Novel Expression of Drosophila Melanogaster Germline Gene Influences Male Fertility

Abstract

In Drosophila melanogaster, the highest number of novel gene expression phenotypes have been observed in male reproductive tract tissues. Recent discovery of  a novel testis expression by a gene, CG14662, in D. melanogaster but not sister species suggests a  possible newly-evolved association with male fecundity. This study examined whether CG14662 is associated with reduced male fertility in D. melanogaster. with the use of RNAi knockdown. Through a series of fertility assays between males and females, we compared the number of progeny produced in crosses where expression of CG14662 was reduced using RNAi knockdown versus crosses where expression remained uninterrupted. We observed a significant decrease in the number of offspring produced by knockdown crosses compared to control crosses. This indicates that CG14662 has a functional association with male fertility in D. melanogaster that may have been co-opted and then rapidly evolved via directional selection. 

Introduction

Natural selection generates phenotypic variation through the modification of traits in various species, often driving population divergence and speciation. These modifications can lead to the gain or loss of traits. One potential source for the appearance of traits is the discovery of novel genes [1]. Such genes result from various mechanisms, including gene duplication [2, 3] and de novo gene evolution [4]. More frequently, novel phenotypes may arise through novel patterns in gene expression facilitated by mutations in cis-regulatory regions or trans-regulatory elements [5, 6, 7]. Variations within cis-regulatory regions result from sequence mutations in promoter/enhancer binding sites, while the trans-regulatory elements result from mutations within transcription factor proteins, leading to novel expression domains, which can cause new genes to arise [6, 7]. These novel genes are then placed under selective pressures, which influence their survival within a species. Directional selection, which favors individuals that deviate from the population, shifts the frequency of trait variation [9] if advantageous to the individual's fitness. Extreme phenotypes are favored with directional selection, ultimately changing the genetic composition of a population leading to evolutionary changes across a species. 

Another mechanism that can generate gene expression variation or novel genes is co-option. Co-option occurs when natural selection discovers an expression pattern that confers a new function for an existing gene and subsequently specializes that gene towards such function [5]. As the initial change does not involve nonsynonymous mutations to the gene itself, this can allow dramatic phenotypic evolution at rapid timeframes. [5]. Examples of co-option include the evolution of a male pregnancy gene in Gulf pipefish [10] and several venom genes in parasitoid wasps [11] from progenitor genes that do not confer such traits. Co-option has been observed across diverse species, driving the emergence of novel evolutionary traits. 

Genes associated with male reproductive success evolve at faster rates than those expressed in nonreproductive tissues within the Drosophila genus [13, 14, 15]. An ongoing interest is whether some reproductive genes have evolved through co-option in the fruit fly Drosophila melanogaster. The highest number of novel gene expression phenotypes in D. melanogaster itself have been observed to occur within male reproductive tract tissues [1]. Yet, both the function and mechanism of evolution of some genes affiliated with male reproduction remain unknown. In particular is CG14662, a germline gene of unknown function present in D. melanogaster and its sister species D. simulans and D. yakuba. The gene was previously found to exhibit a high testis-specific expression rate in D. melanogaster but nearly none in D. simulans and D. yakuba [1]. As D. melanogaster diverged from D. simulans and D. yakuba recently–approximately 1.3 and 3.5 million years ago respectively [12]–this suggests that a novel expression pattern that potentially relates to an evolving testes-specific function arose in D. melanogaster within a short timespan [1]. This raises the following questions: (1) has gene CG14662 co-opted a novel function associated with reproduction, namely fertility of the testes in the D. melanogaster lineage, and (2) if so, is this gene evolution associated with increased reproductive fitness in D. melanogaster males? This study aimed to address these questions by examining whether CG14662 in D. melanogaster indeed expresses a significant function in male fertility. Confirmation of such a function, combined with the gene’s rapid evolution in the species [1][12] would suggest that co-option was a catalyzing mechanism behind such evolution. Better understanding about the extent of co-option’s role in driving species-specific adaptations, such as in reproduction, provides a stronger framework for discovering novel gene expression patterns within both D. melanogaster and other species broadly.

