Abstract
Inherited microorganisms can selfishly manipulate host reproduction to drive through populations. In Drosophila melanogaster, germline expression of the native prophage WO proteins CifA and CifB cause cytoplasmic incompatibility (CI) in which sperms fertilize uninfected embryos that suffer catastrophic mitotic defects and lethality; however in infected females, CifA rescues the embryonic lethality and thus imparts a fitness advantage to Wolbachia. Despite widespread relevance to sex determination, evolution, and vector control, the mechanisms underlying when and how CI impairs male reproduction remain unknown and a topic of debate. Here we use cytochemical, microscopic, and transgenic assays in D. melanogaster to demonstrate that CifA and CifB proteins of wMel localize to nuclear DNA throughout the process of spermatogenesis. Cif proteins cause abnormal histone retention in elongating spermatids and protamine deficiency in mature sperms of CI-causing males. Protamine-deficient sperms travel to the female reproductive tract together with Cif proteins. In female ovaries, CifA localizes to germ cell nuclei and overlaps with Wolbachia in the nurse cell cytoplasm and the oocyte, however Cifs are not present in late-stage oocytes and the embryo. Moreover, CI and rescue are contingent upon a newly annotated CifA bipartite nuclear localization sequence. Our results reveal a previously unrecognized phenomena in which prophage proteins invade animal gametic nuclei and modify the histone-protamine transition of spermatogenesis.
Introduction
Numerous animal species harbor heritable microorganisms that alter host fitness in beneficial and harmful ways. The most common, maternally-inherited bacteria are Wolbachia, and they typically reside intracellularly in reproductive tissues of both male and female arthropods. Here, they induce reproductive modifications with sex specific effects such as cytoplasmic incompatibility (CI) that can rapidly drive the bacteria to high frequencies in host populations. CI also notably yields important consequences on arthropod speciation [1–3] and vector control strategies [4–9] by causing lethality of embryos from Wolbachia-infected males and uninfected females. As CI is rescued by Wolbachia-infected females with the same strain [10,11], the phenotype accordingly imparts a relative fitness advantage to infected females that transmit the bacteria [12].
Two genes, cytoplasmic incompatibility factors cifA and cifB, occur in a Wolbachia prophage WO module enriched with predicted arthropod functions and homology [13–15]. We previously demonstrated that dual, transgenic expression of cifA and cifB from wMel Wolbachia in Drosophila melanogaster males can induce CI, while single expression of cifA in females rescues CI [16,17]. These results form the basis of the Two-by-One genetic model of CI for several, but not all, strains of Wolbachia [13,15,18,19]. At the cellular level, CI-defining lethality associates with chromatin defects and mitotic arrest within the first few hours of embryonic development. Normally after fertilization, the sperm-bound proteins ‘protamines’ are removed in the embryo and replaced by maternally supplied ‘histones’, resulting in the rapid remodeling of the paternal chromatin [20]. However, during CI, there is a delay in the deposition of maternal histones onto the paternal chromatin, resulting in altered DNA replication, failed chromosome condensation, and various mitotic defects that generate embryonic death [13,21–27].
The incipient, pre-fertilization events in the reproductive tissues that establish CI and rescue remain enigmatic and under recent debate, namely whether (i) the Cifs establish CI during spermatogenesis (Host-Modification model) or embryogenesis (Toxin-Antidote model) and whether rescue occurs or does not occur by CifA and CifB binding in the embryo [28,29]. Here, we develop antibodies to localize Cif proteins from wMel Wolbachia during D. melanogaster gametogenesis and embryogenesis and perform genome integrity measurements of developing sperm across transgenic, mutant, and wild type treatment groups. We describe the following cell biological and gametic chromatin events underpinning CI: (i) CifA and CifB proteins localize to the developing sperm nuclei from early spermatogonium stage to late elongating spermatids; (ii) In mature sperm, CifA associates with sperm tails and occasionally occurs in the acrosome, whereas CifB localizes to the acrosome in all of the mature sperms; (iii) Cifs increase histone retention in developing spermatids and decrease protamine levels in the mature sperms; (iv) During copulation, both Cif proteins transfer with the mature sperm exhibiting reduced protamine levels; (v) In the ovaries, CifA occurs in nuclei of germline stem cells and colocalizes with Wolbachia in the nurse cell cytoplasm; (vi) CifA is absent in the embryos, and thus rescue may occur independently of CifA’s presence; (vii) Both CI and rescue are dependent upon a bipartite nuclear localization signal (bNLS) in CifA that impacts nuclear localization. Taken together, we demonstrate the first case, to our knowledge, in which prophage WO-encoded Cif proteins invade gametic nuclei to modify chromatin integrity at the histone to protamine transition stage. Results resolve debate on the mechanistic basis of CI by generally supporting the Host Modification model in the wMel-infected D. melanogaster system.
