The intracellular symbiont Wolbachia enhances recombination in a dose-dependent manner

Wolbachia pipientis is an intracellular alphaproteobacterium that infects 40-60% of insect species and is well known for host reproductive manipulations. Although Wolbachia are primarily maternally transmitted, evidence of horizontal transmission can be found in incongruent host-symbiont phylogenies and recent acquisitions of the same Wolbachia strain by distantly related species. Parasitoids and predator-prey interactions may indeed facilitate the transfer of Wolbachia between insect lineages but it is likely that Wolbachia are acquired via introgression in many cases. Many hypotheses exist as to explain Wolbachia prevalence and penetrance such as nutritional supplementation, protection from parasites, protection from viruses, or straight up reproductive parasitism. Using classical genetics we show that Wolbachia increase recombination in infected lineages across two genomic intervals. This increase in recombination is titer dependent as the wMelPop variant, which infects at higher load in Drosophila melanogaster, increases recombination 5% more than the wMel variant. In addition, we also show that Spiroplasma poulsonii, the other bacterial intracellular symbiont of Drosophila melanogaster, does not induce an increase in recombination. Our results suggest that Wolbachia infection specifically alters host recombination landscape in a dose dependent manner. Article Summary The ubiquitous bacterial symbiont Wolbachia is known to alter host reproduction through manipulation of host cell biology, protect from pathogens, and supplement host nutrition. In this work we show that Wolbachia specifically increases host recombination in a dose dependent manner. Flies harboring Wolbachia exhibit elevated rates of recombination across the 2nd and X chromosomes and this increase is proportional to their Wolbachia load. In contrast, another intracellular symbiont, Spiroplasma, does not lead to an increase in recombination across the intervals tested. Our results point to a specific effect of Wolbachia infection that may have a significant effect on infected insect populations.

Wolbachia pipientis is an intracellular alphaproteobacterium that infects 40-60% of insect 19 species and is well known for host reproductive manipulations. Although Wolbachia are 20 primarily maternally transmitted, evidence of horizontal transmission can be found in 21 incongruent host-symbiont phylogenies and recent acquisitions of the same Wolbachia strain 22 by distantly related species. Parasitoids and predator-prey interactions may indeed facilitate 23 the transfer of Wolbachia between insect lineages but it is likely that Wolbachia are acquired 24 via introgression in many cases. Many hypotheses exist as to explain Wolbachia prevalence and 25 penetrance such as nutritional supplementation, protection from parasites, protection from 26 viruses, or straight up reproductive parasitism. Using classical genetics we show that Wolbachia 27 increase recombination in infected lineages across two genomic intervals. This increase in 28 recombination is titer dependent as the wMelPop variant, which infects at higher load in 29 Drosophila melanogaster, increases recombination 5% more than the wMel variant. In addition, 30 we also show that Spiroplasma poulsonii, the other bacterial intracellular symbiont of 31 Drosophila melanogaster, does not induce an increase in recombination. Our results suggest 32 that Wolbachia infection specifically alters host recombination landscape in a dose dependent 33 manner. 34

