, isoform A
CG9360
Tal
Neuropeptide-like precursor 2
CG10063
CG4546
Porin2
CG13887, isoform C
Odorant-binding protein 99a
CG3301, isoform B
CG6168
CG15116
cecropin C
calreticulin
heat shock protein 27
ribosomal protein L31, isoform B
smt3
CG5402
glycoprotein 93
seminal fluid protein 33A3
accessory gland-specific peptide 26Aa
seminal fluid protein 87B
CG5162
Odorant-binding protein 56e
CG9029
CG18284
accessory gland peptide 62F
Odorant-binding protein 56f
Protein Description
Table B.1. Proteins Identified as differentially expressed between Zim6 and Zim30 using Mass Spectrometry
Figure B.1. Individual protein results of top three candidates (from
Table B.1) showing variability between each of three replicates in total
protein intensity for both zim6 and zim30 samples.
103
References
Agrawal, A. F., L. Hadany and S. P. Otto (2005). "The evolution of plastic
recombination." Genetics 171(2): 803-812.
Akashi, H. and S. W. Schaeffer (1997). "Natural selection and the frequency distributions
of ''silent'' DNA polymorphism in Drosophila." Genetics 146(1): 295-307.
Allers, T. and M. Lichten (2001). "Differential timing and control of noncrossover and
crossover recombination during meiosis." Cell 106(1): 47-57.
Altschul, S. F., T. L. Madden, A. A. Schaffer, J. H. Zhang, Z. Zhang, W. Miller and D. J.
Lipman (1997). "Gapped BLAST and PSI-BLAST: a new generation of protein
database search programs." Nucleic Acids Research 25(17): 3389-3402.
Andolfatto, P., F. Depaulis and A. Navarro (2001). "Inversion polymorphisms and
nucleotide variability in Drosophila." Genetical Research 77(1): 1-8.
Andolfatto, P. and M. Przeworski (2001). "Regions of lower crossing over harbor more
rare variants in African populations of Drosophila melanogaster." Genetics
158(2): 657-665.
Andolfatto, P. and J. D. Wall (2003). "Linkage disequilibrium patterns across a
recombination gradient in African Drosophila melanogaster." Genetics 165(3):
1289-1305.
Andolfatto, P., J. D. Wall and M. Kreitman (1999). "Unusual haplotype structure at the
proximal breakpoint of In(2L)t in a natural population of Drosophila
melanogaster." Genetics 153(3): 1297-1311.
Angilletta, M. J., P. H. Niewiarowski and C. A. Navas (2002). "The evolution of thermal
physiology in ectotherms." Journal of Thermal Biology 27(4): 249-268.
Aquadro, C. F., A. L. Weaver, S. W. Schaeffer and W. W. Anderson (1991). "Molecular
Evolution of Inversions in Drosophila pseudoobscura - the Amylase Gene
Region." Proceedings of the National Academy of Sciences of the United States of
America 88(1): 305-309.
Arnold, M. L. (1997). Natural hybridization and evolution. New York, Oxford University
Press.
Arnold, M. L. (2006). Evolution through genetic exchange. Oxford ; New York, Oxford
University Press.
Ashburner, M. and T. R. F. Wright (1980). The genetics and biology of Drosophila.
London; New York, Academic Press.
Bachtrog, D. (2007). "Reduced selection for codon usage bias in Drosophila miranda."
Journal of Molecular Evolution 64(5): 586-590.
104
Baines, J. F., A. Das, S. Mousset and W. Stephan (2004). "The role of natural selection in
genetic differentiation of worldwide populations of Drosophila ananassae."
Genetics 168(4): 1987-1998.
Bartolome, C., X. Maside and B. Charlesworth (2002). "On the abundance and
distribution of transposable elements in the genome of Drosophila
melanogaster." Molecular Biology and Evolution 19(6): 926-937.
Baudat, F., J. Buard, C. Grey, A. Fledel-Alon, C. Ober, M. Przeworski, G. Coop and B. de
Massy (2010). "PRDM9 Is a Major Determinant of Meiotic Recombination
Hotspots in Humans and Mice." Science.
Becquet, C. and M. Przeworski (2007). "A new approach to estimate parameters of
speciation models with application to apes." Genome Research 17(10): 1505-1519.
Begun, D. J. and C. F. Aquadro (1992). "Levels of Naturally-Occurring DNA
Polymorphism Correlate with Recombination Rates in Drosophila
melanogaster." Nature 356(6369): 519-520.
Begun, D. J. and C. F. Aquadro (1993). "African and North-American populations of
Drosophila melanogaster are very different at the DNA level." Nature
365(6446): 548-550.
Begun, D. J., A. K. Holloway, K. Stevens, L. W. Hillier, Y. P. Poh, et al. (2007).
"Population genomics: Whole-genome analysis of polymorphism and divergence
in Drosophila simulans." Plos Biology 5(11): 2534-2559.
Benson, G. (1999). "Tandem repeats finder: a program to analyze DNA sequences."
Nucleic Acids Research 27(2): 573-580.
Bhutkar, A., S. W. Schaeffer, S. M. Russo, M. Xu, T. E. Smith and W. M. Gelbart (2008).
"Chromosomal rearrangement inferred from comparisons of 12 Drosophila
genomes." Genetics 179(3): 1657-1680.
Birdsell, J. A. (2002). "Integrating genomics, bioinformatics, and classical genetics to
study the effects of recombination on genome evolution." Molecular Biology and
Evolution 19(7): 1181-1197.
Birky, C. W. and J. B. Walsh (1988). "Effects of Linkage on Rates of Molecular
Evolution." Proceedings of the National Academy of Sciences of the United States
of America 85(17): 6414-6418.
Bridges, C. B. (1927). "The relation of the age of the female to crossing over in the third
chromosome of Drosophila melanogaster." Journal of General Physiology 8(6):
689-700.
Bridges, C. B. (1929). "Variation in crossing over in relation to the age of the female in
Drosophila melanogaster." Carnegie Inst. Wash. Publ. 399: 63-89.
105
Brown, J. D. and R. J. O'Neill (2010). "Chromosomes, Conflict, and Epigenetics:
Chromosomal Speciation Revisited." Annual Review of Genomics and Human
Genetics, Vol 11 11: 291-316.