 One way to study gene function is via knockdown. This refers to any technique intended to reduce the gene’s expression levels, impairing its function. If a gene expresses a critical function in male fertility, then its interruption should likewise impair the ability to produce viable offspring. Knockdown of fertility genes in D. melanogaster can be achieved by inducing RNA interference (RNAi) through insertion of the transgene topi-GAL4 into the male genome. Topi-GAL4 contains the GAL4 gene, whose expression is regulated by the topi gene promoter. When expressed, GAL4 activates testes-specific promoters, driving the expression of the target gene in males to produce dsRNA, or hairpins. The dsRNA then triggers the RNAi mechanism by binding to the endogenous mRNA of the target gene, leading to its degradation and thereby reducing the gene’s expression (Figure 1). 

We examined whether knockdown of CG14662 using RNAi in D. melanogaster males is associated with reduced offspring. Topi-GAL4 was introduced by cross-breeding males carrying CG14662 with virgin females carrying topi-GAL4 to produce topi-GAL4/CG14662 male hybrids. We then tested whether reduced expression of CG14662 influences male fertility conducted fertility assays crossing the  topi-GAL4/CG14662 line and a control line with virgin females to assess whether knockdown males produced fewer offspring than control males.

Materials and Methods

Drosophila maintenance and lines

All flies were gathered from the Bloomington Drosophila Stock Center at Indiana University. Flies were fed a standard fly food diet consisting of yeast, cornmeal, and agar medium. Flies were kept in a 25˚C incubator on a 12-hour light, 12-hour dark cycle (Figure 2). Virgin females from topi-GAL4/TM3, Ser  (Bloomington Stock ID #91776) driver stains were collected and crossed to lines carrying UAS-RNAi constructs for CG14662/CyO (Bloomington Stock ID #57803). These genotypes represent a combination of chromosomes, with topi-GAL4 being a transgene, used to drive the expression of the gene of interest, CG14662. We used balancer chromosomes, TM3, Ser and CyO, to prevent recombination from occurring and maintain recessive lethals in a population [17]. Balancers also help to visualize an offspring’s genotype by expressing distinct wing phenotypes that are easy to identify. Flies containing the balancer TM3, Ser will present with serrated wings, while flies containing the balancer CyO will present with curly wings; flies that do not contain either genotype will have straight wings. The topi-GAL4/TM3, Ser driver expresses only in the male germline. Virgin RAL 307 (Bloomington Stock ID #25179) lines were collected and crossed with F1 control and experimental (knockdown) males. RAL 307 is a fully sequenced isofemale inbred line with no known mutations in the female reproductive system, allowing for the detection of sequence variation in offspring. 

Mating Assay

On day 1, three virgin topi-GAL4/TM3 Ser driver females were crossed to three virgin CG14662/CyO hairpin males per vial to produce F1 males (Table 1). Adult flies were kept in vials to mate for five days and then were discarded. Virgin male offspring were collected starting 10 days after mating crosses were set up and separated into different vials according to genotype (Figure 3). The experimental/knockdown genotype was labeled topi-GAL4/CG14662 (straight-winged) and the control genotype was labeled topi-GAL4/CyO (CyO). Experimental and control F1 males were aged 1-3 days for use in fertility crosses. 

 Fertility Assay

Virgin RAL 307 females were collected and aged 2-3 days to ensure virginity before setting up fertility crosses with experimental and control F1 males (Figure 4). Each cross consisted of one virgin RAL 307 female and one F1 virgin male, either experimental or control, per vial. For optimal results, 10 experimental crosses and 10 control crosses were set up and repeated 10 times to obtain a total of 200 fertility crosses. Each pair of adult flies were left in vials to mate for 24 hours and then discarded. Fertility success between the crosses was determined by the number of offspring in each vial, which was counted 16 days after the control and experimental crosses were set up. Drosophila larvae are known to remove mold present in fly food [16]; therefore, we automatically categorized vials with mold as having zero offspring without counting. Vials that showed zero offspring indicated males with decreased fertility and recorded in our data as “zeros.”