Results and Discussion
CifA and CifB invade sperm nuclei during spermatogenesis and spermiogenesis
To evaluate the cellular localization of the Cif proteins, we generated monospecific polyclonal antibodies for visualizing the proteins in reproductive tissues (Fig S1). In D. melanogaster males, the sperm morphogenesis process is subdivided into two events (i) Spermatogenesis including mitotic amplification and meiotic phases, and (ii) Spermiogenesis, a post-meiotic phase. During spermatogenesis, the germline stem cell undergoes four rounds of synchronous mitotic divisions to produce 16 precursor cells called spermatogonia. The spermatogonia then grow and become spermatocytes [30]. After the growth phase, the spermatocytes divide by meiosis and differentiate into 64 haploid round onion spermatids. Post-meiosis, the round sperm nuclei elongate to gradually change their shape accompanied by reorganization of the chromatin during the canoe stage [31]. This results in an individualization complex forming slim, needle-shaped sperm nuclei with reduced volume [30,31]. Elongation and individualization of the spermatids is the final stage of spermiogenesis, after which the mature sperms are transported to the seminal vesicle [30,32].
CifA, but not CifB, localizes in the germline stem cells at the apical end of <8 hr old wMel+ and all cifA and cifB transgene expressing testes (Fig 1, Fig S2). CifA and CifB were both detected in the nuclei of mitotic spermatogonium and spermatocytes. CifA is more abundant than CifB in the spermatogonium stage (Fig S3A, Table S1). In the post-meiotic round onion spermatids, clusters of both CifA and CifB are adjacent to the nuclei. In the elongating canoe-shaped spermatids, CifA and CifB localize to the acrosome, which is the apical position next to the nucleus (Fig 1, Fig S1). CifB is present in all of the spermatid nuclei, whereas CifA is present on an average in 39% of the elongating spermatids per sperm bundle (Fig S3B). During the elongating canoe stage, chromatin-bound histones are typically removed and replaced with protamines to yield compact nuclear packaging and chromatin reorganization of sperms [31]. After nuclear compaction is complete, neither of the Cif proteins are detectable in late spermatid needle-shaped nuclei (Fig 1) and in the mature sperms from the seminal vesicle (Fig S4), indicating either the Cif proteins are fully stripped, or they might not be accessible by the antibodies when the chromatin is tightly compacted [33,34]. To evaluate Cif presence/absence in the mature sperm, we decondensed sperms after isolations from seminal vesicles of <8 hr old males (see methods) and stained them with the respective Cif antibodies. CifA is common along sperm tails in a speckled pattern (Fig 1, Fig S1) and infrequently present in the acrosome region, on average in 45% or 0% of the mature sperm heads depending upon the sampled seminal vesicles (Fig S3B). CifB is present in all of the acrosomal tips of the sperms and not localized to the sperm tails (Fig 1, Fig S1).
During spermatogenesis, Wolbachia are stripped into the cytoplasmic waste bags, which eliminate excess cytoplasmic material during the process of spermatid elongation [23]. Here, we show that some CifA and CifB proteins are also stripped by the individualization complex into the cytoplasmic waste bag (Fig S5). Since Wolbachia are not present in the mature sperms [35,36], these data suggest that the Cif proteins exit Wolbachia cells during spermatogenesis to interact with and modify sperm DNA (see below). Taken together, these findings demonstrate CifA and CifB proteins access and interact with the nuclei of developing Drosophila sperm.
Cifs cause abnormal histone retention and protamine deficiency during spermatogenesis
Since the Cifs localize to sperm nuclei during spermatogenesis, we hypothesized that they may interact with sperm nuclear DNA to impact genome integrity of developing sperm in testes – a central prediction of the Host Modification model of CI [11,29]. At the histone-to-protamine transition stage during spermatogenesis [37], histones normally undergo various post-translational modifications (PTMs) for removal and replacement by smaller protamines for tight chromatin reorganization [37–40]. Lack of PTMs can lead to histone-bound chromatin with improper protamine deposition that causes paternal chromatin defects, male infertility, and embryonic lethality [27,41,42].
