Background

Wolbachia constitutes a diverse group of maternally inherited endosymbionts belonging to the Alphaproteobacteria [1, 2]. To date, 16 different Wolbachia supergroups (A–F and H–Q) have been described [3]. Genomic approaches have been used to classify some of these Wolbachia supergroups as different species [4, 5], although this is still a rather controversial issue [6]. Supergroups A and B are widely spread across many arthropod taxa [7], C and D are found exclusively in filarial nematodes [8] whereas E is found in springtails [9]. Other Wolbachia supergroups are found in different host species. For instance, F supergroup comprises Wolbachia from termites, weevils, true bugs and scorpions [10, 11]. Different genetic markers have been employed to classify Wolbachia in supergroups including the 16S ribosomal RNA (16S rRNA) and the Wolbachia surface protein (wsp) genes [12,13,14]. More recently, two multi locus sequence typing (MLST) approaches and a wsp-based system have been developed for genotyping in addition to phylogenetic and evolutionary analyses of this bacterial taxonomic group [15, 16]. The discovery of supergroups (H-P) is mainly based on the full-length sequence of the 16S rRNA and additional gene markers, such as groEL (heat-shock protein 60), gltA (citrate synthase) and ftsZ (cell division protein) [17]. This is in most cases associated to the lack of positive results (PCR amplification and/or sequencing) of any of the MLST genes in diverse supergroups (see also [3, 18,19,20]).

Wolbachia infections have been reported in the somatic tissues of a wide range of arthropod hosts [21, 22] and filarial nematodes [8, 23, 24]. However, they are mainly known to reside in reproductive tissues and organs [21, 25,26,27]. In arthropods, Wolbachia often behave as reproductive parasites by manipulating the host reproduction to enhance its own vertical transmission [28] giving a reproductive advantage to infected individuals and spreading Wolbachia through natural populations [29,30,31,32,33]. A wide range of reproductive alterations induced by Wolbachia infection has been described in host species, including cytoplasmic incompatibility (CI), parthenogenesis, feminization and male-killing (MK) [1, 2, 34,35,36]. CI is the most common phenotype induced by Wolbachia and is characterized by the induction of an embryonic lethality causing mating incompatibility in the crosses between Wolbachia infected males and uninfected females (unidirectional CI). Similar physiological incompatibilities are observed in crosses between individuals infected by mutually-incompatible Wolbachia strains (bidirectional CI) [35, 37,38,39]. Parthenogenesis is another well-documented Wolbachia-induced mechanism in haplo-diploid species by which the bacterium ‘forces’ unfertilized eggs to develop into females rather than males [40, 41]. Wolbachia-mediated feminization is characterized by the development of infected males into fertile females. This phenotype has been observed in both insects and isopods [42,43,44]. MK is expressed as male lethality during development resulting in a female-biased sex ratio [36, 45, 46]. MK can be elicited early during the embryonic development, or late in the larval or pupal stage [47]. MK is not limited to Wolbachia, as this phenomenon has been described for at least five clades of bacteria associated to the reproductive system (Additional file 1).

Wolbachia-host symbiotic associations are rather complex, since this reproductive microorganism can also be associated with a variety of additional phenotypes. These traits include the protection of insect hosts against pathogens and parasites [48,49,50,51,52,53], mating preference [54,55,56] and the response to olfactory cues [57]. The unique biology of Wolbachia has been explored for the development of novel strategies for the control of pests and diseases [33, 58,59,60,61]. For example, it has been shown that the Incompatible Insect Technique (IIT), which is based on the mechanism of Wolbachia-induced CI, can be used alone or in combination with the Sterile Insect Technique (SIT) to suppress populations of insect pests of agricultural, veterinary or human health importance [58, 62,63,64,65,66,67]. Wolbachia-induced MK has also been suggested as a tool for insect pest control [68, 69].

The South American fruit fly, Anastrepha fraterculus Wiedemann (Diptera: Tephritidae) is a complex of cryptic species [70,71,72,73] that is distributed in subtropical and temperate regions of the American continent, covering a wide geographical range from the United States of America to Argentina [74,75,76]. Recent studies focused on the elucidation of species from the A. fraterculus complex have followed an integrative approach. These scientific works addressed this taxonomic issue using different strategies based on morphology [73, 77], behavior and reproductive isolation [76, 78,79,80,81], and cytology and genetics [82,83,84,85,86]. Based on mating compatibility studies [87,88,89] and population genetic analysis [90, 91], a sinof the A. fraterculus complex was identified in Argentina and southern Brazil. This taxon has been named A. fraterculus sp.1 by Selivon et al. [82] and Brazilian-1 morphotype by Hernández-Ortiz et al. [73]. The presence of Wolbachia has been described in Brazilian populations and in laboratory colonies of A. fraterculus from Argentina and Peru [79, 82, 92]. In addition, a recent publication [93] showed the presence of Wolbachia in A. fraterculus populations belonging to different morphotypes across America.