Buard, J. and B. de Massy (2007). "Playing hide and seek with mammalian meiotic
crossover hotspots." Trends in Genetics 23(6): 301-309.
Buckler, E. S., T. L. Phelps-Durr, C. S. K. Buckler, R. K. Dawe, J. F. Doebley and T. P.
Holtsford (1999). "Meiotic drive of chromosomal knobs reshaped the maize
genome." Genetics 153(1): 415-426.
Bussell, J. J., N. M. Pearson, R. Kanda, D. A. Filatov and B. T. Lahn (2006). "Human
polymorphism and human-chimpanzee divergence in pseudoautosomal region
correlate with local recombination rate." Gene 368: 94-100.
Busza, A., A. Murad and P. Emery (2007). "Interactions between circadian neurons
control temperature synchronization of Drosophila behavior." Journal of
Neuroscience 27(40): 10722-10733.
Caceres, M., J. M. Ranz, A. Barbadilla, M. Long and A. Ruiz (1999). "Generation of a
widespread Drosophila inversion by a transposable element." Science 285(5426):
415-418.
Carneiro, M., J. A. Blanco-Aguiar, R. Villafuerte, N. Ferrand and M. W. Nachman (2010).
"Speciation in the European Rabbit (Oryctolagus Cuniculus): Islands of
Differentiation on the X Chromosome and Autosomes." Evolution 64(12): 34433460.
Carson, H. L. (1946). "The Selective Elimination of Inversion Dicentric Chromatids
during Meiosis in the Eggs of Sciara Impatiens." Genetics 31(1): 95-113.
Charlesworth, B., C. Bartolome and V. Noel (2005). "The detection of shared and
ancestral polymorphisms." Genetical Research 86(2): 149-157.
Charlesworth, B., M. T. Morgan and D. Charlesworth (1993). "The Effect of Deleterious
Mutations on Neutral Molecular Variation." Genetics 134(4): 1289-1303.
Chovnick, A. (1973). "Gene Conversion and Transfer of Genetic Information within
Inverted Region of Inversion Heterozygotes." Genetics 75(1): 123-131.
Cirulli, E. T., R. M. Kliman and M. A. F. Noor (2007). "Fine-scale crossover rate
heterogeneity in Drosophila pseudoobscura." Journal of Molecular Evolution
64(1): 129-135.
Clark, A. G., M. B. Eisen, D. R. Smith, C. M. Bergman, B. Oliver, et al. (2007). "Evolution
of genes and genomes on the Drosophila phylogeny." Nature 450(7167): 203218.
106
Colson, I., S. J. Macdonald and D. B. Goldstein (1999). "Microsatellite markers for
interspecific mapping of Drosophila simulans and D. sechellia." Molecular
Ecology 8(11): 1951-1955.
Comeron, J. M. and M. Kreitman (2002). "Population, evolutionary and genomic
consequences of interference selection." Genetics 161(1): 389-410.
Comeron, J. M., A. Williford and R. M. Kliman (2008). "The Hill-Robertson effect:
evolutionary consequences of weak selection and linkage in finite populations."
Heredity 100(1): 19-31.
Coop, G. and M. Przeworski (2007). "An evolutionary view of human recombination."
Nature Reviews Genetics 8(1): 23-34.
Coop, G., X. Q. Wen, C. Ober, J. K. Pritchard and M. Przeworski (2008). "Highresolution mapping of crossovers reveals extensive variation in fine-scale
recombination patterns among humans." Science 319(5868): 1395-1398.
Copenhaver, G. P. (2005). "Plant genetics: When not to interfere." Current Biology
15(8): R290-R291.
Copenhaver, G. P., E. A. Housworth and F. W. Stahl (2002). "Crossover interference in
Arabidopsis." Genetics 160(4): 1631-1639.
Coyne, J. A., N. H. Barton and M. Turelli (2000). "Is Wright's shifting balance process
important in evolution?" Evolution 54(1): 306-317.
Coyne, J. A., W. Meyers, A. P. Crittenden and P. Sniegowski (1993). "The Fertility Effects
of Pericentric Inversions in Drosophila melanogaster." Genetics 134(2): 487496.
Coyne, J. A. and H. A. Orr (2004). Speciation. Sunderland, Massachusetts, Sinauer
Associates, Inc.
Crow, J. F. and M. Kimura (1970). An introduction to population genetics theory. New
York, Harper & Row.
D'amours, D. and S. P. Jackson (2002). "The Mre11 complex: At the crossroads of DNA
repair and checkpoint signalling." Nature Reviews Molecular Cell Biology 3(5):
317-327.
Darwin, C. (1859). On the origin of the species by means of natural selection. London.
de Massy, B. (2003). "Distribution of meiotic recombination sites." Trends in Genetics
19(9): 514-522.
de Procé, S. M., D. L. Halligan, P. D. Keightley and B. Charlesworth (2009). "Patterns of
DNA-Sequence Divergence Between Drosophila miranda and D.
pseudoobscura." Journal of Molecular Evolution 69(6): 601-611.
107
Dobzhansky, T. and C. Epling (1948). "The Suppression of Crossing over in Inversion
Heterozygotes of Drosophila pseudoobscura." Proceedings of the National
Academy of Sciences of the United States of America 34(4): 137-141.
Dobzhansky, T. G. (1937). Genetics and the origin of species. New York, NY, Columbia
Univ. Press.
Dobzhansky, T. G. (1973). "Is There Gene Exchange between Drosophila pseudoobscura
and Drosophila persimilis in Their Natural Habitats." American Naturalist
107(954): 312-314.
Dobzhansky, T. G. and C. Epling (1944). Contributions to the Genetics, Taxonomy, and
Ecology of Drosophila pseudoobscura and Its Relatives. Washington, D.C.
Dolgin, E. S. and B. Charlesworth (2008). "The effects of recombination rate on the
distribution and abundance of transposable elements." Genetics 178(4): 21692177.
Drouaud, J., C. Camilleri, P. Y. Bourguignon, A. Canaguier, A. Berard, et al. (2006).
"Variation in crossing-over rates across chromosome 4 of Arabidopsis thaliana
reveals the presence of meiotic recombination "hot spots"." Genome Research
16(1): 106-14.