Data analysis

We used the Mann-Whitney U test through the base R function “wilcox.test()” to compare offspring viability between experimental and control groups because the data did  not follow a normal distribution. A Mann-Whitney U Test assesses whether differences in the number of offspring between experimental and control groups are significant. This would ideally test for significant differences in fertility. However, it is possible that some vials with zero offspring did not reflect low fertility but failure of fertile pairs to mate within 24 hours. We therefore applied two Mann-Whitney U tests to assess whether significance differs when accounting for this factor. The first test omitted vials that showed zero offspring by count only. The second test omitted zero-offspring vials based on both count and presence of mold, under the assumption that all zeroes could be due to failed mating.

As removing zero-offspring vials from analysis may lead to artifact results due to sample size imbalance between batches, we also fitted a generalized linear mixed model (GLMM) using the R package “lme4” [18] to assess potential sources of variation within the revised dataset. A GLMM predicts the association between a dependent variable and independent variables (fixed effects) while accounting for potential confounding variables and random effects within the dataset. To predict the number of offspring produced, we included cross-type as a fixed effect and batch as a random effect.

Prior to each analysis, the R package “janitor” [19] was used to format the data by cleaning misformatted column names. Statistical significance was determined using a predetermined p-value (p < 0.05).

Results 

Our assays successfully produced a total of 94 vials with offspring out of 102 vials in the control groups and 95 out of 106 vials in the experimental groups. The control groups showed a higher average quantity of offspring (Figure 5) and nearly half the number of zeroes (Table 2) compared to the experimental groups.

The first Mann-Whitney U test found no significant association between control and experimental males (p-value = 0.4978, W = 2246). The second test yielded a significant association between experimental and control males (p-value = 0.00367, W = 5541.5). The higher ranked sum (W) calculated in the second Mann–Whitney U test suggests that experimental males had more instances where zero offspring were produced when compared to control males (Table 2). 

The GLMM showed significant negative association between cross-type and fertility, with the experimental cross producing fewer offspring than the control cross (cross-type coefficient = -3.659, p-value = 0.00248). 

Discussion

This study sought to address the association between CG14662 and male fecundity, specifically if knockdown of CG14662 in Drosophila melanogaster males is associated with reduced male fertility. We found significantly less offspring was produced by reproducing males where expression of CG14662 was disrupted by RNAi knockdown (topi-GAL4/CG14662) compared to males where knockdown of the gene did not occur (topi-GAL4/CyO). While our first Mann-Whitney U test, which omitted zero-offspring vials that did not present mold, showed a non-significant association between gene CG14662 and male fecundity, significance was regained when all zeroes were removed completely in the second test and our GLMM. Reduction in offspring numbers upon gene knockdown indicates CG14662 plays a role in male fecundity in D. melanogaster. This finding builds upon previous work [1] that identified the gene as possessing a novel testes function by shedding light on the identity of this function as fertility-related.