Utilizing a core histone antibody, we investigated histone abundance within spermatid bundles at the late canoe stage in CI- and non-CI causing males. We show significantly increased histone retention in both wMel+ and cifAB-expressing testes from <8 hr old males compared to the negative controls (Fig 2A, Fig S6, Table S1). Single transgenic expression showed significantly less histone retaining bundles at this stage (Fig S7A, Table S1). To detect if abnormal histone retention is linked with protamine deficiency in mature sperms, we next used the fluorochrome chromomycin A3 (CMA3) stain that fluoresces upon binding to protamine-deficient regions of sperm DNA [43,44]. We show that mature sperms isolated from wild-type seminal vesicles of young (high CI-inducing) wMel+ males exhibit increased protamine deficiency relative to wMel-males (Fig 2B, Table S1). As a positive control, sperm isolated from <8 hr old males with a Protamine A and B knockout mutant, both in the presence (ΔProt+) and absence (ΔProt-) of Wolbachia, also exhibit a significant increase in fluorescence relative to wMel-(Fig S8A, Table S1). Moreover, a key outcome of higher protamine deficiency in ΔProt+ males with Wolbachia is an increase in wMel+ CI compared to wild type wMel+ CI (Fig S8B, Table S2). These findings indicate an additive effect of the protamine deficiency by Wolbachia and ΔProt knock outs on CI penetrance. Notably, ΔProt-males do not recapitulate CI on their own; thus, the protamine deficiency is not the sole cause of CI but works in conjunction with other modifications by Wolbachia. Consistent with these results, 7-day old wMel+ males that express almost no CI (Fig S8C, Fig S8D, Table S2) exhibit a similarly weak protamine deficiency level to wMel-males, as expected. Moreover, transgene analyses specify both single and dual expression of CifA and CifB cause protamine deficiencies at significantly higher levels than negative controls of wMel- and a non-CI transgene (Fig S7B, Table S1). Since singly expressed Cifs do not cause CI in D. melanogaster [16] (Fig S7C, Table S2), thus the additive effects on the protamine deficiency and/or histone retention due to abnormal PTMs may be required to fully establish CI. Interestingly, post-fertilization delays in maternal H3.3 histone deposition occur in CI embryos [45]; thus, we propose that a genome integrity network involving histones, protamines, and possibly other factors in the gametes may be a common and defining feature underpinning the onset of CI and rescue.
Cifs cause a paternally-transferred protamine deficiency
To investigate if Cif proteins or the genome integrity modifications are paternally-transferred to the female reproductive tract, we performed single male:female pairwise mating of CI and rescue crosses and then dissected female tissues for marker visualizations. After 4 hrs of mating, we isolated the whole uterus (Fig 3A) including the sperm storage organs - spermathecae (SP) and seminal receptacle (SR) - of mated females and performed sperm decondensation, antibody staining, and microscopy. We report two key results. First, both CifA and CifB proteins travel with the sperms transferred to Wolbachia-free females (Fig 3B). Second, the sperm protamine deficiency induced by wMel+ and cifAB-expressing males transfers and persists in the sperms isolated from SP and SR of Wolbachia-free females (Fig 3C, Fig 3D, Table S1), thus indicating transmission of the paternal chromatin-packaging defect to females together with the Cifs. In summary, the findings connect the first CI-associated, paternally-transferred sperm modification with the activity of Wolbachia and the Cifs themselves, and they support the Host Modification model since a lasting Cif-induced sperm modification is established in the testes and transferred with the Cif proteins to the female reproductive tract.