Dumont, B. L. and B. A. Payseur (2008). "Evolution of the genomic rate of
recombination in mammals." Evolution 62(2): 276-294.
Duret, L. and P. F. Arndt (2008). "The impact of recombination on nucleotide
substitutions in the human genome." Plos Genetics 4(5): e1000071.
Duret, L., M. Semon, G. Piganeau, D. Mouchiroud and N. Galtier (2002). "Vanishing GCrich isochores in mammalian genomes." Genetics 162(4): 1837-47.
Fan, J. B., A. Oliphant, R. Shen, B. G. Kermani, F. Garcia, et al. (2003). "Highly parallel
SNP genotyping." Cold Spring Harb Symp Quant Biol 68: 69-78.
Faria, R. and A. Navarro (2010). "Chromosomal speciation revisited: rearranging theory
with pieces of evidence." Trends in Ecology & Evolution 25(11): 660-669.
Faul, F., E. Erdfelder, A. G. Lang and A. Buchner (2007). "G*Power 3: a flexible
statistical power analysis program for the social, behavioral, and biomedical
sciences." Behavior Research Methods 39(2): 175-91.
Feder, J. L. and P. Nosil (2009). "Chromosomal Inversions and Species Differences:
When Are Genes Affecting Adaptive Divergence and Reproductive Isolation
Expected to Reside within Inversions?" Evolution 63(12): 3061-3075.
Felsenstein, J. (1974). "Evolutionary Advantage of Recombination." Genetics 78(2): 737756.
108
Findlay, G. D., X. H. Yi, M. J. MacCoss and W. J. Swanson (2008). "Proteomics reveals
novel Drosophila seminal fluid proteins transferred at mating." Plos Biology 6(7):
1417-1426.
Fisher, R. A. (1930). The genetical theory of natural selection. Oxford, The Clarendon
press.
Fitzpatrick, C. L., L. S. Stevison and M. A. F. Noor (2009). "Fine-scale crossover rate and
interference along the XR-chromosome arm of Drosophila pseudoobscura."
Drosophila Information Service 92: 27-29.
Flatt, T. and T. J. Kawecki (2007). "Juvenile hormone as a regulator of the trade-off
between reproduction and life span in Drosophila melanogaster." Evolution
61(8): 1980-1991.
Flatt, T., M. P. Tu and M. Tatar (2005). "Hormonal pleiotropy and the juvenile hormone
regulation of Drosophila development and life history." Bioessays 27(10): 9991010.
Foguera, G., J. L. Munoz, S. Ceballos, F. Bozinovic and E. Hasson (2007). "Testing the
interplay between physiological and life-history traits: an experimental study in
Drosophila." Evolutionary Ecology Research 9(7): 1211-1222.
Foss, E., R. Lande, F. W. Stahl and C. M. Steinberg (1993). "Chiasma Interference as a
Function of Genetic-Distance." Genetics 133(3): 681-691.
Frazer, K. A., D. G. Ballinger, D. R. Cox, D. A. Hinds, L. L. Stuve, et al. (2007). "A second
generation human haplotype map of over 3.1 million SNPs." Nature 449(7164):
851-U3.
Gaut, B. S., S. I. Wright, C. Rizzon, J. Dvorak and L. K. Anderson (2007). "Opinion Recombination: an underappreciated factor in the evolution of plant genomes."
Nature Reviews Genetics 8(1): 77-84.
Gerton, J. L., J. DeRisi, R. Shroff, M. Lichten, P. O. Brown and T. D. Petes (2000).
"Global mapping of meiotic recombination hotspots and coldspots in the yeast
Saccharomyces cerevisiae." Proceedings of the National Academy of Sciences of
the United States of America 97(21): 11383-11390.
Gilliland, W. D. (2003). Population Genetics of two Female-Specific Meiotic Loci of
Drosophila melanogaster. Population Biology, University of California, Davis.
PhD.
Gloor, G. B. and W. R. Engels (1992). "Single-fly DNA preps for PCR." Drosophila
Information Service 71: 148-149.
Gordo, I. and B. Charlesworth (2001). "Genetic linkage and molecular evolution."
Current Biology 11(17): R684-R686.
109
Groenen, M. A. M., P. Wahlberg, M. Foglio, H. H. Cheng, H. J. Megens, et al. (2009). "A
high-density SNP-based linkage map of the chicken genome reveals sequence
features correlated with recombination rate." Genome Research 19(3): 510-519.
Gruneberg, H. (1935). "A new inversion of the X-chromosome in Drosophila
melanogaster." Journal of Genetics 31(2): 163-184.
Haber, J. E., G. Ira, A. Malkova and N. Sugawara (2004). "Repairing a double-strand
chromosome break by homologous recombination: revisiting Robin Holliday's
model." Philosophical Transactions of the Royal Society of London Series BBiological Sciences 359(1441): 79-86.
Hadany, L. and T. Beker (2003). "On the evolutionary advantage of fitness-associated
recombination." Genetics 165(4): 2167-2179.
Haddrill, P. R., B. Charlesworth, D. L. Halligan and P. Andolfatto (2005). "Patterns of
intron sequence evolution in Drosophila are dependent upon length and GC
content." Genome Biology 6(8): R67.
Haldane, J. B. S. (1922). "Sex ratio and unisexual sterility in hybrid animals." Journal of
Genetics 12(2): 101-109.
Hasson, E. and W. F. Eanes (1996). "Contrasting histories of three gene regions
associated with In(3L)Payne of Drosophila melanogaster." Genetics 144(4):
1565-1575.
Hellmann, I., I. Ebersberger, S. E. Ptak, S. Paabo and M. Przeworski (2003). "A neutral
explanation for the correlation of diversity with recombination rates in humans."
American Journal of Human Genetics 72(6): 1527-1535.
Hellmann, I., K. Prufer, H. K. Ji, M. C. Zody, S. Paabo and S. E. Ptak (2005). "Why do
human diversity levels vary at a megabase scale?" Genome Research 15(9): 12221231.
Herrel, A., R. S. James and R. Van Damme (2007). "Fight versus flight: physiological
basis for temperature-dependent behavioral shifts in lizards." Journal of
Experimental Biology 210(10): 1762-1767.
Hey, J. (2006). "On the failure of modern species concepts." Trends in Ecology &
Evolution 21(8): 447-450.
