Endosymbiont and gut bacterial communities of the brown-banded cockroach, Supella longipalpa

Introduction

The brown-banded cockroach, Supella longipalpa, is an invasive nuisance and public health pest native to northern Africa that infests built environments across all inhabited continents. Its general biology has been reviewed in detail relatively recently (Nasirian, 2016) and therefore will only be briefly discussed here. Like its more prevalent relative, the German cockroach (Blattella germanica), S. longipalpa is an omnivorous species that completes its life cycle strictly indoors. However, S. longipalpa differs from B. germanica in numerous aspects of its physiology and behavior, including nutrient selection (Cohen et al., 1987), thermal preference (Tsai & Chi, 2007), and oviposition patterns (Benson & Huber, 1989), among others. Moreover, relative to B. germanica, fundamental biological studies of S. longipalpa are lacking.

In recent years, understanding the roles of commensal and symbiotic microbes in the biology of pest cockroaches has garnered significant interest across scientific disciplines (Latorre et al., 2022; Pan, Wang & Zhang, 2020), and the microbiomes of over 19 different omnivorous species have been sequenced (Tinker & Ottesen, 2020). The microbiome of the German cockroach (Kakumanu et al., 2018) has been the most extensively characterized and studied, revealing its involvement in olfaction and dietary preference (Pérez-Cobas et al., 2015; Zhu et al., 2022), aggregation behavior (Wada-Katsumata et al., 2015), susceptibility to insecticides (Pietri, Tiffany & Liang, 2018; Wolfe & Scharf, 2022), and colonization resistance against ingested human pathogenic bacteria (Ray, Potts & Pietri, 2020; Turner, Van Hulzen & Pietri, 2023). These findings have highlighted the potential applications of targeting the microbial communities of cockroaches for pest control and disease prevention, spurring additional work investigating the use of antimicrobials to effectively manipulate said communities (Li et al., 2020; Rosas et al., 2018; Wolfe & Scharf, 2021; Zha et al., 2023).

In contrast to B. germanica, little is known about the microbial communities associated with S. longipalpa. Several reports have described the isolation of diverse Gram-positive and Gram-negative bacteria from field-collected S. longipalpa (Nasirian, 2016). For example, Citrobacter spp., Enterobacter spp., Serratia spp., Klebsiella spp., and Pseudomonas spp. have all been successfully cultured from S. longipalpa (Le Guyader, Rivault & Chaperon, 1989). However, the abundances of these microbes and whether they represent contaminants or stable associations are unclear. Further, as culture independent high throughput sequencing methods have not been used to assess the microbial communities of S. longipalpa, the vast majority of taxa have likely been overlooked.

An additional intriguing microbial aspect of S. longipalpa is its relationship with the vertically transmitted insect endosymbiont Wolbachia. Most cockroaches are known to be colonized by Blattabacterium, a cockroach specific, vertically transmitted endosymbiont (Noda et al., 2020), but not by Wolbachia. On the other hand, S. longipalpa is one of few cockroach species that has been shown to be colonized by Wolbachia at a high prevalence, in addition to Blattabacterium (Choubdar et al., 2023; Gibson & Hunter, 2009; Oladipupo et al., 2023; Vaishampayan et al., 2007). Yet, the relative balance of the two endosymbionts throughout development, and their possible connections to bacteria in the gut, are unknown.

To add to the scarce body of knowledge of microbial interactions in S. longipalpa, in the current study we sought to quantitatively examine and characterize both its endosymbiont and gut bacterial communities. To do so, bacterial 16S rRNA gene amplicon sequencing was conducted on DNA extracts from whole adult females and males, early instar nymphs, and late instar nymphs. We then analyzed community composition and differences in relative taxonomic abundances and diversity indices across the sexes and developmental stages.

Materials and Methods

Results

Taxonomic signatures among nymph and adult S. longipalpa

Taxonomic assessment of the most abundant bacterial families among S. longipalpa identified Blattabactericeae (59.3%), Anaplasmataceae (24.9%), Desulfovibrionaceae (2.5%), Akkermansiaceae (2.2%), Bacteroidaceae (1.7%), Lactobacillaceae (1.6%) and Tannerellaceae (1.3%), among a few others (Fig. 1A). As Blattabacteriaceae and Anaplasmataceae are endosymbionts of S. longipalpa that do not reside in the gut, we performed analysis without these taxa, which identified Lactobacillaceae (21.6%), Desulfovibrionaceae (13.9%), Akkermensiaceae (13.5%), Bacteroidaceae (7.8%), Rikenellaceae (7.7%), Tannerellaceae (6.2%) and Ruminococceae (4.9%) as the most abundant bacterial families (Fig. 1B).