CifA localizes to ovarian germline nuclei and colocalizes with Wolbachia in the cytoplasm
Expression of cifA alone in the ovaries rescues CI [16,17]. Using CifA antibodies, we show that in both wMel+ and cifA expressing ovaries, CifA protein localizes to the cyst DNA in region 1 of the germarium (Fig 4A), indicative of nuclear access in ovaries similar to that in testes (Fig 1). Cystoblast in the germarium undergoes rounds of mitotic divisions to produce oocyte and nurse cells [46,47]. Along egg chamber stages 2-8 of wMel+ females, CifA colocalizes with Wolbachia in the nurse cells and oocyte cytoplasm (Fig 4A). While Wolbachia was still detected in the stage 10 egg chamber, CifA was not present. In cifA expressing females, CifA was primarily detected in the germarium and cytoplasm of the early egg chambers (Fig 4A). Presence of high levels of CifA in the egg is proposed to rescue CI. Thus, we hypothesized that loading of CifA protein and/or maternally-derived proteins/transcripts into the oocyte may enable CifA’s interaction with the embryonic pronuclei and thus rescue. Notably, we did not detect CifA in ~30-60 min old embryos obtained from the rescue cross (Fig 4B), though positive control histone signals were detected colocalizing with host DNA, suggesting that CifA ovarian expression may impart rescue-capable egg modifications that eventually rescue the fertilized embryo from CI sperms. Taken together, CifA localization in ovaries suggests that future research may focus on how interactions with nuclear DNA enable rescue by either altering chromatin remodeling or impacting maternally-derived products that are essential for repairing paternal chromatin defects and consequently embryogenesis. This finding is also consistent with a Host Modification model for rescue.
Both CI and rescue are dependent upon a CifA bipartite nuclear localization signal
Based on cNLS mapping tool for nuclear localization signals [48], CifA amino acids notably harbor a predicted bipartite nuclear localization sequence (bNLS) (Table S3) in the most conserved region of the protein [13,49,50] that is under strong purifying selection [17]. As bNLSs bind to the extended surface groove of nuclear transport protein importin-α, also known as karyopherin-α [51], we hypothesized that sperm nuclear localization of CifA and CI and rescue are dependent on the bNLS. To test this hypothesis, we mutagenized two bNLS sequences with alanine substitutions (aa189-190 for NLS1 (denoted cifA189), aa224-225 for NLS2 (denoted cifA224)), and we additionally deleted the entire bNLS region (cifAΔbNLS) (Fig 5A). The bNLS deletion also corresponds to the weakly predicted catalase-rel domain in CifA [49,50].
Each bNLS mutant, individually and together (cifA189;224), was dually expressed in testes with transgenic cifB to assess CI and singly expressed in females to assess rescue. Transgenic cifA189 expression yielded a significant reduction in CI (~20%) and rescue (~35%) as we previously reported (Fig 5B, Fig 5C) [50]. Conversely, transgenic cifA224 expression showed no significant difference from the controls in either CI or rescue, suggesting this region has little to no impact and the 189 site alone is crucial for CI and rescue phenomenon. However, when both mutants are expressed in cifA189;224 or when the entire bNLS is deleted, CI and rescue are strongly inhibited (Fig 5B, Fig 5C). These results highlight the importance of the bNLS in inducing CI as well as rescue. To determine if the lack of CI induction is due to non-nuclear localization of CifA protein, we used the deletion mutant cifAΔbNLS with wild type cifB to demonstrate that it mislocalizes to the cytoplasm of onion stage spermatids rather than the nuclei (Fig 5D). Additionally, to test if lack of bNLS can impact sperm genomic integrity, we performed CMA3 based protamine-deficiency assay (as described above) and found reduced protamine deficiency levels due to non CI-causing cifAΔbNLS;B line compared to CI-causing cifAB (Fig S9, Table S1). Overall, these data provide previously unknown findings that a functional bNLS and thus CifA nuclear localization are required for CI and rescue.
Conclusions
Discovery of nuclear-targeting Cif proteins in fly gametes establishes new insights on the incipient cell biological and genomic steps that underpin a worldwide symbiotic drive system with relevance to arthropod speciation and pest control [11]. As there are no previous reports of phage proteins invading animal gametic nuclei and impairing the histone-to-protamine transition during spermatogenesis, the findings have significant implications for expanding understanding of the tripartite cell biology of prophage-bacteria-eukaryote interactions to include the realm of animal reproduction. In addition to disentangling the germline events of the Cif proteins that control gametogenesis and embryogenesis, the evidence specifies that the Cif proteins modify sperm genomic integrity and paternally transfer, but they themselves do not bind each other in the embryo. These findings strongly support the Host Modification model of CI.