Hey, J. and R. M. Kliman (2002). "Interactions between natural selection,
recombination and gene density in the genes of drosophila." Genetics 160(2):
595-608.
Hey, J. and R. Nielsen (2004). "Multilocus methods for estimating population sizes,
migration rates and divergence time, with applications to the divergence of
Drosophila pseudoobscura and D. persimilis." Genetics 167(2): 747-760.
110
Hey, J. and R. Nielsen (2007). "Integration within the Felsenstein equation for improved
Markov chain Monte Carlo methods in population genetics." Proceedings of the
National Academy of Sciences of the United States of America 104(8): 27852790.
Hill, W. G. and A. Robertson (1966). "Effect of linkage on limits to artificial selection."
Genetical Research 8(3): 269-294.
Hillers, K. J. (2004). "Crossover interference." Current Biology 14(24): R1036-R1037.
Hilliker, A. J., G. Harauz, A. G. Reaume, M. Gray, S. H. Clark and A. Chovnick (1994).
"Meiotic Gene Conversion Tract Length Distribution within the Rosy Locus of
Drosophila melanogaster." Genetics 137(4): 1019-1024.
Hoffmann, A. A. and L. H. Rieseberg (2008). "Revisiting the Impact of Inversions in
Evolution: From Population Genetic Markers to Drivers of Adaptive Shifts and
Speciation?" Annual Review of Ecology Evolution and Systematics 39: 21-42.
Housworth, E. A. and F. W. Stahl (2003). "Crossover interference in humans." American
Journal of Human Genetics 73(1): 188-197.
Hudson, R. R. and N. L. Kaplan (1985). "Statistical properties of the number of
recombination events in the history of a sample of DNA sequences." Genetics 111:
147-164.
Huxley, J. S. (1928). "Sexual difference of linkage in Gammarus chevreuxi." Journal of
Genetics 20(2): 145-156.
Ihara, N., A. Takasuga, K. Mizoshita, H. Takeda, M. Sugimoto, et al. (2004). "A
comprehensive genetic map of the cattle genome based on 3802 microsatellites."
Genome Research 14(10A): 1987-1998.
Ishii, K. and B. Charlesworth (1977). "Associations between Allozyme Loci and Gene
Arrangements Due to Hitch-Hiking Effects of New Inversions." Genetical
Research 30(2): 93-106.
Jackson, B. C. (in press). "Recombination-suppression: how many mechanisms for
chromosomal speciation?" Genetica.
Joyce, E. F. and K. S. McKim (2010). "Chromosome Axis Defects Induce a CheckpointMediated Delay and Interchromosomal Effect on Crossing Over during
Drosophila Meiosis." Plos Genetics 6(8): e1001059.
Keeney, S., C. N. Giroux and N. Kleckner (1997). "Meiosis-specific DNA double-strand
breaks are catalyzed by Spo11, a member of a widely conserved protein family."
Cell 88(3): 375-384.
Keightley, P. D. and S. P. Otto (2006). "Interference among deleterious mutations
favours sex and recombination in finite populations." Nature 443(7107): 89-92.
111
Khil, P. P. and R. D. Camerini-Otero (2010). "Genetic crossovers are predicted accurately
by the computed human recombination map." PLoS Genet 6(1): e1000831.
King, R. C. (1970). Ovarian development in Drosophila melanogaster. New York,,
Academic Press.
Kirkpatrick, M. and N. Barton (2006). "Chromosome inversions, local adaptation and
speciation." Genetics 173(1): 419-434.
Kliman, R. M. and J. Hey (1993). "Reduced Natural-Selection Associated with Low
Recombination in Drosophila melanogaster." Molecular Biology and Evolution
10(6): 1239-1258.
Kliman, R. M. and J. Hey (2003). "Hill-Robertson interference in Drosophila
melanogaster: reply to Marais, Mouchiroud and Duret." Genetical Research
81(2): 89-90.
Kong, A., D. F. Gudbjartsson, J. Sainz, G. M. Jonsdottir, S. A. Gudjonsson, et al. (2002).
"A high-resolution recombination map of the human genome." Nature Genetics
31(3): 241-247.
Kosambi, D. D. (1944). "The estimation of map distances from recombination values."
Annals of Eugenics 12: 172-175.
Krimbas, C. V. and J. R. Powell (1992). Drosophila inversion polymorphism. Boca Raton,
Fla., CRC Press.
Kucherlapati, R. and G. R. Smith (1988). Genetic recombination. Washington, D.C.,
American Society for Microbiology.
Kulathinal, R. J., S. M. Bennett, C. L. Fitzpatrick and M. A. F. Noor (2008). "Fine-scale
mapping of recombination rate in Drosophila refines its correlation to diversity
and divergence." Proceedings of the National Academy of Sciences of the United
States of America 105(29): 10051-10056.
Kulathinal, R. J., L. S. Stevison and M. A. F. Noor (2009). "The Genomics of Speciation
in Drosophila: Diversity, Divergence, and Introgression Estimated Using LowCoverage Genome Sequencing." Plos Genetics 5(7): e1000550.
Laayouni, H., E. Hasson, M. Santos and A. Fontdevila (2003). "The evolutionary history
of Drosophila buzzatii. XXXV. Inversion polymorphism and nucleotide
variability in different regions of the second chromosome." Molecular Biology
and Evolution 20(6): 931-944.
Leman, S. C., Y. Chen, J. E. Stajich, M. A. Noor and M. K. Uyenoyama (2005).
"Likelihoods from summary statistics: recent divergence between species."
Genetics 171(3): 1419-36.
112
Lercher, M. J. and L. D. Hurst (2002). "Human SNP variability and mutation rate are
higher in regions of high recombination." Trends in Genetics 18(7): 337-340.
Levine, R. P. (1956). "Crossing over and Inversions in Coadapted Systems." American
Naturalist 90(850): 41-45.
Lucchesi, J. T. and D. T. Suzuki (1968). "The Interchromosomal control of
recombination." Annu Rev Genet 2: 53-86.
Machado, C. a., T. S. Haselkorn and M. A. F. Noor (2007). "Evaluation of the genomic
extent of effects of fixed inversion differences on intraspecific variation and
interspecific gene flow in Drosophila pseudoobscura and D. persimilis." Genetics
175(3): 1289-1306.