At the genus level, with the endosymbionts, Blattabacterium was highly abundant (59.1% ± 5.2 SEM), followed by Wolbachia (24.7% ± 3.9 SEM), Akkermansia (2.2% ± 0.67 SEM), Desulfolvibrio (2.1% ± 0.57 SEM), Bacteroides (1.7% ± 0.43 SEM), Lactobacillus (1.6% ± 0.46 SEM), Alistipes (1.6% ± 0.32 SEM), and Parabacteroides (1.2% ± 0.36 SEM) (Fig. 1C). Without the endosymbionts, Lactobacillus was the most abundant (21.7% ± 5.8 SEM), followed by Akkermansia (13.5% ± 3.3 SEM), Desulfolvibrio (11.4% ± 1.9 SEM), Bacteroides (7.8% ± 1.2 SEM), Alistipes (7.7% ± 0.85 SEM), Parabacteroides (5.8% ± 0.93 SEM), Erysipelatoclostridium (2.5% ± 1.1 SEM), and Dysgonomonas (2.4% ± 0.32 SEM), among a few others (Fig. 1D). Results of the most abundant bacterial genera in each developmental stage are summarized in heatmaps in Fig. 1E.

Minor differences in gut microbiome diversity and composition between nymph and adult stages of S. longipalpa

Alpha diversity analyses of the S. longipalpa gut microbiome revealed some significant differences between stages. After removing the endosymbionts, no significant differences were found in the number of observed genera between the early nymphs, late nymphs and adults (Fig. 2A; Kruskal-Wallis test; P-value = 0.48). However, gut microbiome alpha diversity (Shannon Index) was significantly higher in the late nymphs compared to the early nymphs (Fig. 2B; Kruskal-Wallis with Dunn’s Test; P-value = 0.03). No significant differences in alpha diversity were detected between the adult males and females (Figs. S1A and S1B). While principal coordinate analysis (PCoA plot) based on weighted Bray-Curtis distances showed that the early nymphs may have more similar gut microbiome composition compared to the late nymphs and adults, a PERMANOVA test indicated no significant differences (Fig. 2C; R2 = 0.12; F-statistic = 1.48; P-value = 0.172). Similarly, PCoA based on unweighted Bray-Curtis distances showed no significant differences (Fig. S2A; PERMANOVA: R2 = 0.06; F-statistic = 0.69; P-value = 0.77) either with or without the symbionts and broken down by males and females (Figs. S2B and S2C). To further probe differences in gut microbiome beta diversity between the stages of development, we investigated the interindividual variation. Analysis of the distance from centroid in ordination space, which indicates how dispersed microbiomes are relative to the average distance in their group (Sharma et al., 2022), found that the adults have higher inter-individual variation in bacterial community composition compared to the early nymphs (Fig. 2D; Wilcoxon Rank Sum test; P-value = 0.02). Overall, alpha and beta diversity analyses showed some notable differences between the early nymphs and later developmental stages (i.e., late nymphs and adults), which were more similar to each other.

Akkermansia is significantly more abundant in early nymphs

Since minor differences were found in gut microbiome diversity and variation, we further analyzed what bacterial taxa may be driving these differences. Differential abundance analysis found that Akkermansia was the only genus to be significantly different among the developmental stages in S. longipalpa, showing it was much higher in abundance in the early nymph stages relative to the late nymph stages (Fig. 3A; ANOVA; F-value = 20.55; P-value = 1.14e-05). Analysis of the other most abundant genera including Lactobacillus, Desulfolvibrio and Bacteroides did not show any significant differences between the developmental stages (Figs. S3A–S3C). Considering that heatmaps displayed possible differences in the abundance of the endosymbionts in each developmental stage (particularly Wolbachia), we also conducted a differential analysis of these taxa. However, no statistical significance was detected (Fig. 3B; ANOVA; F-value = 0.038; P-value = 0.97 and Fig. 3C; ANOVA; F-value = 1.487; P-value = 0.25). Further analysis of differential abundances between adult males and females found only one statistically significant difference; the females displayed significantly higher abundances of Parabacteroides than males (Fig. S4B; Student’s t-test; P-value = 0.04). Analysis of the differential abundances of the other most abundant genera found in adult males and females can be seen in Fig. S4.