Material and Methods
Cif proteins antibody development
Conserved amino acid regions of CifA and CifB proteins from wMel Wolbachia were previously identified [49]. Using these regions, monospecific polyclonal antibodies were commercially generated by Pacific Immunology through injection of three synthesized and conserved short (20 aa) peptides for each protein into rabbits. Sequences of peptides were Cys-EYFYNQLEEKDKEKKLTE for CifA, and Cys-DENPPENLLSDQTRENFRR for CifB. The resulting α-CifA and α-CifB antibodies were evaluated using an enzyme-linked immunosorbent assay, and titers were determined to be higher than 1:500,000 for each antibody. Using standard protocols of the Invitrogen WesternDot kit, antibody specificity to wMel+ samples was verified using western blots (1:1000-fold antibody dilution) on protein isolated from homogenates of 50 testes pairs (0-8 hour old males) and 10 ovary pairs (6 day old females) from wMel+ (positive), wMel-(negative control), and cifAB transgenic (positive) flies. The correct size band was only detected from wMel+ and cifAB reproductive tissues (Fig S1). Because the antibodies were generated in the same animal, all subsequent labeling was done with individual antibodies.
NLS identification
CifA amino acid sequences from known Wolbachia and close relatives were input into the cNLS Mapper software [52] to identify putative NLS sequences within each protein (Table S3). cNLS Mapper identifies sequences specific to the importin α/β pathway. A cut-off score of 4 was applied to all sequences. Higher scores indicate stronger NLS activities. Scores >8 indicate exclusive localization to the nucleus, 7-8 indicate partial localization to the nucleus, 3-5 indicate localization to both the nucleus and the cytoplasm, and score 1-2 indicate localization exclusively to the cytoplasm. Predicted NLS sequences are divided into monopartite and bipartite classes. Monopartite NLSs contain a single region of basic residues, and bipartite NLSs contain two regions of basic residues separated by a linker region.
Development of transgenic lines
A cifA variant was synthesized de novo at GenScript and cloned into a pUC57 plasmid as described previously [50]. Site-directed mutagenesis was performed by GenScript to produce the mutants outlined in Figure 5. The cifA189 variant was first described in Shropshire et al. [50] as cifA2. UAS transgenic cifA mutant flies were then generated using previously established protocols [13]. Briefly, GenScript sub-cloned each gene into the pTIGER plasmid, a pUASp-based vector designed for germline-specific expression. Transgenes were then integrated into y1 M{vas-int.Dm}ZH-2A w*; P{CaryP}attP40 attachment sites into the D. melanogaster genome using PhiC31 integrase via embryonic injections by BestGene. At least 200 embryos were injected per transgenic construct, and successful transformants were identified based on red eye color gene included on the pTIGER plasmid containing the transgene. All sequences are reported in Table S4.
Hatch rates
Parental flies were either wild type uninfected (wMel-) or infected (wMel+) with Wolbachia or transgene-expressing with no Wolbachia infection. Uninfected transgenic flies were generated previously [13,17]. Paternal grandmother age was controlled to 9-11 days for expression of naturally high penetrance of wMel CI [53]. Parental transgenic males were generated through crossing nanos-Gal4:VP16 virgin females (aged 9-11 days) to UAS-cif transgenic, uninfected males [53]. Mothers were aged 6-9 days before crossing, while father males first emerged between 0-8 hours were used in hatch rates and tissue collections to control for the younger brother effect associated with lower CI penetrance [13].
Hatch rates were set up as described previously [13,17]. Briefly, a male and female pair was placed in an 8oz, round bottom, polypropylene Drosophila stock bottle with a grape juice-agar plate containing a small amount of yeast placed at the base and secured with tape. These bottles were then placed in a 25°C incubator overnight to allow for courting and mating. The following day, these plates were discarded and replaced with new grape juice-agar plates with fresh yeast. After an additional 24 hours, the plates were removed, and the embryos were counted. The embryo plates were then incubated for 36 hours at 25°C before the total number of unhatched embryos were counted. Any crosses with fewer than 25 embryos laid were discarded from the analyses. Statistical significance (p<0.05) was determined by a Kruskal-Wallis test and Dunn’s multiple test correction in GraphPad Prism 7. All p-values are listed in Table S2.