Machado, C. A., R. M. Kliman, J. A. Markert and J. Hey (2002). "Inferring the history of
speciation from multilocus DNA sequence data: The case of Drosophila
pseudoobscura and close relatives." Molecular Biology and Evolution 19(4): 472488.
Maddox, J. F., K. P. Davies, A. M. Crawford, D. J. Hulme, D. Vaiman, et al. (2001). "An
enhanced linkage map of the sheep genome comprising more than 1000 loci."
Genome Research 11(7): 1275-1289.
Mancera, E., R. Bourgon, A. Brozzi, W. Huber and L. M. Steinmetz (2008). "Highresolution mapping of meiotic crossovers and non-crossovers in yeast." Nature
454(7203): 479-U1.
Marais, G. (2003). "Biased gene conversion: implications for genome and sex evolution."
Trends in Genetics 19(6): 330-338.
Marais, G., D. Mouchiroud and L. Duret (2001). "Does recombination improve selection
on codon usage? Lessons from nematode and fly complete genomes."
Proceedings of the National Academy of Sciences of the United States of America
98(10): 5688-5692.
Marais, G., P. Nouvellet, P. D. Keightley and B. Charlesworth (2005). "Intron size and
exon evolution in Drosophila." Genetics 170(1): 481-485.
Margulies, M., M. Egholm, W. E. Altman, S. Attiya, J. S. Bader, et al. (2005). "Genome
sequencing in microfabricated high-density picolitre reactors." Nature
437(7057): 376-380.
Mark, H. F. L. and S. Zimmering (1977). "Centromeric Effect on Degree of Nonrandom
Disjunction in Female Drosophila-Melanogaster." Genetics 86(1): 121-132.
Marsolier-Kergoat, M. C. and E. Yeramian (2009). "GC Content and Recombination:
Reassessing the Causal Effects for the Saccharomyces cerevisiae Genome."
Genetics 183(1): 31-38.
113
Maynard Smith, J. and J. Haigh (1974). "Hitch-Hiking Effect of a Favorable Gene."
Genetical Research 23(1): 23-35.
Maynard Smith, J. and S. Maynard Smith (1954). "Genetics and Cytology of Drosophila
subobscura .VIII. Heterozygosity, Viability and Rate of Development." Journal of
Genetics 52(1): 152-164.
Mayr, E. (1942). Systematics and the origin of species from the viewpoint of a zoologist.
New York, Columbia University Press.
Mayr, E. (1963). Animal species and evolution. Cambridge,, Belknap Press of Harvard
University Press.
McVean, G. (2007). "The structure of linkage disequilibrium around a selective sweep."
Genetics 175(3): 1395-1406.
Menotti-Raymond, M., V. A. David, L. A. Lyons, A. A. Schaffer, J. F. Tomlin, M. K.
Hutton and S. J. O'Brien (1999). "A genetic linkage map of microsatellites in the
domestic cat (Felis catus)." Genomics 57(1): 9-23.
Modrich, P. and R. Lahue (1996). "Mismatch repair in replication fidelity, genetic
recombination, and cancer biology." Annual Review of Biochemistry 65: 101-133.
Muller, H. J. (1964). "The relation of recombination to mutational advance." Mutation
Research 1(1): 2-9.
Myers, S., L. Bottolo, C. Freeman, G. McVean and P. Donnelly (2005). "A fine-scale map
of recombination rates and hotspots across the human genome." Science
310(5746): 321-324.
Myers, S., R. Bowden, A. Tumian, R. E. Bontrop, C. Freeman, T. S. MacFie, G. McVean
and P. Donnelly (2010). "Drive Against Hotspot Motifs in Primates Implicates the
PRDM9 Gene in Meiotic Recombination." Science.
Myers, S., C. Freeman, A. Auton, P. Donnelly and G. McVean (2008). "A common
sequence motif associated with recombination hot spots and genome instability
in humans." Nature Genetics 40(9): 1124-1129.
Nachman, M. W. (1997). "Patterns of DNA variability at X-linked loci in Mus
domesticus." Genetics 147(3): 1303-1316.
Nachman, M. W. (2002). "Variation in recombination rate across the genome: evidence
and implications." Current Opinion in Genetics & Development 12(6): 657-663.
Navarro, A., E. Betran, A. Barbadilla and A. Ruiz (1997). "Recombination and gene flux
caused by gene conversion and crossing over in inversion heterokaryotypes."
Genetics 146(2): 695-709.
114
Neel, J. V. (1941). "A relation between larval nutrition and the frequency of crossing over
in the third chromosome of Drosophila melanogaster." Genetics 26(5): 506-516.
Niehuis, O., J. D. Gibson, M. S. Rosenberg, B. A. Pannebakker, T. Koevoets, et al. (2010).
"Recombination and Its Impact on the Genome of the Haplodiploid Parasitoid
Wasp Nasonia." Plos One 5(1): e8597.
Nobrega, C., M. Khadem, M. Aguade and C. Segarra (2008). "Genetic exchange versus
genetic differentiation in a medium-sized inversion of Drosophila: The A(2)/A(st)
arrangements of Drosophila subobscura." Molecular Biology and Evolution
25(8): 1534-1543.
Noor, M. A. F. (1997). "Environmental effects on male courtship intensity in Drosophila
pseudoobscura (Diptera: Drosophilidae)." Journal of Insect Behavior 10(2): 305312.
Noor, M. A. F. (2008a). "Connecting recombination, nucleotide diversity and species
divergence in Drosophila." Fly 2(5): 255-256.
Noor, M. A. F. (2008b). "Mutagenesis from meiotic recombination is not a primary
driver of sequence divergence between Saccharomyces species." Mol Biol Evol
25(11): 2439-44.
Noor, M. A. F. and S. M. Bennett (2009). "Islands of speciation or mirages in the desert?
Examining the role of restricted recombination in maintaining species." Heredity
103(6): 439-444.
Noor, M. A. F., D. A. Garfield, S. W. Schaeffer and C. A. Machado (2007). "Divergence
between the Drosophila pseudoobscura and D. persimilis genome sequences in
relation to chromosomal inversions." Genetics 177(3): 1417-1428.