In the present study, we initiated a comprehensive study to detect and characterize Wolbachia infections in A. fraterculus from Argentina including a laboratory colony and three wild populations. After the detection and molecular characterization of the symbiont, we raised the hypothesis that Wolbachia infection may be associated with the induction of reproductive phenotypes, which could be a contributing factor in the speciation of A. fraterculus species complex. This hypothesis was tested with a series of crossing experiments assessing pre- or post-mating incompatibility, and these phenomena are discussed.

Materials and methods

Samples collection and DNA isolation

Wild A. fraterculus individuals were obtained from infested fruits collected in three different localities of Argentina: Horco Molle (Tucumán province); Villa Zorraquín (Entre Ríos province) and Puerto Yeruá (Entre Ríos province) (Table 1). Larvae and pupae obtained from each locality were maintained under standard laboratory conditions [94, 95] until emergence. In addition, individuals from the laboratory colony reared at IGEAF (INTA-Castelar, Buenos Aires, Argentina) were obtained, processed and stored under the same conditions until DNA extraction (Table 1). A. fraterculus IGEAF strain was established in 2007 with approximately 10,000 pupae from the semi-mass rearing colony kept at Estación Experimental Agroindustrial Obispo Colombres, San Miguel de Tucumán, Tucumán, Argentina [96] and maintained to date (70 generations) under artificial rearing.

Table 1 Sampling locations and number of individuals used for Wolbachia characterization
Full size table

All insects were washed with TE buffer (10 mM Tris–HCl, 10 mM EDTA, pH 8) and stored at − 20 °C until DNA extraction. Total DNA was individually isolated from adult flies (whole body) based on the protocol described by Baruffi et al. [97]. The quality of DNA samples was tested by electrophoresis in agarose gels 0.8% w/v in buffer TBE 0.5 X and stained with ethidium bromide [98]. Images were captured with an UVP reveler (Fotodyne Inc. Hartland, WI, USA). Quality and quantity of DNA samples were also analyzed with Nanodrop 1000 (Thermo Scientific).

Detection and genotyping of Wolbachia strains

Wolbachia detection was based on the amplification and sequencing of a 16S rRNA gene fragment (438 bp) using the Wolbachia-specific primers wspecF and wspecR [99] and a wsp gene fragment (590 to 632 bp long) using primers 81F/691R [13]. The sequence characterization of a wsp gene from each Wolbachia-nucleotide variant found in this study was performed by wsp hypervariable regions (HVRs) analysis using the Wolbachia MLST database (pubmlst.org/Wolbachia). HVR alleles were determined based on comparisons among available translated nucleotide sequences [100]. Laboratory colony (37 individuals; 24 females, 13 males) and insects from natural populations (39 individuals; 22 females, 17 males) were analyzed. A subset of DNA samples (Table 1) were genotyped using the MLST scheme proposed by Baldo et al. [15] to characterize Wolbachia. Partial regions of gatB (aspartyl/glutamyl-tRNA(Gln) amidotransferase, subunit B), coxA (cytochrome c oxidase, subunit I), hcpA (conserved hypothetical protein), fbpA (fructose-bisphosphate aldolase) and ftsZ genes were amplified, using the standard protocols provided in the Wolbachia MLST database [15]. PCR products were purified using a Wizard SV Gel and PCR Clean-Up System (Promega) and forward and reverse sequences were obtained using an Abi 3130XL Genetic Analyzer (Applied Biosystem, SIGYSA- INTA, Argentina). Sequences were manually edited and aligned using Bioedit 7.0.9.0 [101] and Staden Package [102].

A Neighbor-joining tree was reconstructed based on the concatenated MLST datasets (gatB, coxA, hcpA, fbpA and ftsZ; 2079 bases long) using sequences generated in the present study and a batch of representative nucleotide sequences belonging to A, B and D Wolbachia supergroups published by Baldo and Werren [103] available through the Wolbachia MLST webpage. The phylogenetic tree was constructed using Mega Version 5.1 software [104] based on the Jukes and Cantor [105] genetic distance model after 1000 bootstrap resamples.

Seven additional gene markers previously described for the genotyping of Wolbachia were utilized to distinguish Wolbachia genetic variants infecting the A. fraterculus Argentinean populations. Partial regions of groEL and gltA [17], dnaA (Chromosomal replication initiator protein) [106], aspC (aspartate aminotransferase) atpD (ATP synthase) sucB (dihydrolipoamide succinyltransferase) and pdhB (E1 component of the pyruvate dehydrogenase complex) [16] genes were amplified using primer sequences and PCR conditions described by the cited authors. At least three individuals of each A. fraterculus IGEAF strain harboring different genetic variants of Wolbachia were analyzed.