The exhibition of a distinct expression pattern by CG14662 absent in sister species [1] suggests that this function rapidly evolved in D. melanogaster testes, which indicates selective pressures acted on regulatory regions within the gene. Natural selection affects phenotypic variation in two primary ways: stabilizing selection or directional selection [20]. Stabilizing selection acts against individuals that exhibit extreme variations from the average population phenotype, promoting the stability of that phenotype [8]. Directional selection favors individuals that deviate from the population mean, thereby resulting in phenotypic change within the population [9]. Novel expression of CG14662 possibly arose from mutations in regulatory regions that led to the discovery of a new advantageous fertility function, leading to a fitness advantage for D. melanogaster males carrying such mutations paired with increased expression levels. This adaptation was likely further shaped by directional selection; stabilizing selection maintains an existing stable phenotype throughout the population and would have disfavored novel variants. Mutations in regulatory regions of the gene along with directional selective pressures likely resulted in the co-option of a new expression of CG14662 in D. melanogaster. Expanding on the molecular structure of the gene could unveil its involvement in various mechanisms within D. melanogaster. Further research should also explore specific mechanisms by which CG14662 influences male fertility, including its role in spermatogenesis or interaction in genetic pathways. These would shed further insight into the context of this gene’s evolution and how it currently mediates variation in reproduction-related phenotypes, adding to our body of knowledge about how rapid selection can occur at the molecular level. This will strengthen the broader framework for discovering novel gene expression patterns.

The main limitation of our study was our inability to ensure successful mating was always identified. Since our categorization of zeroes could not rule out crosses that simply did not mate within the 24 hour timeframe, there was ambiguity as to whether zero offspring in our data is reliable to use in assessing fertility. While we were able to account for this problem by removing zeroes from analysis, this came at the cost of removing every potential instance of sterile males which probably skewed the distribution of progeny count per vial in both groups to higher averages. While our analyses nevertheless showed a significant reduction in progeny count based solely on males that successfully reproduced, the magnitude of that reduction is probably an underestimate. In addition, since we assumed that mold is indicative of no larvae, it is possible that some zeroes were false negatives. Future studies should document when mating occurs and how quickly to distinguish  vials that yield zero offspring due to low fertility from vials with unsuccessful mating, and verify the absence of larvae when mold is present. Larger sample sizes would also assist in mitigating the impact of zeroes on the balance of data in studies that necessitate their omission like ours, as well as give a more accurate representation of the D. melanogaster species as a whole.

Figures & Tables

The outcome of the crosses done between topi-GAL4/TM3, Ser virgin females and CG14662/CyO virgin males. The experimental males are topi-GAL4/CG14662 and the control males are topi-GAL4/CyO. 

 Experimental groups produced 50 vials with zero offspring while control groups produced 27 vials. The proportion of zero offspring produced in experimental crosses is higher (0.2647059) compared to control (0.4716981).

RNA interference (RNAi) is used to knockdown gene CG14662 in D. melanogaster males. The genotype topi-GAL4/TM3, Ser contains a GAL4 gene, whose expression is regulated by the topi gene promoter. When expressed, GAL4 transcription factor targets the UAS binding site and recruits RNA Polymerase II to drive the expression of gene CG14662 in knockdown males to produce dsRNA. The dsRNA is cut up by dicer and small fragments bind to the endogenous target mRNA using the RNA-induced silencing complex (RISC), leading to the degradation of CG14662.

Flies were kept in vials and fed a standard fly food diet consisting of yeast, cornmeal, and agar medium and kept in a 25˚C incubator on a 12-hour light, 12-hour dark cycle. Each assay used this vial containing the fly food diet.

One virgin topi-GAL4/TM3, Ser female was crossed to one virgin CG14662/CyO male per vial to produce F1 experimental (topi-GAL4/CG14662) and control (topi-GAL4/CyO) males to be used in the fertility assays.

For the fertility assay, one virgin RAL 307 female was crossed to one experimental (a) male and one control (b) male per vial. Each pair of adult flies were left in vials to mate for 24 hours and then discarded; offspring were counted 16 days later.

Control groups have a higher median of 8, Q1 equal to 2.25, Q3 equal to 15.75, and an IQR equal to 13.5. While, experimental groups have a median of 4, Q1 equal to 0.00, Q3 equal to 12.50, and an IQR equal to 12.5.