Noor, M. A. F., K. L. Grams, L. A. Bertucci and J. Reiland (2001). "Chromosomal
inversions and the reproductive isolation of species." Proceedings of the National
Academy of Sciences of the United States of America 98(21): 12084-12088.
Noor, M. A. F. and R. M. Kliman (2003). "Variability on the dot chromosome in the
Drosophila simulans clade." Genetica 118(1): 51-58.
Novitski, E. and G. Braver (1954). "An Analysis of Crossing over within a Heterozygous
Inversion in Drosophila melanogaster." Genetics 39(2): 197-209.
Orr-Weaver, T. L., J. W. Szostak and R. J. Rothstein (1981). "Yeast Transformation - a
Model System for the Study of Recombination." Proceedings of the National
Academy of Sciences of the United States of America-Biological Sciences 78(10):
6354-6358.
Ortiz-Barrientos, D., A. S. Chang and M. A. F. Noor (2006). "A recombinational portrait
of the Drosophila pseudoobscura genome." Genetical Research 87(1): 23-31.
115
Page, S. L. and R. S. Hawley (2003). "Chromosome choreography: The meiotic ballet."
Science 301(5634): 785-789.
Paigen, K., J. P. Szatkiewicz, K. Sawyer, N. Leahy, E. D. Parvanov, S. H. S. Ng, J. H.
Graber, K. W. Broman and P. M. Petkov (2008). "The Recombinational Anatomy
of a Mouse Chromosome." Plos Genetics 4(7): e1000119.
Palumbi, S. R. (1996). PCR and molecular systematics. Molecular systematics. D. M.
Hillis, C. Moritz and B. K. Mable. Sunderland, Mass., Sinauer Associates: 205247.
Parsch, J. (2003). "Selective constraints on intron evolution in Drosophila." Genetics
165(4): 1843-1851.
Parsch, J., S. Novozhilov, S. S. Saminadin-Peter, K. M. Wong and P. Andolfatto (2010).
"On the Utility of Short Intron Sequences as a Reference for the Detection of
Positive and Negative Selection in Drosophila." Molecular Biology and Evolution
27(6): 1226-1234.
Parsons, P. A. (1988). "Evolutionary rates - effects of stress upon recombination."
Biological Journal of the Linnean Society 35(1): 49-68.
Pearson, W. R. and D. J. Lipman (1988). "Improved Tools for Biological Sequence
Comparison." Proceedings of the National Academy of Sciences of the United
States of America 85(8): 2444-2448.
Pegueroles, C., V. Ordonez, F. Mestres and M. Pascual (2010). "Recombination and
selection in the maintenance of the adaptive value of inversions." Journal of
Evolutionary Biology 23(12): 2709-2717.
Petes, T. D. (2001). "Meiotic recombination hot spots and cold spots." Nature Reviews
Genetics 2(5): 360-369.
Plough, H. H. (1917). "The effect of temperature on crossing over in Drosophila." Journal
of Experimental Zoology 24(2): 147-209.
Plough, H. H. (1921). "Further studies on the effect of temperature on crossing over."
Journal of Experimental Zoology 32(2): 187-202.
Portin, P. and M. Rantanen (1990). "Further-Studies on the Interchromosomal Effect on
Crossing over in Drosophila melanogaster Affecting the Preconditions of
Exchange." Genetica 82(3): 203-207.
Powell, J. R. (1992). Inversion Polymorphisms in Drosophila pseudoobscura and
Drosophila persimilis. Drosophila inversion polymorphism. C. V. Krimbas and J.
R. Powell. Boca Raton, Fla., CRC Press: 73-126.
Presgraves, D. C. (2005). "Recombination enhances protein adaptation in Drosophila
melanogaster." Current Biology 15(18): 1651-1656.
116
Priest, N. K., D. A. Roach and L. F. Galloway (2007). "Mating-induced recombination in
fruit flies." Evolution 61(1): 160-167.
Ptak, S. E., D. A. Hinds, K. Koehler, B. Nickel, N. Patil, D. G. Ballinger, M. Przeworski, K.
A. Frazer and S. Paabo (2005). "Fine-scale recombination patterns differ between
chimpanzees and humans." Nature Genetics 37(4): 429-434.
Ram, K. R. and M. F. Wolfner (2007). "Seminal influences: Drosophila Acps and the
molecular interplay between males and females during reproduction." Integrative
and Comparative Biology 47(3): 427-445.
Ranz, J. M., D. Maurin, Y. S. Chan, M. von Grotthuss, L. W. Hillier, J. Roote, M.
Ashburner and C. M. Bergman (2007). "Principles of genome evolution in the
Drosophila melanogaster species group." Plos Biology 5(6): 1366-1381.
Redfield, H. (1966). "Delayed mating and relationship of recombination to maternal age
in Drosophila melanogaster." Genetics 53(3): 593-607.
Reed, F. A. and S. A. Tishkoff (2006). "Positive selection can create false hotspots of
recombination." Genetics 172(3): 2011-2014.
Richards, S., Y. Liu, B. R. Bettencourt, P. Hradecky, S. Letovsky, et al. (2005).
"Comparative genome sequencing of Drosophila pseudoobscura: Chromosomal,
gene, and cis-element evolution." Genome Research 15(1): 1-18.
Robbins, L. G. (1974). "Exchange within Heterozygous Inversions in Drosophila
melanogaster." Genetics 77(1): 105-114.
Roberts, P. A. (1976). The genetics of chromosome aberration. The genetics and biology
of Drosophila. M. Ashburner, E. Novitski and T. R. F. Wright. London ; New
York, Academic Press. 1a: 67-184.
Rockman, M. V. and L. Kruglyak (2009). "Recombinational Landscape and Population
Genomics of Caenorhabditis elegans." Plos Genetics 5(3): e1000419.
Roeder, G. S. (1997). "Meiotic chromosomes: it takes two to tango." Genes &
Development 11(20): 2600-2621.
Roselius, K., W. Stephan and T. Stadler (2005). "The relationship of nucleotide
polymorphism, recombination rate and selection in wild tomato species."
Genetics 171(2): 753-763.
Rozas, J., J. C. Sanchez-DelBarrio, X. Messeguer and R. Rozas (2003). "DnaSP, DNA
polymorphism analyses by the coalescent and other methods." Bioinformatics 19:
2496-2497.