Detection of other reproductive symbionts

A. fraterculus DNA samples were also screened for the presence of other reproductive symbionts (Spiroplasma sp. [107], Cardinium sp. [108], Rickettsia sp. [109], Arsenophonus sp. [110] and Hamiltonella sp. [111]) using the primers and conditions described by the authors cited in Table 2. In case of successful amplification, PCR products of expected size (according to the previously published works) were purified and sequenced.

Table 2 Additional primers used for the detection of Wolbachia and other symbionts
Full size table

New DNA sequences were deposited in public databases as is described in “Availability of data and material” section.

Establishment of A. fraterculus Af-Cast-1 and Af-Cast-2 strains

At least 20 single pairs (female and male) from A. fraterculus IGEAF strain (IGEAF, INTA Castelar, Argentina) were maintained in standard conditions (25 °C temperature; 50% humidity and 12:12 Light: dark photoperiod) from the day of emergence to ensure that flies were virgin, since A. fraterculus reaches sexual maturity between 4 and 10 days after emergence [112]. At day 10 after emergence, egg collection devices (described by Vera et al. [94]) were offered to each couple continuously, either for a month or until at least 100 eggs were obtained. Total DNA was individually extracted from the parents of the families to determine the Wolbachia wsp nucleotide variant present in each one of them by PCR and sequencing of the amplicon as described above. Families sharing the same Wolbachia nucleotide variant (either wAfraCast1_A or wAfraCast2_A) were pooled and maintained as discrete strains under laboratory conditions of rearing. These A. fraterculus strains were named Af-Cast-1 and Af-Cast-2.

Evaluation of Wolbachia genomic integration in A. fraterculus

The two laboratory strains of A. fraterculus (Af-Cast-1 and Af-Cast-2 strains) were treated with antibiotics. Eggs were deposited in plastic containers with larval diet [95] containing 0.01% rifampicin (Richet). After adult emergence, Wolbachia infection status was assessed by wsp and 16S rRNA based PCR assays using the specific primers described above. DNA extracted from individuals of the Af-Cast-1 and Af-Cast-2 A. fraterculus strains reared without antibiotic treatment was used as a positive control.

Singly-infected A. fraterculus strains (Af-Cast-1 or Af-Cast-2) were maintained in our laboratory under standard rearing conditions [94].

Mating experiments

In order to detect whether the presence of Wolbachia is associated with reproductive isolation, we carried out mating tests crossing A. fraterculus strains Af-Cast-1 and Af-Cast-2. Pre-zygotic isolation (which occurs before fertilization of gametes) as well as post-zygotic isolation (which occurs after fertilization) tests were performed as described below.

Pre-zygotic isolation test

Individual crosses in every possible combination (i.e., female x male: Af-Cast-1 x Af-Cast-1, Af-Cast-1 x Cast-2, Af-Cast-2 x Af-Cast-1 and Af-Cast-2 x Af-Cast-2) were carried out in no-choice mating arenas under laboratory conditions following standard procedures [113]. Each arena consisted of a 1 L plastic cylindrical container with a screen lid. The day before the test, 10 day-old (sexually mature) and virgin males were individually transferred to the mating arenas with no food or water. The next morning, under semidarkness, 15 day-old (sexually mature) and virgin females were released in the experimental arenas. Once the experiment was set up, the room lights were turned on (8:30 am). Experiments were conducted under laboratory conditions (T: 25 ± 1 °C and 70 ± 10% RH). The number of replicates was 59 ± 5 per cross type. The number of mated couples (percentage of mating), latency to mate and mating duration time were recorded for each type of cross. After the mating trial was completed, flies were removed from the mating arenas. Mated flies were preserved for post-zygotic tests (see below) whereas unmated flies were stored at − 20 °C.

Post-zygotic isolation test

Mated couples were maintained with food and water under controlled conditions and allowed to lay eggs in an artificial egg-laying device. Eggs were collected, placed on a piece of black filter paper, counted and transferred to Petri dishes (3 cm diameter) with larval diet [94, 95]. The Petri dishes were placed in a larger container on top of a layer of vermiculite (pupation substrate). After 5 days, the number of hatched eggs was recorded. After all developing larvae had exited the diet and pupated in the vermiculite pupae were collected, counted and placed under controlled conditions until emergence. The number and sex of emerged adults from each cross were recorded. Once the post-zygotic test ended, parental flies were stored at − 20 °C and subsequently checked for the presence of Wolbachia (using the wsp-based PCR assay described above).

Ten F1 couples from each family (sibling mating) were randomly selected and kept under standard laboratory conditions with food and water and allowed to lay eggs to obtain F2, following the procedures described above for the parental generation.

Data analysis

The percentage of mating recorded in the pre-zygotic test was compared among the four types of crosses by means of a chi-square test of homogeneity. The latency to mate and the mating duration time were compared among treatments using a one-way analysis of variance (ANOVA) followed by a post hoc Tukey’s multiple comparisons test.

