Infectious speciation revisited: impact of symbiont-depletion on female fitness and mating behavior of Drosophila paulistorum

Abstract

The neotropical Drosophila paulistorum superspecies, consisting of at least six geographically overlapping but reproductively isolated semispecies, has been the object of extensive research since at least 1955, when it was initially trapped mid-evolution in flagrant statu nascendi. In this classic system females express strong premating isolation patterns against mates belonging to any other semispecies, and yet uncharacterized microbial reproductive tract symbionts were described triggering hybrid inviability and male sterility. Based on theoretical models and limited experimental data, prime candidates fostering symbiont-driven speciation in arthropods are intracellular bacteria belonging to the genus Wolbachia. They are maternally inherited symbionts of many arthropods capable of manipulating host reproductive biology for their own benefits. However, it is an ongoing debate as to whether or not reproductive symbionts are capable of driving host speciation in nature and if so, to what extent. Here we have reevaluated this classic case of infectious speciation by means of present day molecular approaches and artificial symbiont depletion experiments. We have isolated the α-proteobacteria Wolbachia as the maternally transmitted core endosymbionts of all D. paulistorum semispecies that have coevolved towards obligate mutualism with their respective native hosts. In hybrids, however, these mutualists transform into pathogens by overreplication causing embryonic inviability and male sterility. We show that experimental reduction in native Wolbachia titer causes alterations in sex ratio, fecundity, and mate discrimination. Our results indicate that formerly designated Mycoplasma-like organisms are most likely Wolbachia that have evolved by becoming essential mutualistic symbionts in their respective natural hosts; they have the potential to trigger pre- and postmating isolation. Furthermore, in light of our new findings, we revisit the concept of infectious speciation and discuss potential mechanisms that can restrict or promote symbiont-induced speciation at post- and prezygotic levels in nature and under artificial laboratory conditions.

Figure 1. Schematic presentation of incipient speciation among D. paulistorum semispecies.

Crosses between semispecies give rise in both directions to high hybrid embryonic mortality and complete hybrid male sterility (after as reviewed by [42]).

Figure 2. Presence of germline-associated microbes in D. paulistorum semispecies and their sterile hybrids.

(A) PCR on total DNA extracted from each 15 pairs of ovaries (♀) and testes (♂) from ten day-old Amazonian (AM), Orinocan (OR) semispecies, and from hybrids (AxO) derived from matings between AM females and OR males. Controls are testes of D. simulans infected with wRi Wolbachia (+) and D. melanogaster strain w ¹¹¹⁸ uninfected (−). PCRs were performed with universal 16S rRNA (upper gel) and Wolbachia-specific wsp (lower gel) primer. (B) Wolbachia-specific wsp PCR on total DNA from ten flies of D. paulistorum semispecies and hybrids of Amazonian (AM), Centroamerican (CA), Orinocan (OR) semispecies, and F1 hybrids derived from crossings of AM females to OR males (AxO) and CA females to OR males (CxO). PCR controls are adults of D. simulans infected with wRi Wolbachia (+) and two uninfected strains (−), i.e., D. simulans STC and D. melanogaster strain w ¹¹¹⁸. DNA was extracted from 10-day-old females (f) and males (m).

Figure 3. Distribution of Wolbachia during development of D. paulistorum semispecies.

OR (A–C), AM (D–F,L) and AxO hybrids (G–K) of AM females and OR males. Whole-mount immunostainings with the Wolbachia-specific anti-WSP antibody (yellow/green) were performed on embryos (A,B,D,E, and G,H), testes (J,K), and ovaries (L); DNA is counter-stained with DAPI (blue), or propidium iodide (red) following the protocol of . Transmission electron microscopy of rod-shaped pleomorphic Wolbachia cells (arrows) in the testes of fertile OR (C) and AM semispecies (F), plus sterile AxO hybrid males (I). Symbiont morphology and density in reproductive host tissues corroborate earlier pictures published in , . In OR semispecies high-titer Wolbachia accumulate in early blastodermal stages (A); in further differentiating embryos symbionts selectively target the primordial germ cells of OR (B). Presence of Wolbachia in OR testes associated with developing spermatids (C). In AM semispecies Wolbachia are present at very low-titer levels presumably in somatic and germline cells during blastodermal, gastrulating and late embryonic development (D). Higher magnification of bastodermal AM embryos clearly shows low density of the symbiont (E). Signal intensities are clearly enhanced in F1s of AxO derived from matings between AM females and OR males, suggesting overreplication of the maternally-trasmitted symbiont in hybrids (G,H). Wolbachia immunostainings on cryosections of testes of AxO hybrids in transversal sections (J), and during spermatid development (K). Presence of Wolbachia during D. paulistorum oogenesis (L).

Figure 4. Effect of mild antibiotic treatment on female Drosophila fecundity.

(A) Mean numbers and standard deviations (SD) of mature eggs per pair of ovaries of eight-day-old females (blue bars) and after 15 generations of mild Tetracycline of 0.01% (red bars). Total n = 80. Seven D. paulistorum strains were assayed belonging to the six described semispecies Amazonian (AM), Centroamerican (CA), two Interior (IN and Ll), Andean-Brazilian (AB), Orinocan (OR), and Transitional (TR). D. willistoni (Pan98) is a control strain that was collected in 1998 in Panama naturally infected with wWil Wolbachia . (B,C) Effect of mild antibiotic treatment on D. paulistorum oogenesis in 10-day-old females in (B) untreated OR control and (C) after three generations on 0.1% Rifampicin. (D,E) DAPI staining of egg chambers of (D) untreated and (E) treated OR females showing highly abnormally shaped nurse cells (white arrows).

Figure 5. Mating preferences in combinations between untreated and treated heterogamic pairs of D. paulistorum semispecies.

y-axis represents sexual isolation index (SII); number of mating assay (1–18) is shown on x-axis (corresponding to assay numbers in Table S2). Grey bars indicate untreated controls; black bars indicate assays with Rifampicin treated flies: Tested lines were kept on 0.01% Rifampicin for ten generations; or on 0.1% and 0.2% Rifampicin for five generations. SII and standard error (SE) were determined following . For each array five replicates and 120 matings were scored (12A ♀♀ +12B ♀♀ +12A ♂♂ +12B ♂♂ differentiated by rotated wing clips) for each row, totaling 2,160 matings. Two-tailed P values were calculated by comparing SIIs of untreated and treated pairs of mating choice assays by Fisher's exact test. Statistically significant results are indicated by one, two or three asterisks, i.e., P<0.05; P<0.01 and P<0.001, respectively. Abbreviations: Amazonian (AM); Centroamerican (CA); Orinocan (OR) and Andean-Brazilian (AB) semispecies (lines POA1 and POA10). U =  untreated; T =  treated with antibiotics.

Figure 6. Frequencies of successful heterogamic matings of D. paulistorum females.

Box plots represent distribution of mating frequencies obtained from interstrain mating choice assays with Amazonian (AM), Centroamerican (CA), Orinocan (OR) and Andean-Brazilian (AB; lines POA1 and POA10, respectively) semispecies (females are first named). Tested lines were kept on normal food, or on 0.01% (A,B), 0.1% (C–F) and 0.2% Rifampicin (G–I). Mating frequency was determined by number of successful heterogamic matings out of twelve females in five replicas each (Table S2). Combinations of heterogamic mating choice pairs are indicated by U/U  =  both partners untreated; T/T  =  both treated; T/U  =  females treated, males untreated; and U/T  =  females untreated, males treated. Statistical significant values are indicated by one, two or three asterisks, i.e., P<0.05, P<0.01 and P<0.001, respectively; and statistical outliers by black stars.