Samollow, P. B., C. M. Kammerer, S. M. Mahaney, J. L. Schneider, S. J. Westenberger, J.
L. VandeBerg and E. S. Robinson (2004). "First-generation linkage map of the
117
gray, short-tailed opossum, Monodelphis domestica, reveals genome-wide
reduction in female recombination rates." Genetics 166(1): 307-329.
Schaeffer, S. W. and W. W. Anderson (2005). "Mechanisms of genetic exchange within
the chromosomal inversions of Drosophila pseudoobscura." Genetics 171(4):
1729-1739.
Schultz, J. and H. Redfield (1951). "Interchromosomal Effects on Crossing over in
Drosophila." Cold Spring Harbor Symposia on Quantitative Biology 16: 175-197.
Schweitzer, M. D. (1934). "Coincidence and interference in Drosophila melanogaster."
Proceedings of the National Academy of Sciences of the United States of America
20: 154-158.
Schweitzer, M. D. (1935). "An analytical study of crossing over in Drosophila
melanogaster." Genetics 20(5): 0497-0527.
Sella, G., D. A. Petrov, M. Przeworski and P. Andolfatto (2009). "Pervasive Natural
Selection in the Drosophila Genome?" Plos Genetics 5(6): e1000495.
Shifman, S., J. T. Bell, R. R. Copley, M. S. Taylor, R. W. Williams, R. Mott and J. Flint
(2006). "A high-resolution single nucleotide polymorphism genetic map of the
mouse genome." PLoS Biol 4(12): e395.
Singer, T., Y. P. Fan, H. S. Chang, T. Zhu, S. P. Hazen and S. P. Briggs (2006). "A highresolution map of Arabidopsis recombinant inbred lines by whole-genome exon
array hybridization." Plos Genetics 2(9): 1352-1361.
Smith, P. D., V. G. Finnerty and A. Chovnick (1970). "Gene Conversion in Drosophila Non-Reciprocal Events at Maroon-Like Cistron." Nature 228(5270): 442-444.
Spencer, C. C., P. Deloukas, S. Hunt, J. Mullikin, S. Myers, B. Silverman, P. Donnelly, D.
Bentley and G. McVean (2006). "The influence of recombination on human
genetic diversity." PLoS Genet 2(9): e148.
Spurway, H. and U. Philip (1952). "Genetics and Cytology of Drosophila subobscura .7.
Abnormal Gene Arrangements in Element A." Journal of Genetics 51(1): 198-215.
Stahl, F. W. (1994). "The Holliday Junction on Its 30th Anniversary." Genetics 138(2):
241-246.
Stahl, F. W., H. M. Foss, L. S. Young, R. H. Borts, M. F. F. Abdullah and G. P.
Copenhaver (2004). "Does crossover interference count in Saccharomyces
cerevisiae?" Genetics 168(1): 35-48.
Steiner, W. W., E. M. Steiner, A. R. Girvin and L. E. Plewik (2009). "Novel Nucleotide
Sequence Motifs That Produce Hotspots of Meiotic Recombination in
Schizosaccharomyces pombe." Genetics 182(2): 459-469.
118
Stephan, W. and C. H. Langley (1998). "DNA polymorphism in Lycopersicon and
crossing-over per physical length." Genetics 150(4): 1585-1593.
Stevison, L. S. (2008). "Hybridization and Gene Flow." Nature Education 1(1).
Stevison, L. S. and M. H. Kohn (2009). "Divergence population genetic analysis of
hybridization between rhesus and cynomolgus macaques." Molecular Ecology
18(11): 2457-2475.
Stevison, L. S. and M. A. F. Noor (2010). "Genetic and Evolutionary Correlates of FineScale Recombination Rate Variation in Drosophila persimilis." Journal of
Molecular Evolution 71(5-6): 332-345.
Strasburg, J. L., C. Scotti-Saintagne, I. Scotti, Z. Lai and L. H. Rieseberg (2009).
"Genomic Patterns of Adaptive Divergence between Chromosomally
Differentiated Sunflower Species." Molecular Biology and Evolution 26(6): 13411355.
Strathern, J. N., B. K. Shafer and C. B. Mcgill (1995). "DNA-Synthesis Errors Associated
with Double-Strand-Break Repair." Genetics 140(3): 965-972.
Stump, A. D., M. C. Fitzpatrick, N. F. Lobo, S. Traore, N. F. Sagnon, C. Costantini, F. H.
Collins and N. J. Besansky (2005). "Centromere-proximal differentiation and
speciation in Anopheles gambiae." Proceedings of the National Academy of
Sciences of the United States of America 102(44): 15930-15935.
Stump, A. D., M. Pombi, L. Goeddel, J. M. C. Ribeiro, J. A. Wilder, A. D. Torre and N. J.
Besansky (2007). "Genetic exchange in 2La inversion heterokaryotypes of
Anopheles gambiae." Insect Molecular Biology 16(6): 703-709.
Sturtevant, A. H. (1913). "The linear arrangement of six sex-linked factors in Drosophila,
as shown by their mode of association." Journal of Experimental Zoology 14(1):
43-59.
Sturtevant, A. H. (1919). "Contributions to the genetics of Drosophila melanogaster. III.
Inherited linkage variation in the second chromosome." Carnegie Inst. Wash.
Publ. 278: 305-341.
Sturtevant, A. H. and G. W. Beadle (1936). "The relations of inversions in the X
chromosome of Drosophila melanogaster to crossing over and disjunction."
Genetics 21(5): 554-604.
Sturtevant, A. H. and T. Dobzhansky (1936). "Geographical distribution and cytology of
"sex ratio" in Drosophila pseudoobscura and related species." Genetics 21(4):
473-490.
Symington, L. S. (2002). "Role of RAD52 epistasis group genes in homologous
recombination and double-strand break repair." Microbiology and Molecular
Biology Reviews 66(4): 630-670.
119
Szostak, J. W., T. L. Orrweaver, R. J. Rothstein and F. W. Stahl (1983). "The DoubleStrand-Break Repair Model for Recombination." Cell 33(1): 25-35.
Tenaillon, M. I., J. U'Ren, O. Tenaillon and B. S. Gaut (2004). "Selection versus
demography: A multilocus investigation of the domestication process in maize."