Post-zygotic tests involved the analysis of the following parameters both in F1 and F2 generations: % of egg hatch (number of hatched eggs/total number of eggs*100); % of pupation (number of recovered pupae/number of eclosed larvae*100); % of adult emergence (number of emerged adults/number of recovered pupae*100); female sex ratio (number of adult females/number of emerged adults). These variables were analyzed by means of a one-way ANOVA. Normality and homoscedasticity assumptions were met for all variables, except for the percentage of pupation in the F1. In this case, data were arcsine square transformed to meet homogeneity of variances assumptions. In all cases, ANOVA were followed by post hoc Tukey’s multiple comparisons tests. Deviations from a 0.5 sex ratio were evaluated by means of a G-test of goodness of fit, applying the Bonferroni correction for multiple comparisons.

Additionally, we analyzed: 1. Percentage of mated females that produced eggs (number of females that laid > 10 eggs/number of mated females*100); 2. Percentage of females that produced viable eggs (number of females for which > 5% of eclosed eggs were found / number of females that produced eggs*100); 3. Percentage of females with descendants (number of females which produced > 5 emerged F1 adults/number of females that produced viable eggs *100); 4. Percentage of mated females that produced viable eggs (number of females for which > 5% of eclosed eggs were found/number of mated females*100; i.e., considering all mated females); 5. Percentage of mated females with descendants (number of females which produced > 5 emerged F1 adults/number of mated females *100; i.e., considering all mated females). These variables were compared among types of crosses by means of a Chi-Square test of homogeneity; first among the four types of crosses, and later between Af-Cast-1 and Af-Cast-2 females.

Statistical analyses were performed using STATISTICA for Windows [114].

Cytological analysis

Mated females that did not produce descendants (females that did not lay eggs or that laid unviable eggs) were dissected under a stereoscope microscope (Olympus SZ30, Tokyo, Japan) to check both for any developmental abnormalities in the ovaries and the presence of sperm in spermathecae. The two ovaries and three spermathecae from each female were removed and placed on a slide. Preparations were stained with 2% acetic-orcein and observed under a phase contrast microscope Olympus BX40 (Olympus, Tokyo, Japan) using a 20X magnification objective. The general appearance, shape and structure of ovaries were analyzed as previously described [115, 116] and, the presence of sperm inside each one of the three spermathecae was visualized as previously described [112]. Sperm presence was determined whenever we visualized conspicuous bundles of sperm. For each female, the content of each spermatheca (presence/absence of sperm) was recorded.

Results

Molecular characterization of Wolbachia

Wolbachia was positively detected in all the A. fraterculus adults tested (N = 76; Table 1) using the 16S rRNA and wsp gene PCR-based assays. 16S rRNA sequence analysis showed identical base composition among samples (76 DNA samples, 380 bases). Basic Local Alignment Search Tool (BLAST) searches against the European Nucleotide Archive (ENA, EMBL, EBI) showed 100% identity with a large number of sequences including wMel (Wolbachia endosymbiont of Drosophila melanogaster; GenBank accession DQ412083.1).

In the case of wsp gene sequences (507 bases) a unique non-synonymous nucleotide change (C/T) was detected among the 76 samples analyzed (Fig. 1). The wsp nucleotide variants detected were named wAfraCast1_A and wAfraCast2_A respectively. BLAST nucleotide search of wsp gene sequence from wAfraCast1_A showed 100% identity with A. fraterculus isolate wAfBrazil_A (EU651897.1) and A. fraterculus isolate wAfPeru_A (EU651893.1) among others. The wsp nucleotide sequence of wAfraCast2_A showed 100% identity only with A. fraterculus isolate wAfArgentina_A (EU651896.1).

Fig. 1
figure 1

Identification of the single nucleotide substitution in 507 bp alignment of Wolbachia wsp sequences. The figure shows a section of the wsp nucleotide sequences alignment including Wolbachia sequences described here (AN KC589026.1 and KC589027.1 GenBank) corresponding to wAfraCast1_A or wAfraCast2_A respectively and, sequences from GenBank (NCBI) corresponding to the A. fraterculus isolate wAfArgentina_A (EU651896.1); A. fraterculus isolate wAfBrazil_A (EU651897.1); A. fraterculus isolate wAfPeru_A (EU651893.1) and Wolbachia strain wMel infecting D. melanogaster (DQ412100.1)

Full size image

The analysis of the HVRs of the wsp gene performed through the Wolbachia MLST webpage, evidenced different wsp allele and allelic profiles in HVR4 for the Wolbachia nucleotide variants identified here (Table 3). Further HVRs allelic profiles comparison revealed perfect match between wAfraCast1_A and several Wolbachia strains including Wolbachia strains infecting Rhagoletis cerasi (Diptera: Tephritidae) and Leucophenga maculosa (Diptera: Drosophilidae), whereas wAfraCast2_A showed no perfect match in this database.