Blattabacterium abundance is negatively correlated with microbiome alpha diversity and abundance of the secondary endosymbiont Wolbachia

To consider whether the highly abundant endosymbionts may affect each other or other constituents of the gut microbiota, we performed several correlations. Pearson’s correlation analysis performed with the ten most abundant bacterial taxa among all the developmental stages found Blattabacterium to be negatively correlated with them all. Blattabacterium and Wolbachia showed a particularly strong and statistically significant negative correlation (Fig. 4B; Pearson correlation coefficient r = −0.82; Multiple R2 = 0.68; P-value = 8.131e−07). Meanwhile, Wolbachia was positively correlated with all the other bacterial taxa except for Akkermansia and Parabacteroides (Fig. 4A). Furthermore, we also observed Blattabacterium to be strongly negatively correlated with gut microbiome alpha diversity (Fig. 4C; Pearson correlation coefficient r = −0.78; Multiple R2 = 0.62; P-value = 5.749e−06), while Wolbachia showed no correlation with gut microbiome alpha diversity (Fig. 4D; Pearson correlation coefficient r = 0.31; Multiple R2 = 0.09; P-value = 0.14).

Discussion

Here, we present the first molecular study comprehensively characterizing the microbiota of the brown-banded cockroach (S. longipalpa), both with and without its endosymbionts, across developmental stages and sexes. Our results indicate that the microbial community in early nymphs is distinct from later stages, which suggests developmentally driven changes in the gut microbiome. In addition, out data suggest that primary Blattabacterium endosymbionts play an important role in shaping the total microbial community, as an increase in Blattabacterium relative abundance was specifically associated with lower gut microbiota diversity and lower relative abundance of the secondary endosymbiont Wolbachia.

It is well established that pest cockroaches, particularly B. germanica and the American cockroach, Periplaneta americana, harbor a wide variety of microorganisms in their gut that assist a number of important physiological processes. However, the microbial composition of S. longipalpa had not been previously elucidated. While we found that the Blattabacterium and Wolbachia endosymbionts make up over 80% of the microbial community composition in S. longipalpa, we also found some similarities to other cockroach species in the composition of the gut microbiome. For example, some of the most abundant bacterial taxa at the family level detected in B. germanica were also present in S. longipalpa, including Desulfovibrionaceae, Bacteroidaceae, Ruminococcaceae, and Rikenellaceae (Kakumanu et al., 2018; Pérez-Cobas et al., 2015). However, S. longipalpa also harbored Lactobacillaceae and Akkermansiaceae in higher abundances. At the genus level, microbial taxa discovered in S. longipalpa were also found in three species of Periplaneta cockroaches including Lactobacillus, Desulfovibrio, Dysgonomonas, Bacteroides, Tennerella, Parabacteroides and Alistipes (Lee et al., 2020).

Overall, the microbial communities in the early nymphs were observed to be different from the late nymphs and adults, which were more similar to each other. Microbial alpha diversity was found to be significantly lower in the early nymphs compared to the late nymphs, while lower inter-individual variation was also observed in early nymphs relative to adults. Fully deciphering diversity patterns in S. longipalpa may require future analysis of individual gut sections, but our observations are similar to developmental changes discovered in B. germanica (Carrasco et al., 2014). In ticks (Amblyomma americanum) (Menchaca et al., 2013) and several species of crustaceans (Mente et al., 2016; Miao et al., 2020; Zhang et al., 2021), changes in the microbiome during development have also been reported. Although our data do not address the mechanisms behind the patterns we describe, we speculate several possibilities. Developmental differences in the gut microbiome may be due to differences in feeding behavior that introduce bacteria into the gut (Pérez-Cobas et al., 2015; Zhu et al., 2022), including coprophagy, which is more prevalent in early instar nymphs (Cruden & Markovetz, 1987; Kakumanu et al., 2018; Pietri & Kakumanu, 2021). Additionally, biochemical changes in the gut microenvironment that occur as the gut matures may exert selection on the microbial community, inhibiting the survival of some taxa while promoting the growth of others, resulting in ecological succession.