Immunofluorescence: testes and seminal vesicles
Siblings from the hatch rate (males 0-8 hours) were collected for testes dissection. Tissues were fixed in 4% formaldehyde in PBS-T for 30 min at room temperature and washed in 1x PBS-T three times for 5 min each. Tissues were then blocked in 1% BSA in PBS-T for 1 hour at room temperature. They were then incubated with 1° antibody (α-CifA 1:500 OR α-CifB 1:500) overnight at 4°C rotating. After washing in 1X PBS-T three times for 5 min each at room temperature, they were incubated with 2° antibody (AlexaFluor 592) for 4 hours at room temperature in the dark. Tissues were then washed three times for 5 minutes each in 1X PBS-T and mounted on slides. To stain the nuclear DNA, 0.2mg/mL of DAPI was added to the mounting media before the coverslip was gently placed over the tissue and excess liquid wiped away. Slides were allowed to dry overnight in the dark before viewing on the Zeiss LSM 880 confocal microscope. All images were acquired with the same parameters for each line and processed in ImageJ as described in [54].
Decondensation of Mature Sperm Nuclei
Squashed seminal vesicles collected from male flies (aged 0-8 hours) were treated with 10 mM DTT, 0.2% Triton X-100, and 400 U heparin in 1X PBS for 30 min [33]. The slides were then washed quickly in 1X PBS before immunofluorescence staining (see above).
Immunofluorescence: Histones
Testes from male flies (aged 0-8 hours) were fixed and stained as described above for testes. The tissues were stained with a core histone antibody (MABE71) (1:1000) and imaged on a Keyence BZ-800 microscope. Total late canoe stage sperm bundles were quantified in each testis, and those that retained histones were determined. Ratios of late canoe stage bundles containing histones relative to total bundles from each individual testis were graphed in GraphPad Prism 7. Statistical significance (p<0.05) were determined by pairwise comparisons based on Kolmogorov-Smirnov test and multiple comparisons based on a Kruskal-Wallis test and Dunn’s multiple test correction in GraphPad Prism 7.
Sperm isolation and Chromomycin A3 staining
Seminal vesicles were collected from male flies (aged 0-8 hours for 1-day old flies and 7-days for older flies) and placed on a microscope slide in 1X PBS. Sperm was extracted on the slide using forceps and fixed in 3:1 vol/vol methanol:acetic acid at 4°C for 20 min. Excess solution was then removed, and the slide was air dried. Each slide was treated in the dark for 20 min with 0.25mg/mL of CMA3 in McIlvain’s buffer, pH 7.0, with 10mM MgCl2. Sperm was then washed in 1x PBS, mounted, and imaged using a Keyence BZ-X700 Fluorescence microscope. All images were acquired with the same parameters for each line and did not undergo significant alteration. Fluorescence quantification was performed by scoring fluorescent pixels in arbitrary units (A.U.) within individual sperm heads using ImageJ as per the details described in [54], and calculated fluorescence intensity per sperm head was graphed. Statistical significance (p<0.05) was determined by a Kruskal-Wallis test and Dunn’s multiple test correction in GraphPad Prism 7. All of the experiments involving CMA3 staining were performed at 21°C instead of 25°C. CI hatch rate assays were run in parallel to ensure that CI and rescue phenotypes are not impacted due to changed temperature conditions.
Immunofluorescence: ovaries
Ovaries from females (6 days old) were dissected in PBS on ice and processed as described previously [55]. Tissues were then blocked in 1% BSA in PBS-T for 1 hour at room temperature. They were then incubated with 1° antibody (α-CifA 1:500 OR α-CifB 1:500) overnight at 4°C rotating. For staining Wolbachia, a polyclonal antibody against ftsZ protein generated in rabbit was used at 1:150 dilution (a kind gift from Dr. Irene Newton). After washing in 1X PBS-T three times for 5 min each at room temperature, they were incubated with 2° antibody (AlexaFluor 592) for 4 hours at room temperature in the dark. Tissues were then washed three times for 5 minutes each in 1X PBS and mounted on slides. To stain the nuclear DNA, 0.2mg/mL of DAPI was added to the mounting media before the coverslip was gently placed over the tissue and excess liquid wiped away. Slides were allowed to dry overnight in the dark before viewing on the Zeiss LSM 880 confocal microscope.
Immunofluorescence: embryos
After 24 hours of mating, plates were switched, and embryos were collected every 30 minutes. Embryos were collected in a 100μm mesh basket in embryo wash solution. To remove the chorion, the basket was placed in 50% bleach for 3 min and then rinsed with 1X PBS. The embryos were then transferred to 50:50 4% paraformaldehyde (PFA) and heptane in a microcentrifuge tube and rotated for 20 min at room temperature. Tubes were then removed from the rotator, and the heptane and PFA were allowed to separate before the bottom PFA phase was carefully removed. Methanol was added to the remaining heptane, and the tube was shaken vigorously for 20 seconds before the embryos settled to the bottom and solution was removed. A new volume of methanol was added to the embryos, and they were allowed to settle to the bottom of the tube. Methanol was removed, and all blocking, staining, and imaging steps were carried out for testes and ovary tissues above.