Table 3 Characterization of the wsp HVRs. HVR allele definition is based on the amino acid motifs analysis of the wsp gene sequence (61–573 bp) in respect of wMel (Wolbachia databases - webpage pubmlst.org/Wolbachia/). Assigned alleles to wsp nucleotide sequences are also showed (wsp allele)
Full size table

MLST analysis showed identical nucleotide sequences in 22 DNA samples from the different A. fraterculus populations evaluated (Table 1). The MLST allelic profile obtained corresponds to gatB:1, coxA:1, hcpA:1, ftsZ:3 and fbpA:1 and sequence type (ST) 13. Phylogenetic analysis based on a concatenated dataset of 5 MLST loci (2079 bases) including the nucleotide sequences obtained here and a dataset of representative sequences from A, B and D Wolbachia supergroups from Baldo and Werren [103] revealed that Wolbachia found in Argentinean A. fraterculus populations belong to supergroup A (Fig. 2).

Fig. 2
figure 2

Neighbor-joining tree reconstructed based on concatenated MLST data (2079 bases). Phylogenetic tree reconstructed using a dataset including 30 MLST concatenated sequences published by Baldo and Werren [103] and a unique sequence corresponding to the concatenated MLST from wAfraCast1/2_A. Branch name is identified as Wolbachia sequence type (ST)– Wolbachia strain (if known)– host species name. Numbers in nodes indicate bootstrap support percentage (1000 replicates). Wolbachia supergroups are shown to the right of the tree. Similar topology was observed using Maximum Likelihood analysis (Additional file 4)

Full size image

In addition to MLST analysis, we evaluated the polymorphisms in seven additional loci from the Wolbachia genome (groEL, gltA, dnaA, sucB, aspC, atpD and pdhB) in at least three individuals of Af-Cast-1 and Af-Cast-2 strains. After the analysis of at least 370 b from each locus (see details in Table 2) no polymorphism was identified showing a high similarity between wAfraCast1_A and wAfraCast2_A at genomic level (see sequence alignments in Additional file 2). The sequence comparisons using BLAST also evidenced similarities among sequences from Wolbachia infecting Drosophila species (wMel, wRi, wHa) for the five genes evaluated, confirming the results obtained by MLST and phylogenetic analyses of wAfraCast1/2_A clustered with wMel group from supergroup A (Fig. 2).

Prevalence of Wolbachia

Wolbachia was detected in 100% of A. fraterculus individuals through PCR amplification and sequencing of wsp and 16S rRNA genes. A different prevalence of the two Wolbachia sequence variants identified in A. fraterculus populations was observed (Table 4). We found wAfraCast1_A in 16% and wAfraCast2_A in 84% of the A. fraterculus individuals from our laboratory colony (37 individuals; 24 females, 13 males). In addition, we identified wAfraCast2_A in 95% of the insects from natural populations (39 individuals; 22 females, 17 males) while only two individuals from Puerto Yeruá (Entre Rios) showed the presence of wAfraCast1_A (Table 4). Based on PCR and direct sequencing, no evidence of double infections was detected in the 76 A. fraterculus DNA samples analyzed.

Table 4 Prevalence of Wolbachia in A. fraterculus from Argentina
Full size table

Cytoplasmic Wolbachia in A. fraterculus

The presence of cytoplasmic Wolbachia and the lack of obvious Wolbachia integrations into the host genome (at least detectable with the molecular methods used in the present study) were confirmed in both A. fraterculus strains (Af-Cast-1 and Af-Cast-2) by means of antibiotic treatment followed by PCR assays. Wolbachia was not detected in any of the individuals treated with antibiotic (10 flies), whereas, control individuals (10 flies belonging to Af-Cast-1 and Af-Cast-2 strains reared without antibiotic treatment) resulted in a positive Wolbachia-specific amplicon in 100% of the cases.

Mating experiments

We followed the scheme of crossing experiments described in Fig. 3. Parental crosses and filial crosses (sibling matings) were performed to analyze the existence of pre- and post-zygotic sexual isolation barriers associated to Wolbachia. Specific PCR bands of approximately 430 bp corresponding to Wolbachia wsp gene were successfully amplified in all Wolbachia-infected A. fraterculus individuals used in the crossing experiments (parental flies, Fig. 3). Additionally, the absence of PCR amplicons was evidenced for all Wolbachia-cured parental pairs used as a control of our experiments.

Fig. 3
figure 3

Mating scheme of Af-Cast-1 and Af-Cast-2 individuals harboring different variant of Wolbachia (wAfraCast1_A and wAfraCast2_A, respectively). ∆ A. fraterculus harbouring wAfraCast1_A and □ A. fraterculus harbouring wAfraCast2_A. Individuals in the parental crosses were the subjects of the pre-zygotic tests. Their offspring were the subjects of the post-zygotic tests (F1)

Full size image

Pre-zygotic isolation test: We observed similar percentages of mating among the four possible types of cross (Chi-square test: χ2 = 6.637, P = 0.084, d.f. = 3) with a relatively high mean percentage of mated couples (72%) compared to previous results for this species. The latency and mating duration time did not differ among the types of crosses [ANOVA: Latency: F (3,165) = 1.831, P = 0.143; Mating duration time: F (3,165) = 2.597, P = 0.054] (Table 5). These results showed a lack of pre-zygotic isolation between the A. fraterculus sp1 strains described here.