Molecular Biology and Evolution 21(7): 1214-1225.
van Os, H., S. Andrzejewski, E. Bakker, I. Barrena, G. J. Bryan, et al. (2006).
"Construction of a 10,000-marker ultradense genetic recombination map of
potato: Providing a framework for accelerated gene isolation and a genomewide
physical map." Genetics 173(2): 1075-1087.
Weinstein, A. (1918). "Coincidence of crossing over in Drosophila melanogaster
(Ampelophila)." Genetics 3(2): 135-U2.
Wesley, C. S. and W. F. Eanes (1994). "Isolation and Analysis of the Breakpoint
Sequences of Chromosome Inversion in(3l)Payne in Drosophila melanogaster."
Proceedings of the National Academy of Sciences of the United States of America
91(8): 3132-3136.
White, B. J., C. D. Cheng, D. Sangare, N. F. Lobo, F. H. Collins and N. J. Besansky
(2009). "The Population Genomics of Trans-Specific Inversion Polymorphisms in
Anopheles gambiae." Genetics 183(1): 275-288.
White, M. J. D. and F. H. W. Morley (1955). "Effects of Pericentric Rearrangements on
Recombination in Grasshopper Chromosomes." Genetics 40(5): 604-619.
Winckler, W., S. R. Myers, D. J. Richter, R. C. Onofrio, G. J. McDonald, et al. (2005).
"Comparison of fine-scale recombination rates in humans and chimpanzees."
Science 308(5718): 107-111.
Wolfner, M. F. (2002). "The gifts that keep on giving: physiological functions and
evolutionary dynamics of male seminal proteins in Drosophila." Heredity 88: 8593.
Wong, A. K., A. L. Ruhe, B. L. Dumont, K. R. Robertson, G. Guerrero, et al. (2010). "A
Comprehensive Linkage Map of the Dog Genome." Genetics 184(2): 595-U436.
Wright, S. (1982). "The Shifting Balance Theory and Macroevolution." Annual Review of
Genetics 16: 1-19.
Zeng, K. and B. Charlesworth (2010). "Studying Patterns of Recent Evolution at
Synonymous Sites and Intronic Sites in Drosophila melanogaster." Journal of
Molecular Evolution 70(1): 116-128.
Zenger, K. R., L. M. McKenzie and D. W. Cooper (2002). "The first comprehensive
genetic linkage map of a marsupial: The tammar wallaby (Macropus eugenii)."
Genetics 162(1): 321-330.
120
Zwick, M. E., J. L. Salstrom and C. H. Langley (1999). "Genetic variation in rates of
nondisjunction: Association of two naturally occurring polymorphisms in the
chromokinesin nod with increased rates of nondisjunction in Drosophila
melanogaster." Genetics 152(4): 1605-1614.
121
Biography
I was born on July 10, 1982, in Dallas, Texas. I received a Bachelor of Science in
Biophysics from Centenary College in Shreveport, LA. I obtained a Master of Arts in
Ecology and Evolutionary Biology at Rice University, completing my thesis in May 2007.
As a masters student, I served as a teaching assistant (TA) for Animal Behavior and
Evolution of Genes and Genomes and gave guest lectures for both courses as well as the
Evolution course. I continued on to Duke University where I have been pursuing a PhD
in the Biology Department. As a PhD student, I served as a TA for the courses Genetics
and Molecular Biology and Introduction Biology and guest lectured for the Evolutionary
Genetics course. I have maintained memberships in the Society for the Study of
Evolution and the Society for Molecular Biology and Evolution, as well as served as an
associate member of the Faculty of 1000.
Fellowships, Grants, and Awards:
2010 Katherine Goodman Stern Fellowship, Duke University
2010 Dean's Award for Excellence in Mentoring, Duke University
2009 NSF Doctoral Dissertation Improvement Grant ($15,000)
2009 Walter M. Fitch Student Award Finalist, SMBE Meeting, Iowa State University
2008 Duke Biology Dept. Grant-in-aid of Research ($1,000)
2007 Sigma Xi Grant-in-aid of Research ($400)
2007 Summer Institute in Statistical Genetics Scholarship, University of Washington
($2,500)
2006 Rice EEB Travel Grant, Rice University ($500)
2005 Bob E. & Lore Merten Watt Fellowship, Rice University ($1,000)
2005 Texas Rodeo and Livestock Fellowship, Houston, TX ($2,000)
2005 Outstanding Biophysics Student of the Year, Centenary College
Publications:
1. L.S. Stevison and M.A.F. Noor. 2010. Genetic and evolutionary correlates of finescale recombination rate variation in Drosophila persimilis. Journal of Molecular
Evolution. 71(5): 332-345.
2. Kulathinal R.J., L.S. Stevison, M.A.F. Noor. 2009. The genomics of speciation:
Diversity, divergence and introgression on a genome-wide scale. PLoS Genetics. 5(7):
e1000550.
3. Stevison, L.S. and M.H. Kohn. 2009. Divergence population genetic analysis of
hybridization between rhesus and cynomolgus macaques. Molecular Ecology. 18(11):
2457-2475.
122
4. Fitzpatrick, C.F., Stevison, L.S. and M.A.F. Noor. 2009. Fine-scale crossover rate and
interference along the XR-chromosome arm of Drosophila pseudoobscura.
Drosophila Information Service, 92:27-29.
5. Stevison, L.S. and M.A.F. Noor. 2009. Recombination Rates in Drosophila. In:
Encyclopedia of Life Sciences, Chichester: John Wiley & Sons, Ltd.
6. Stevison, L.S. and M.H. Kohn. 2008. Inferring genetic background of cynomolgus
macaques (Macaca fascicularis). Journal of Medical Primatology. 37(6): 311-7.
7. Stevison, L.S. 2008. Hybridization and gene flow. Nature Education 1(1).
8. Bourgeois G., L.S. Stevison, D. Ortiz-Barrientos, and M.A.F. Noor. 2005.
Characterization of UDP-glycosyltransferase genes in Drosophila pseudoobscura.
Drosophila Information Service, 88: 12-15.
9. Stevison L.S., B.A. Counterman, and M.A.F. Noor. 2004. Molecular Evolution of Xlinked Accessory Gland Proteins in Drosophila pseudoobscura. Journal of Heredity
95 (2): 114-118.
123