Funding
Digestive Disease Research Center Scholarships S1848284, S1848300, S1883559, Vanderbilt-Ingram Cancer Center Scholarships S1871288, S1848887, S1848952, National Institutes of Health Awards R01 AI132581 and AI143725 to S.R.B., F32 Ruth Kirschstein Postdoctoral Fellowship to B.A.L., and the Vanderbilt Microbiome Initiative to S.R.B.
Author contributions
Conceptualization, B.A.L. and S.R.B.; Methodology, B.A.L., R.K. and S.R.B.; Investigation, B.A.L., R.K., and I.T.R.; Writing – Original Draft, B.A.L., R.K. and S.R.B.; Writing – Review & Editing, B.A.L., R.K., I.T.R. and S.R.B.; Funding Acquisition, B.A.L. and S.R.B.; Resources, S.R.B.; Supervision, S.R.B.
Competing interests
Authors declare no competing interests.
Data and materials availability
All data are available in the main text or the supplementary materials. Unique biological materials will be available upon request.
Supplementary figure legends.
Fig S1. Western blots using Cif antibodies reveal protein bands at the proper size. Western blots were run on protein extracted from both testes and ovaries of wild type infected wMel+, uninfected wMel- and dual cifAB expressing transgenic lines. Cifs are absent in wMel-control and present at accurate size in wMel+ and cifAB lines.
Fig S2. Cifs invade spermatid nuclei in wild type wMel+ testes and absent in wMel-control line. Testes (n=20) from <8 hrs old males of wild type wMel+ and wMel-lines were dissected and immunostained to visualize CifA (green) and CifB (red) during sperm morphogenesis. DAPI stain (blue) was used to label nuclei. CifA and CifB localization patterns in wild type lines are similar to that of transgenic cifAB (Fig 1) and signals are absent in wMel-uninfected control line. The experiment was conducted in parallel to the one shown in Fig 1.
Fig S3. CifA and CifB vary in abundance levels in spermatids and mature sperms.(A) ImageJ-based signal intensity quantification indicates CifA (green) is more abundantly expressed than CifB (red) in the spermatogonium stage of the spermatogenesis. Mean of individual data points with standard deviation is plotted on the graph. Letters indicate statistically significant (p<0.05) differences as determined by pairwise comparisons based on a Mann-Whitney test. P-values are reported in Table S1. (B) In the decondensed mature sperms isolated from seminal vesicles, CifB is present in the acrosomal tip of canoe-shaped spermatids and mature sperm heads, whereas CifA is present in only 40% and 20% of them, respectively. Quantification was performed on the images obtained in Fig 1 data. Each dot represents percentage of Cifs present in spermatids or mature sperms per testes examined.
Fig S4. CifA and CifB are not detectable in the condensed mature sperms in seminal vesicles due to technical limitations. Seminal vesicles (n=20) from <8 hrs old males of transgenic cifAB, wild type wMel+ and wMel-lines were dissected and immunostained to visualize CifA (green) and CifB (red) in the mature condensed sperms (indicated by white arrows). DAPI stain (blue) was used to label nuclei. Absence of both CifA and CifB indicates that proteins are not accessible to the antibodies when the sperm chromatin is condensed and tightly packed.
Fig S5. CifA and CifB are also removed in the cytoplasmic waste bag. Testes (n=20) from <8 hrs old males of transgenic cifAB, and wild type wMel-lines were dissected and immunostained to visualize CifA (green) and CifB (red) in the cytoplasmic waste bags (WB) that are present near the basal end of sperm tail bundles. Some of the Cif proteins strip down in the WB in cifAB line and absent in wMel-control testes. Brightfield is shown to highlight the morphology of sperm tail bundles and waste bags, which are otherwise not visible using Cif antibodies and DAPI stain. The experiment was run in parallel to the ones shown in Fig 1 and Fig S2.
Fig S6. Full uncropped fluorescent images are shown related to Fig 2A.