Table 5 Mean values of percentage of mating, latency, and mating duration time of each type of cross, and female proportion obtained in the offspring (F1 and F2)
Full size table

Post-zygotic isolation analysis

We did not observe any statistically significant differences among the types of crosses regarding the percentage of eggs that hatched and adults that emerged in the F1 generation [%Egg hatch: F (3,82) = 0.52, P = 0.67; % Adults emergence: F (3,48) = 0.28, P = 0.84]. In contrast, the percentage of pupation showed statistically significant differences among crosses [ANOVA: F (3,46) = 4.78, P < 0.01]. Multiple comparison analysis showed that the Af-Cast-1 x Af-Cast-1 cross had a statistically significant lower percentage of pupation than the Af-Cast-2 x Af-Cast-2 cross. The other two types of crosses (Af-Cast-1 x Af-Cast-2 and Af-Cast-2 x Af-Cast-1) showed intermediate pupation values (Fig. 4 A-C).

Fig. 4
figure 4

Survival across development - Parameters measured for each type of cross (female x male). The crosses Af-Cast-1 x Af-Cast-1, Af-Cast-2 x Af-Cast-1, Af-Cast-1 x Af-Cast-2, Af-Cast-2 x Af-Cast-2 are mentioned in the figure as 1 × 1, 2 × 1, 1 × 2 and 2 × 2 respectively. a, b and c - F1 offspring analysis. d, e and f – F2 offspring analysis. (A/D) mean (± S.E.) % egg hatch; (B/E) mean (± S.E.) % pupation = percentage of larvae that reached pupae stage; (C/F) mean (± S.E.) % adult emergence = percentage of pupae that reached the adult stage. Points sharing a letter did not present any statistically significant differences

Full size image

In the F2 generation, we observed that the percentage of egg hatch and the percentage of pupation showed no statistically significant differences among crosses [F (3,30) = 2.15, p = 0.18; and F (3,29) = 1.49, p = 0.24, respectively] (Fig. 4 D and E). However, the percentage of adult emergence showed statistically significant differences among crosses (F (3,28) = 3.46; p = 0.029). Furthermore, Af-Cast-1 x Af-Cast-1 families showed the lowest percentages of adult emergence and Af-Cast-2 x Af-Cast-1 families the highest (Tukey test) (Fig. 4 F).

A sex ratio distortion that significantly favored females (both in F1 and F2 offspring) was detected in Af-Cast-1 x Af-Cast-2 crosses, whereas, in the case of Af-Cast-1 x Af-Cast-1 crosses, significant deviation of this parameter was observed only in F2 descendants. No bias was evidenced in crosses involving Af-Cast-2 females (Table 5).

Further analysis of data obtained from the parental crosses gave no statistically significant differences regarding percentage of mated females that produce eggs (χ2 = 2.321; p = 0.508, d.f. = 3), percentage of females that produce viable eggs (χ2 = 2.322, p = 0.508, d.f. = 3), percentage of females with descendants (χ2 = 0.396, p = 0.941, d.f. = 3), percentage of females that produce viable eggs (χ2 = 4.893, p = 0.180, d.f. = 3) and percentage of females with descendants (χ2 = 5.778, p = 0.123, d.f. = 3), (Fig. 5 A-E). Since data were homogeneous, results were pooled and compared between types of female. Again, the percentage of mated females that produced eggs did not differ between type of female (χ2 = 1.956, p = 0.162, d.f. = 1) (Fig. 5 F). Similarly, the percentage of females that produce viable eggs and the percentage of females with descendants were not statistically different between types of female (χ2 = 0.632, p = 0.427, d.f. = 1 and χ2 = 0.070, p = 0.791, d.f. = 1, respectively) (Fig. 5 G and H). In contrast, both the percentage of mated females that produced viable eggs and the percentage of mated females with descendants were significantly higher for the Af-Cast-2 females (χ2 = 4.706, p = 0.030, d.f. = 1; and χ2 = 5. 560, p = 0.018, d.f. = 1, respectively) (Fig. 5 I and J).

Fig. 5
figure 5

Mating experiments - additional analyses. a-e represent comparisons that included the four types of crosses. f-j, data coming from the same female were pooled irrespectively of the type of male they mated. Asterisks indicate statistically significant differences (p < 0.05) when percentages were compared by means of a Chi-square test of homogeneity

Full size image

Cytology of mated females

For each type of cross, we dissected the ovaries of at least 10 mated females that did not lay eggs and five mated females that laid unviable eggs. In all cases (77 females), we observed ovaries with a normal shape (fully developed and conserved size and structure), similar to the ones observed in reproductively mature females (control females, 15–20 days old) from A. fraterculus IGEAF strain (data not shown). In addition, the cytological analysis of spermathecae showed a high density of sperm (bundles) present in control females (Fig. 6 A) and absence of sperm in females that did not lay eggs and females that produced unviable eggs from the crossing experiments (77 females analyzed) (Fig. 6 B). It is worth mentioning that A. fraterculus is able to lay unfertilized eggs even in the absence of mating (virgin females). The results obtained here highlight the absence of sperm in the spermathecae as the main cause of the lack of descendants in the analyzed crosses.