Author's Note

My time at Dr. David Begun’s Lab at UC Davis allowed me to explore various evolutionary processes and gain a deeper understanding of how changes in gene expression generate phenotypic variation in the common fruit fly, Drosophila melanogaster. I have always been interested in evolution and recently I have found myself in the field of genetics looking at various biological mechanisms responsible for speciation and population divergence. With that, I hope this article exposes more individuals to the field of evolution and the never ending discovery of the gene pool.

References

1. Cridland JM, Majane AC, Sheehy HK, Begun DJ. 2020. Polymorphism and Divergence of Novel Gene Expression Patterns in Drosophila melanogaster. Genetics. 216(1): 79-93.
2. Ohno S. 1970. Evolution by Gene Duplication. Springer-Verlag, New York.
3. Zhang J. 2003. Evolution by gene duplication: an update. Science Direct. 18(6): 292–298. 
4. Begun DJ, Lindfors HA, Thompson ME, Holloway AK. 2006. Recently evolved genes identified from Drosophila yakuba and D. erecta accessory gland expressed sequence tags. Genetics. 172(3): 1675–1681.
5. True JR & Carroll SB. 2002. Gene Co-Option in Physiological and Morphological Evolution. Annual Review of Cell and Developmental Biology. 18: 53-80. 
6. Emerson JJ & Li WH. 2010. The genetic basis of evolutionary change in gene expression levels. The Royal Society. 365(1552): 2581–2590. 
7. Osada N, Miyagi R, Takahashi A. 2017. Cis- and Trans-regulatory Effects on Gene Expression in a Natural Population of Drosophila melanogaster. Genetics. 206(4): 2139-2148. 
8. Romero I, Ruvinsky I, Gilad Y. 2012. Comparative studies of gene expression and the evolution of gene regulation. Nature Reviews Genetics. 13: 505-516. 
9. Sanjaka JS, Sidorenko J, Robinson MR, Thornton KR, Visscher PM. 2018. Evidence of directional and stabilizing selection in contemporary humans. PNAS. 115(1): 151-156. 
10. Harlin-Cognato A, Hoffman EA, Jones AG. 2006. Gene Cooption Without Duplication During the Evolution of a Male-Pregnancy Gene in Pipefish. PNAS. 103(51): 19407-19412. 
11. Martinson EO, Mrinalini, Kelkar YD, Chang CH, Werren JH. 2017. The evolution of venom by co-option of single-copy genes. Current Biology. 27(13): 2007–2013.e8.
12. Obbard DJ, Maclennan J, Kim K-W, Rambaut A, O’Grady PM, Jiggins FM. 2012. Estimating Divergence Dates and Substitution Rates in the Drosophila Phylogeny. Molecular Biology and Evolution. 29(11):3459–3473.
13. Begun DJ, Whitley P, Todd BL, Waldrip-Dali HM, Clark AG. 2000. Molecular population genetics of male accessory gland proteins in Drosophila. Genetics. 156(4): 1879–1888.
14. Swanson WJ & Vacquier VD. 2002. The Rapid Evolution of Reproductive Proteins. Nature Reviews Genetics. 3: 137-144. 
15. Zhang Z, Hambuch TM, Parsch J. 2004. Molecular Evolution of Sex-Biased Genes in Drosophila. Molecular Biology and Evolution. 21(11): 2130-2139.
16. Yang D. 2018. Carnivory in the larvae of Drosophila melanogaster and other Drosophila species. Scientific Reports. 8, 15484. 
17. Cox RD, Brown SDM. 2003. Rodent models of genetic disease. ScienceDirect 13(3): 278-283.
18. Bates D, Mächler M, Bolker B, Walker S. 2015. Fitting Linear Mixed-Effects Models Using lme4. Journal of Statistical Software. 67(1): 1-48.
19. Firke S. 2024. janitor: Simple Tools for Examining and Cleaning Dirty Data. R package version 2.2.0.9000.
20. Kingsolver JG and Diamond SE. 2011. Phenotypic Selection in Natural Populations: What Limits Directional Selection? The American Naturalist. 177(3): 346-357.