Fig S7. Individual CifA and CifB expressing lines do not show abnormal histone retention but are protamine deficient.(A) Testes (n=15) from <8 hrs old males of single transgene-expressing lines cifA, cifB and a non CI-causing control gene WD0508 were dissected to quantify spermatid bundles with histone retention (purple) during late canoe stage of spermatogenesis. DAPI stain (blue) was used to label spermatid nuclei. Total spermatid bundles with DAPI signals and those with retained Histones were manually counted and graphed. Single transgenic expressing lines showed significantly less histones similar to the negative control wMel-at the late canoe stage. Vertical bars represent mean and error bars represent standard deviation. Letters indicate statistically significant (p<0.05) differences as determined by multiple comparisons based on a Kruskal-Wallis test and Dunn’s multiple test correction. (B) Mature sperms isolated from seminal vesicles (n=15) of <8 hr old males reared at 21°C were stained with fluorescent CMA3 (green) for detection of protamine deficiency in each individual sperm nucleus. Individual sperm head intensity was quantified in ImageJ (see methods) and graphed. cifA- and cifB-expressing lines showed significantly higher fluorescence indicative of reduced levels of protamines compared to wMel- and WD0508 control lines. Vertical bars represent mean and error bars represent standard deviation. Letters indicate statistically significant (p<0.05) differences as determined by multiple comparisons based on a Kruskal-Wallis test and Dunn’s multiple test correction. All of the P-values are reported in Table S1. The experiments were performed in parallel to the ones shown in Fig 2. (C) CI hatch rate analyses of transgenic male siblings used in CMA3 assays (Figure 2B, S6) validate that CI crosses (black circles) yielded significant less embryonic hatching compared to non CI-inducing ones, when reared at 21°C. Letters to the right indicate statistically significant (p<0.05) differences as determined by multiple comparisons based on a Kruskal-Wallis test and Dunn’s multiple test correction. All of the P-values are reported in Table S2.
Fig S8. Protamine mutants show significantly increased levels of protamine deficiency in mature sperms compared towMel-flies.(A and C) Sperms from the Wolbachia-infected (ΔProt+) and -uninfected (ΔProt-) protamine mutant (w[1118]; ΔMst35B[floxed], Sco/CyO) males exhibit significantly increased CMA3 fluorescence indicative of protamine deficiency compared to both wild type wMel+ and wMel-. 7 day-old wMel+ males that do not cause CI show similar level of protamine levels as of 7 day-old wMel-. Vertical bars represent mean, and error bars represent standard deviation. Letters indicate statistically significant (p<0.05) differences as determined by multiple comparisons based on a Kruskal-Wallis test and Dunn’s multiple test correction. All of the P-values are reported in Table S1. (B and D) CI hatch rate analyses of male siblings used in CMA3 assays (panel A and C) validate that ΔProt+ males with increased sperm protamine deficiency causes stronger (rescuable) CI levels than wMel+. ΔProt-males do not cause CI. 7d old wMel+ do not induce CI in a manner that correlates with their reduced levels of sperm protamine deficiency. Letters to the right indicate statistically significant (p<0.05) differences as determined by pairwise Mann-Whitney test and multiple comparisons calculated using a Kruskal-Wallis test and Dunn’s multiple test correction. All of the P-values are reported in Table S2.
Fig S9. Deletion of CifA bNLS reduces protamine deficiency levels. Mature sperms isolated from seminal vesicles (n=15) of <8 hr old males of transgenic cifAΔbnlsB line shows reduced fluorescence indicative of less Protamine deficiency compared to cifAB. To control for any background confounding effects of nos-Gal4VP16 driver line, wMel-fathers were prior crossed to nos- mothers to generate males with nos;wMel- genotype. CMA3 fluorescence levels of sperms isolated from nos;wMel- males were similar to wMel-wild type lines used in previous assays in this study. Vertical bars represent mean and error bars represent standard deviation. Letters indicate statistically significant (p<0.05) differences as determined by multiple comparisons based on a Kruskal-Wallis test and Dunn’s multiple test correction. All of the P-values are reported in Table S1.
Acknowledgments
The authors thank Jennifer Battle for her assistance in fly collections and staining for sperm integrity assays, Alex Mansueto for his assistance in hatch rates, and both Sarah Bordenstein and Dylan Shropshire for providing helpful feedback on the manuscript. We thank Dr. Irene Newton for providing Wolbachia antibody and Dr. Janna McLean for sending the Protamine mutant fly line for conducting the experiments.
Footnotes
↵* Co-first authors