Fig. 6
figure 6

Cytological analysis of A. fraterculus spermathecae (20X). a: spermatheca of A. fraterculus showing the presence of sperm bundles, which are indicated by an arrow b spermatheca of A. fraterculus showing no sperm in its content

Full size image

Detection of other reproductive symbionts

We evaluated the presence of Spiroplasma sp., Cardinium sp., Rickettsia sp., Arsenophonus sp. and Hamiltonella sp. by using specific PCR assays (Table 2). After the analysis of at least ten DNA samples from each of the A. fraterculus IGEAF strains, no symbiont-specific amplicons were obtained.

Discussion

The presence of Wolbachia in both laboratory and wild A. fraterculus populations from Argentina was evidenced and characterized in this study. Mating experiments showed a slight deficit of males in F1 and F2 progenies and a detrimental effect on larval survival, suggesting that some kind of male-killing phenotype may be associated with the presence of one of the two Wolbachia strains detected in A. fraterculus sp. 1.

The analysis of the wsp gene at a nucleotide level allowed the identification of two sequence variants of Wolbachia in the host populations (named as wAfraCast1_A and wAfraCast2_A). Sequence analysis of concatenated MLST dataset showed that these Wolbachia variants share the same MLST allelic profile. Furthermore, phylogenetic analysis clustered these variants in the same group (ST13) with wMel (Wolbachia infecting D. melanogaster), along with other Wolbachia strains belonging to supergroup A. Our findings using MLST in the identification of Wolbachia (and its clustering in supergroup A) were also supported by 16S rRNA sequence analysis.

Further characterization of Wolbachia using an antibiotic treatment, allowed the confirmation of an active cytoplasmic infection of this endosymbiont. We did not find evidence of insertion in the A. fraterculus genome, as antibiotic-treated flies showed lack of a specific amplicons for wsp and 16S rRNA Wolbachia genes. In addition, the prevalence analysis of the Wolbachia variants shows the absence of double infections under the experimental design and standard conditions used in the present study. Single infections of Wolbachia have also been described in other A. fraterculus populations [79, 117, 118].

The presence of Wolbachia in Argentinean populations of A. fraterculus was first reported by Cáceres et al. [79]. These authors analyzed two laboratory strains of A. fraterculus established at Insect Pest Control Laboratory (Seibersdorf, Austria), originally derived from wild flies collected from Argentina and Peru. Each laboratory population harbored closely related Wolbachia strain (wArg and wPer, respectively), with the presence of one nucleotide substitution in wArg based on wsp gene sequencing. In the present work, we found identical results at a nucleotide level with these previously reported Wolbachia wsp gene sequences (wAfraCast1_A identical to wPer and wAfraCast2_A identical to wArg). Moreover, we found that the wAfraCast1_A wsp sequence presented an identical nucleotide composition compared to a partial wsp sequence detected in a Brazilian population of Anastrepha sp.1, (GenBank AN EU 116325) reported by Coscrato and colleagues [117]. The presence of the same wsp gene sequence in different populations of A. fraterculus does not necessarily mean that they are infected with identical Wolbachia strains [16, 119]. The Wolbachia infection status of several morphotypes of the A. fraterculus cryptic species complex (including A. fraterculus sp.1) was recently published by Prezotto et al. [93]. The information provided by these authors with respect to Wolbachia sequence variants infecting different A. fraterculus populations from Argentina (either using MLST or wsp HVR analyses) differs from our findings. More knowledge regarding the origin of the samples and the number of individuals analyzed by Prezotto et al. [93] are needed in order to compare the results obtained in the two studies. Moreover, the same authors suggested a potential association between specific Wolbachia strains and distinct A. fraterculus morphotypes, which could act as a reinforcing factor in the diversification processes, providing also, some evidence of the possible way of transmission of Wolbachia. Further characterization of Wolbachia strains infecting members of the A. fraterculus complex, taking into account crossing experiments and deeper molecular analysis could provide insight to the speciation process in this complex, unraveling the genetic entities present and their phylo-geographic distribution.