Interestingly, a major difference observed in the gut microbiome of S. longipalpa across development was the greater relative abundance of Akkermansia in early nymphs. Some species in this genus (Akkermansia muciniphila) have been similarly found to colonize the gut early in life in humans (Derrien et al., 2008; Rodrigues et al., 2022), but Akkermansia has not been widely observed in the gut microbiome of cockroaches. To our knowledge, only one metagenomic study of the gut microbiome of B. germanica detected Akkermansia in both the early nymph and adult stages (Domínguez-Santos et al., 2021). Akkermansia spp. are Gram-negative anaerobic bacteria that can colonize the gut of a broad range of mammals (Belzer & de Vos, 2012). Although our study could not identify the species of Akkermansia present, different species, particularly A. muciniphila, have been identified as potential probiotic bacteria, similar to Lactobacillus spp. (Rodrigues et al., 2022), suggesting that both Akkermansia and Lactobacillus could be necessary for optimal fitness of S. longipalpa in the indoor environment.

An additional, particularly notable finding was the apparent relationship between the primary endosymbiont Blattabacterium and the secondary endosymbiont Wolbachia. Considering all life stages and sexes, there was a strong, statistically significant negative correlation between the relative abundances of the two endosymbionts, as higher abundance of Blattabacterium was linked to lower abundance of Wolbachia. This link was much stronger than the link between Blattabacterium and the levels of other abundant taxa such as Akkermansia and Lactobacillus. The biological significance of this correlation remains unclear, but it suggests competition between the two endosymbionts for intracellular space and/or resources, as both are known to occupy fat body and germline tissues to enable vertical transmission (Pietri, DeBruhl & Sullivan, 2016). As such, S. longipalpa may be a promising model for the study of dual endosymbiont interactions at both the organismal and cellular levels. A similar negative relationship was also observed between Blattbacterium and the gut microbiome, as higher relative abundance of this endosymbiont was specifically significantly negatively correlated with gut microbiome alpha diversity (Shannon Index), while the abundance of Wolbachia was not. In studies of experimentally generated aposymbiotic B. germanica, other investigators found that Blattabacterium and gut microbiota had distinct physiological roles and did not compensate for each other nor interact significantly (Cazzaniga et al., 2023; Muñoz-Benavent et al., 2021). Our data indicate that S. longipalpa differs in this regard. However, fully establishing these links will require targeted follow-up studies of the absolute abundances of both endosymbionts across development.

From a practical standpoint, the new insight into the microbiome of S. longipalpa presented here has several important implications. There is ample evidence that the gut microbiota plays important positive roles in the physiology of other cockroaches, including insecticide resistance (Chao et al., 2020; Jiang et al., 2021; Pietri, Tiffany & Liang, 2018; Wolfe & Scharf, 2022; Zhang et al., 2018, 2022; Zhang & Yang, 2019). While it remains to be determined if this is the case in S. longipalpa, understanding the composition of its gut microbiome will inform the design of studies to test the links between the microbiome and physiology and suggest potential bacterial taxa to target, either chemically or environmentally, to mitigate any microbe-mediated resistance or other microbe-mediated benefits. In addition, it is well known that cockroaches have the ability to acquire, maintain, and transmit human enteric bacterial pathogens (Donkor, 2020; Turner, Peta & Pietri, 2022). Although S. longipalpa has been found to harbor such pathogens in the field (Nasirian, 2016), little is known about the mechanistic aspects of pathogen transmission by this species. Recent work has revealed that the transmission of enteric bacterial pathogens such as Escherichia coli and S. Typhimurium by German cockroaches is a complex, biologically active process that is influenced in part by the gut microbiota (Ray, Potts & Pietri, 2020; Turner, Van Hulzen & Pietri, 2023; Turner, Peta & Pietri, 2022). Therefore, the microbiota may also be a potential target for mitigating pathogen transmission by S. longipalpa, and its role in the ability to become infected by and transmit pathogens is intriguing. Both of the above avenues should be investigated further, and the information provided by this study is a pivotal first step towards extending our knowledge of the microbial communities of S. longipalpa beyond characterization and into the functional and practical realms.

Supplemental Information