Researchers from the University of Washington have used mass spectrometry to identify 63 new seminal fluid proteins and 19 previously unannotated genes in the fruit fly.
The method they developed, which combines isotopic labeling with mass spec-data searches against an entire translated genome, are applicable to a number of taxa and organisms, including human beings, and “illustrate the power of combining proteomics with evolutionary biology to address fundamental questions about reproduction,” the authors write in an article published in the July 29 issue of PLoS Biology describing their work.
In addition to novel proteins, the authors confirmed that 75 proteins that had previously been predicted as seminal fluid proteins, or Sfps, are truly transferred during mating, and showed that Sfps have evolved in a comparatively short timeframe, suggesting the creation of new genes.
Work in the 1960s looking into the sperm of fruit flies found that proteins found in semen played multiple roles in the reproductive process including sperm capacitation, sperm storage and competition, and fertilization. Researchers also found that in some cases, the proteins affect the behavior and physiology of the female fly.
While some Sfps have been identified, the authors estimated that before their research, less than one-third of all of the predicted fruit fly Sfps had been detected in mated female flies.
Geoffrey Findlay, first author on the study and a doctoral candidate in genome sciences at the University of Washington, told ProteoMonitor this week that most of the earlier work in the area involved screening transcripts from male reproductive tissue, and then based on the predicted protein sequences from the transcripts, the proteins secreted and transferred during mating would be computationally determined.
Unlike the transcriptional approach, his and his colleagues’ approach is unbiased, which is “of critical importance, since it is these proteins that are the most likely to influence post-mating processes,” they write in the article.
The most recent estimate of the number of Sfps in Drosophila melanogaster is about 112, Findlay said. Based on their work, he put a new estimate of roughly 200. Because the paper is a proof-of-principle, follow-up work into the role and functions of the proteins identified in the study is needed.
“One thing that we’re starting … to ask [is] ‘If you knock down using RNA interference one of the proteins, does that have an effect on how well the male performs in mating experiments?’” Findlay said.
Distinguishing Transferred Proteins
The first part of their work involved modifying a technique to label the females’ proteins in such a way that they would be distinct from Sfps. They fed the female flies yeast that was grown in media enriched with 15N isotopes to create an “isotopically ‘heavy’ form of the female proteins,” according to the article.
By doing so, they could differentiate Sfps from other proteins under a mass spec by the fact that the Sfps were unlabeled. The researchers chose to label female flies, they said, because mass spec resolution works best for unlabeled peptides and they were interested in identifying male Sfps.
While the labeling strategy is not unique — it was first described by Dutch researchers in 2003 — “what’s new … is that it was applied to a biological problem,” Findlay said.
“One thing that we’re starting … to ask [is] ‘If you knock down using RNA interference one of the proteins, does that have an effect on how well the male performs in mating experiments?’”
In an e-mail, Michael MacCoss, a co-author on the study and an assistant professor of genome sciences at the University of Washington, said, “the original paper used the method specifically to measure changes in protein abundance. In our case, we used the labeling to qualitatively identify proteins within the female that originated from the male during mating.”
That, he said, was the breakthrough technology of his and his colleagues’ work.
The team also used multiple biological replicates of mating experiments with different strains of males. For the control set, the team used DTA-E males, which are spermless and do not produce main cell accessory gland proteins. In two experiments, the team found 11 transferred proteins.
Six experiments using Canton S and tudor males resulted in a set of 138 high-confidence Sfps. Of that, 75 had been previously predicted, but only 19 were previously confirmed to be transferred at mating. Only five previously documented sperm proteins were found by the Washington researchers, “confirming that our protein-preparation protocol effectively selected for soluble, extracellular proteins,” they said.
In all, they found 63 novel Sfps, including 45 found in at least two biological replicates. While the majority fell into the same functional categories including proteases, protease inhibitors, immune-response inhibitors, and lipid metabolism, several new classes were identified, including six members of the odorant-binding protein family.
Using their mass spec data, the authors set out to quantify the relative abundance of each Spf. They calculated a normalized spectral abundance factor — based on the number of spectra associated with each Sfp in a given experiment and standardized by the length of the protein and the total number of Sfp spectra detected in the experiment — for each protein, which was then averaged across all experiments.
While several of the most abundant proteins had been previously characterized Sfps, some novel proteins were also in the top quartile for abundance. The measurements were only approximate, but the data is “the first proteomic-scale view of the relative amount of each transferred Sfp, which may be useful for selecting candidates for further investigation,” the authors wrote.
What’s New in Sperm? 19 Genes
Next, they examined the cross-species evolution of seminal fluid content, comparing their findings from D. melanogaster with the predicted annotations of D. simulans and D. yakuba. They repeated their mating experiments with wild-type strain of each species and found 63 Spfs in all three species, including 19 that had never been identified as seminal fluid components.
Some proteins detected in D. simulans and D. yakuba did not have annotated orthologs for D. melanogaster, and the researchers reasoned that other Sfps may not be annotated as genes in D. melanogaster, which would make them impossible to detect by searching the mass spectra against the annotated proteome.
To overcome this, they constructed a six-reading frame translation of the D. melanogaster euchromatic genome, producing more than 5.8 million potential open reading frames. They applied the Hardklör algorithm to predict which spectra from a tudor experiment “came from male peptides containing only natural abundance isotopes,” the authors said.
Searching against the six-frame database and discarding the spectra that matched an ORF corresponding to an annotated protein, they came up with 23 novel, putative ORFs that did not match any D. melanogaster gene annotation in the database FlyBase. For each putative ORF, rapid amplification of cDNA ends, or RACE, and reverse-transcriptase PCR, the team found 19 unannotated genes.
The discovery of the 19 genes “in a system that has been studied for over a century and for which comparative genomic analysis is now straightforward underscores the limitations of both computational gene prediction programs and the ‘whole-proteome’ databases that are routinely used during shotgun MS analyses,” according to the authors.
In contrast, their approach combining new proteomic methods with the vast amounts of genetic sequence data can provide “considerable insight into the molecular players of a specific biological process,” they said. The approach, they add, should be applicable to many taxa, including worms, plants, rodents, and microorganisms.
Furthermore, their mass spec- and RACE-based method for identifying novel genes should be applicable to other organisms with sequenced genomes “particularly if their genome sizes are no more than 1-2 orders of magnitude greater than the D. melanogaster genome,” including humans, rodents, and Arabidopsis, the researchers said.
In collaboration with researchers at the University of Arizona, Findlay and his co-authors are trying to do the six-frame translation search in mouse. So far, because the mouse genome is better annotated than fruit flies, they “aren’t finding many new things. There are a couple of [findings] we’re following up on right now to see if they’re bona fide new genes,” he said.
Additional work into Sfps could provide new insight into their potential role in the formation of new species, according to the authors. “Determining the transferred Spfs, and subsequently identifying their functions and evolutionary patterns, could therefore be important steps in identifying potential ‘speciation genes,” they said.
The Sfps identified in the article provide a basis for studying long-standing evolutionary questions and identifying specific molecules and functional allelic variants “that affect both sperm competition and male-female co-evolution and conflict,” the researchers conclude. “The challenge ahead will be to apply the combination of genetic, biochemical, and evolutionary methods that have already yielded many insights into Drosophila reproduction to this novel collection of transferred proteins.”