Our crossing experiments showed a detrimental effect during the development for crosses involving Af-Cast-1 females. This is suggested by a statistically significant lower percentage of pupation in F1 offspring and a lower percentage of adults emergence in F2 descendants observed in the crosses involving Af-Cast-1 flies. Despite the lack of differences between females in the percentage of mated females that laid eggs, those that laid viable eggs, and those that successfully produced progeny, we were able to find a tendency to lower values in Af-Cast-1 females, which was statistically significant when these percentages were computed considering the total number of mated females, which allowed these small, non-significant effects, to accumulate. These results might point to a negative effect of a Wolbachia variant on the reproductive biology of its host. We also found that some parameters associated to immature development varied in some crosses between F1 and F2. For instance, Af-Cast-2 x Af-Cast-2 cross yielded higher egg hatch and pupation in the F1 than in the F2. Because these crosses involved flies with equivalent genetic background and Wolbachia infection status, this result suggests that unidentified experimental conditions probably varied between the F1 and F2.

Cytological analysis showed the absence of sperm in the spermathecae of females that did not lay eggs and females that produced unviable eggs, showing that the lack of sperm transfer is the main cause of unviable embryos production in some families. This result combined with the lack of differences in the % of hatched eggs allowed us to rule out the presence of a bidirectional cytoplasmic incompatibility associated to Wolbachia infection in the tested crosses. Also, it supports the hypothesis that detrimental effects in the survival associated to Wolbachia would occur later in the developmental stages rising new questions regarding possible effects of this bacterium on the host’s reproductive behavior that should be further addressed.

The analysis of sex ratio in each type of cross and generation showed a distortion in favor of females in crosses involving Af-Cast-1 females. Particularly, we observed this type of distortion in F1 and F2 of Af-Cast-1 x Af-Cast-2 pairs, and F2 offspring from Af-Cast-1 x Af-Cast-1 crosses. Additionally, individual analyses of each family showed that only a few paired-crosses appear to contribute with this sex ratio distortion (Additional file 3).

Our finding indicates that the effect of Wolbachia may not be homogeneous among different individuals belonging to the same host strain and requires further analysis. Studies including the quantification of Wolbachia titers in the parental couples and the measurement of biologically important parameters, in connection with genetic studies of the offspring, including cytological (cytogenetic) analysis will provide more evidence of the phenotype elicited by this endosymbiont in A. fraterculus. In this regard, previous studies described the importance of bacterial densities in the expression of a phenotype and the presence of different Wolbachia densities during host development [2, 36] using sensitive tools as the quantitative real time PCR (qPCR) and other methods for the detection of low titer reproductive symbionts [120,121,122,123,124]. Moreover, the action of non-bacterial, maternal-inherited microorganisms [125] must also be taken into account for future studies. Detection of low titer endosymbionts using more sensitive methods and the inclusion of crossing experiments involving antibiotic treatments will contribute to a better understanding of our findings.

Detrimental effects (lower % of pupation and % of adult emergence in the F1 and F2, respectively) and the sex ratio distortion observed in crosses involving Af-Cast-1 females, potentially elicited by the presence of Wolbachia and associated to a male killing phenotype, have been previously described in insect species by Hurst et al. [45], Dyer and Jaenike [46] and Kageyama and Traut [126]. A larger set of crossing experiments combined with the analysis of several biologically important parameters from the host populations (e.g. fecundity, % egg hatch, and/or differences in larval and/or pupal survival) are needed to better understand the effects Wolbachia may be inducing to this host species.

The results obtained here display the differences between the phenotype elicited by two Wolbachia sequence variants on their hosts, revealing some disparity in the cross-talk involving the bacteria and its hosts. This may include genetic variability in the bacterium as well as in the host species. In our study, we evidenced a significant similarity between the two Wolbachia strains analyzed, based on the identical MLST allelic profile and the identical sequences of 16S rRNA gene and seven additional Wolbachia genes (groEL, gltA, dnaA, sucB, aspC, atpD and pdhB). It is also worth noting that several studies have demonstrated the importance of the host genetic background associated to the molecular mechanisms involved in the phenotype induced by Wolbachia [39, 58, 118, 127]. Microsatellite analyses have shown a high genetic variability and differentiation among Argentinean populations of A. fraterculus [90, 128, 129]. Genetic evaluations using this kind of markers could be potentially useful to identify variation between the A. fraterculus strains harboring different Wolbachia variants under study in the present work. These studies may contribute to our understanding of the different reproductive effects displayed by Wolbachia in these singly-infected A. fraterculus strains.

Conclusion

This work contributes to the characterization of Wolbachia infection in A. fraterculus sp.1 from Argentina. We gained a first insight on possible mechanisms associated to the Wolbachia - A. fraterculus interaction by crossing singly-infected A. fraterculus strains. We found a potential deleterious effect on immature stages and a sex ratio distortion (male-killing) associated to one of the detected Wolbachia variants (wAfraCast1_A). Further mating experiments, coupled with quantification of Wolbachia titers and including cured lines, will shed light on the phenotype elicited by Wolbachia in A. fraterculus. Our findings are important for the characterization of A. fraterculus populations from Argentina, and as a contribution to develop environmentally-friendly and species-specific control strategies against this pest.