Introduction: 35
Recombination, the exchange of genetic material during meiosis is thought to be largely 36 beneficial, as it increases the efficacy of natural selection 1,2 . Because of chromosome 37 architecture, loci that are physically linked to each other can interfere with selection such that 38 selection at one locus reduces the effective population size, and therefore the efficacy of 39 selection, at linked loci. This phenomenon, termed "Hill-Robertson interference," means that 40 positive or negative selection at one site can interfere with selection at another site. By 41 allowing loci to shuffle between chromosomes, recombination mitigates Hill-Robertson 42 interference 3 . As a result of this re-shuffling, areas of the genome subject to high 43 recombination rates show higher nucleotide diversity, either because of the inherent 44 mutagenic effect of recombination or by the indirect influence of recombination on natural 45 selection in a population. Overall, a large body of literature supports the assertion that 46 recombination increases efficacy of selection and enhances adaptation in animals, as studied in 47 various Drosophila species 1-3 . 48 One factor that may influence recombination is bacterial infection. For example, injection of 49 flies with the bacterial pathogen Serratia increases recombination post infection 4 . Many 50 Drosophila species are colonized persistently by Wolbachia pipientis, an alpha-proteobacterium 51 within the Rickettsiales and the most common infection on the planet, found in 40-60% of all 52 insects. Wolbachia's prevalence in populations is likely modulated by its reproductive 53 manipulations, induced to benefit infected females 5 . However, this reproductive parasitism 54 alone is not sufficient to explain Wolbachia infection prevalence; indeed there are many 55 recently discovered, insect infecting strains which do not seem to induce any reproductive 56 phenotype at all, suggestive of other potential benefits provided by the symbiont 6-8 . One 57 known benefit is pathogen blocking, where Wolbachia repress the virus replication within the 58 insect host 9-11 . This phenomenon has important implications for the use of Wolbachia in vector 59 control 12 . In addition to protecting its host from pathogens, Wolbachia also generally improves 60 the fitness and fecundity of some hosts, increased recombination? Here we answer these questions using classical genetics in 70 Drosophila melanogaster with different Wolbachia variants and using another intracellular 71 symbiont, Spiroplasma poulsonii. We confirm that Wolbachia significantly increases 72 recombination across two intervals, one on the X and one on the 2 nd chromosome, but we 73 could not detect any effect on the 3 rd chromosome interval queried. In addition, there is a clear 74 correlation between Wolbachia load and recombination events, suggesting Wolbachia itself is 75 the cause of the elevated recombination; clearing the host of Wolbachia restores 76 recombination rate to a basal level while infection with a high-titer variant increases 77 recombination. Another intracellular symbiont, Spiroplasma, does not increase recombination 78 rate, suggesting this phenomenon is not simply due to the presence of a bacterial infection in 79 the gonads, but is Wolbachia specific. These results suggest that Wolbachia specifically elevates 80 host recombination, providing a previously unknown benefit to its host. 81

Wolbachia infection increases host recombination rate 83
We reasoned that if Wolbachia infected flies exhibited increased recombination rate, this 84 would be evident in natural populations. We took advantage of a set of isogenized flies, 85 sampled from a wild-caught population in North Carolina, the Drosophila Genetic Reference 86 Panel 17 . Virgin females from two backgrounds (DGRP-320, infected with Wolbachia and DGRP-87 83, Wolbachia-free) were crossed independently to three different lines carrying chromosomal 88 markers, allowing us to distinguish recombinants along certain genomic intervals based on 89 presence of dominant markers (Fig. 1). carrying ebony and rough. 93 As a control, we also cleared the Wolbachia infection from line DGRP-320 by rearing the flies on 94 tetracycline for 3 generations and then repopulating the extracellular microbiome for 1 95 generation. When we compared Wolbachia infected and tetracycline cleared individuals, 96 controlling for genetic background, we observed a statistically significant increase in 97 recombination in F2 progeny derived from Wolbachia-infected mothers (Fig. 2). Specifically, for 98 the X and 2 nd chromosomes we observed an increase in mutation rate of 6.4% and 6.1%, 99 respectively. No statistically significant effect of Wolbachia infection was observed on the 3 rd 100 . before use in experiments. We looked specifically at the X chromosome intervals as we had 113 already established that Wolbachia significantly increased recombination across that genomic 114 interval (Fig. 2). Again, we observed a significant effect of Wolbachia infection on 115 recombination rate in this experiment (one-way ANOVA; df = 15, χ2 = 15.14, p = 0.015) (Fig. 3). 116 Interestingly, we observed a significant effect of Wolbachia titer on recombination rate -the 117 high titer wMelPop variant increases recombination on the X chromosome in F2 progeny by 118 9.5% compared to the 6.3% observed for wMel (df = 11, χ2 = 12.65, p = 0.044) (Fig. 3). 119

Spiroplasma does not increase host recombination rate 120
We hypothesized that Wolbachia may be a stress on the host cell, increasing recombination 121 rate as a result of increased reactive oxygen species or other immune activation pathways. We 122 therefore reasoned that any bacterial infection may increase recombination rate. To test this 123 hypothesis, we procured Spiroplasma poulsonii MSRO (a gift from John Jaenike), which we used 124 to infect a Wolbachia-free OreR lab stock (Oregon-R-modENCODE, BDSC #25211). We used the 125 same crossing scheme as above to introduce Spiroplasma into the y[1] v [1] background, 126 carrying phenotypic markers on the X chromosome. As a genetic control, we used stock 127 #25211. Counter to our hypothesis, we observed no significant increase in recombination based 128 on Spiroplasma infection (Fig. 4). 129

Discussion: 130
For sexually reproducing organisms, recombination is both a source of genetic diversity within a 131 population and a mechanism by which to decouple differential selection on sites across the 132 chromosome. Therefore, recombination is thought to be beneficial. Here, we observed that 133 Wolbachia infection significantly increased the recombination rate observed across two 134 genomic intervals (for both the X and the 2 nd chromosome). Importantly, two lines of evidence 135 presented here support the hypotheses that this increase is Wolbachia specific: we observed a 136 dose dependency to the recombination rate and we did not identify a significant effect of 137 Spiroplasma infection on recombination rate. Recombination rates vary dramatically across 138 animals, even within a genus, as best illustrated within the Drosophila clade 19,20 . The 139 mechanism behind this difference is not well understood but our data suggest that the 140 symbiont Wolbachia may influence recombination rate of infected Drosophila. This result 141 suggests a previously unknown benefit to Wolbachia infection and may help explain the 142 prevalence of Wolbachia in certain insect populations. 143 The mechanism by which Wolbachia infection elevates recombination is an active area of 144 inquiry in our lab. Wolbachia have an active type IV secretion system that they use to secrete 145 proteins into the host and modulate host cell biology. It is possible that some of these proteins 146 may influence recombination rate directly or indirectly, although no effectors have yet been 147 identified that bind to host DNA. Here we used two different Wolbachia variants to support the 148 hypothesis that Wolbachia increases host recombination rate. However, it is possible that 149 strains outside of the wMel clade do not increase host recombination and a comparative 150 genomic framework could be used to identify loci in Wolbachia that confer the phenotype. 151 Finally, a recent publication suggested Wolbachia wMel infected Drosophila prefer cooler 152 temperatures 21 . Increases in temperature modulate recombination in Drosophila 22 and it is 153 possible that Wolbachia infection elevates host temperature enough to generate an increase in 154 the number of detected recombinants, in a laboratory setting where flies are kept at a constant 155 temperature. 156 Wolbachia is known for its ability to transfer between species. This process can occur through 157 horizontal transmission of the strain or through introgression via hybridization 23 . One 158 particular strain, wRi, has been identified as having globally spread across highly divergent 159 Drosophila species, and in a few cases, instances of introgression between species are known to 160 have facilitated this transfer 23 . Wolbachia facilitates its own maintenance in populations 161 through reproductive manipulations 5 and potentially through mutualistic benefits offered to 162 the host 6,11 . Wolbachia has been shown to facilitate divergence of hosts, through manipulation 163 of sperm-egg compatibility, strengthening species boundaries 24 . It is therefore tempting to 164 suggest, based on these results, that Wolbachia may also increase introgression between 165 species to facilitate their own spread. This hypothesis, however, awaits further research. 166 167

Acknowledgements: 168
We would like to thank anonymous reviewers for their feedback. We also thank MaryAnn 169 Martin for their assistance and guidance throughout the course of this project and Nadia Singh 170 for discussions at earlier stages of data analysis. This work was funded in part by the National selected as recombination trackers for the X, 2 nd , and 3 rd chromosomes in Drosophila 185 melanogaster: #1509, which is marked with yellow (y) and vermillion (v) on the X chromosome; 186 #433, which is marked with vestigial wings (vg) and brown (bw) on the 2 nd chromosome; and 187 #496, which is marked with ebony (e) and rough (ro) on the 3 rd chromosome. Two stocks, one 188 Wolbachia infected and one uninfected, were selected at random from the Drosophila Genetic 189 Reference using an Applied Biosystems StepOne Real-time PCR system and iTaq universal SYBR Green 208 supermix. The Wolbachia primers used are as follows: wspF 5'-CATTGGTGTTGGTGTTGGTG -3' 209 and wspR 5'-ACCGAAATAACGAGCTCCAG -3'. The host primers used are as follows: Rpl32F 5'-210 CCGCTTCAAGGGACAGTATC -3' and Rpl32R 5'-CAATCTCCTTGCGCTTCTTG -3'. The cycling 211 conditions are as follows: 50⁰C for 2 minutes, 95⁰C for 10 minutes, followed by 40 cycles of 212 95⁰C for 30 seconds and 59⁰C for 1 minute. The reaction was carried out in a 96-well plate.

213
Gene expression was determined by the Livak and Pfaffl methods. 214

Recombination Assay 215
To determine if recombination events had occurred, a two-step crossing method was devised, 216 shown in Figure 1. Ten virgin DGRP females aged 3-5 days were housed with ten phenotypically 217 marked males and were allowed to mate for 10 days, after which parentals were cleared from 218 the bottle. Virgin female F1 progeny were collected and crossed to the male parental line in the 219 same ratio as before and allowed to mate for 10 days before being cleared from the bottle. All 220 F2 progeny from this cross were collected and frozen after 10 days of the clearing.