Microbiota of the Whitefly Bemisia tabaci (Hemiptera: Aleyrodidae) by 16S rDNA Illumina Sequencing

Abstract

Bemisia tabaci (Aleyrodidae family) is one of the most damaging pests of numerous crops worldwide. Insecticides, namely pyrethroids and organophosphates, have long been the primary control tools against this pest, resulting in several resistance cases. In Tunisia, the two most damaging biotypes of B. tabaci, MEAM1-B and MED-Q, are sympatric, and more concerns about developing resistance keep rising due to the extensive use of insecticides. Here, we aimed to elucidate the molecular mechanism of resistance to pyrethroids and organophosphorus insecticides in two Tunisian populations of B. tabaci, collected respectively on Capsicum annuum and Lantana camara, and then determine the bacterial community associated with insecticide resistance and susceptible biotypes based on 16S rRNA Illumina sequencing. The results showed that the population collected on Capsicum annuum belonged to the MEAM1-B biotype with an insecticide resistance profile. In contrast, the population collected on the Lantana camara belonged to the MED-Q biotype with a sensitive profile. The bacterial communities of the two biotypes were predominantly structured by the Proteobacteria phylum and three genera, including Candidatus Portiera, the secondary facultative symbiont, and Hamiltonella, which were unevenly distributed between the two biotopes. Our results provide the first evidence for insecticide resistance alleles in Tunisian MEAM1-B populations and suggest an association between bacterial community composition within susceptible biotypes and insecticide resistance.

1. Introduction

The whitefly B. tabaci (Gennadius) (Hemiptera: Aleyrodidae) [1] is a major pest of many crops worldwide and a vector of over 300 plant viruses [2]. Its genetic diversity has generated an impressive list of host plants and a wide range of biological data [3]. B. tabaci has multiple genetic groups, also called biotypes, which are morphologically identical but have different biological and genetic characteristics [4]. Early work that assigned B. tabaci to various biotypes (denoted using the Roman alphabet) was based on biological characteristics such as fecundity, behavior, host range, and virus transmission ability [5]. Subsequently, the focus shifted to genetic variation through various molecular markers such as protein markers, RAPD-PCR (random amplification of polymorphic DNA), and sequencing of ribosomal and mitochondrial genes [6]. As of 2011, all the molecular data available from the sequencing of the mitochondrial cytochrome oxidase 1 (mtCOI) gene made it possible to group these biotypes into 11 genetic groups, and the naming was based on their geographical origin [7]. New groups were added from this date, and 45 groups have been described [8,9]. Currently, the global status of this pest refers to two highly invasive groups. This is the case for the former biotype B (now known as “Middle East-Asia Minor 1” or MEAM1) and the former biotype Q (now known as “Mediteranean” or MED), which are the most invasive and damaging in the world [7,10]. MEAM1 is an aggressive colonizer and efficient vector of viruses, while MED exhibits strong resistance to insecticides [7,10]. In Tunisia, Ben Abdelkrim and his collaborators [11] reported the presence of MEAM1-B and MED-Q biotypes with MEAM1-B predominating.

Chemical insecticides are a major control strategy in managing the whitefly B. tabaci. Two classes of insecticides, including organophosphates and pyrethroids, have been successfully used for many years to control this insect pest [12,13,14]. However, over time, the extensive use of insecticides has promoted the eventual selection of resistant individuals in treated populations. This development is an obvious constraint to effective pest management. Farmers often increase the frequency of treatments and the quantity of products used to cope with this constraint. However, this strategy has resulted in the opposite of the desired effect and has increased resistant individuals compared to susceptible ones [15,16,17,18]. Both organophosphates and pyrethroids act on the insect nervous system by altering the functioning of the voltage-gated sodium channel. One well-studied example of resistance that involves this type of alteration is resistance to pyrethroids. As described previously, this class of insecticides acts primarily by binding to Csvd and causes a knockdown (kd) effect. Kd resistance is the earliest known form of resistance to pyrethroids, characterized by target site insensitivity attributed to structural changes. This type of resistance is linked to a recessive gene named kdr, which encodes for a Csvd protein. The first kdr mutation was detected by Busvine in the housefly (Musca domestica) in 1951; it is translated at the protein level by a substitution of leucine (L) by phenylalanine (F) (L1014F). This same substitution was then identified in other insects, such as the green peach aphid Myzus persicae and the cabbage moth Plutella xylostella. This mutation is usually found in association with a second “super-kdr” mutation, which results in a substitution of a methylation substitution of methionine (M) to a threonine (T) (M918T), conferring higher levels of resistance [12,14,19].

Another example is the resistance to organophosphates. The insensitivity of AChE is a widespread mechanism in insects; mutations affect the enzyme’s structure by limiting insecticide access to the catalytic site. Two genes encoding AChE are present in most insects, ace1 and ace2, and a combination of mutations is generally found [19,20].

Recently, several studies have directly linked insect symbionts to the detoxification of insecticides and/or conferring insecticide tolerance on their hosts. The detoxification of organophosphates by a bacterial symbiont was first reported in the apple maggot Rhagoletis pomonella by Boush and Matsumura [21], who reported the degradation of six different insecticidal active ingredients by an insect-associated bacterium. Further, Kikuchi and collaborators [22] showed that the Burkholderia bacteria colonizing the guts of the bean bug confer resistance to the organophosphorus pesticide fenitrothion on its insect hosts. Other findings have also indicated that bacterial symbionts, particularly Sphingomonas, play a vital role in enhancing the resistance of cotton aphids (Aphis gossypii) to insecticides, especially imidacloprid [23]. Furthermore, Hamiltonella defensa, a secondary bacterial symbiont, significantly decreases insecticide susceptibility in the wheat aphid (Sitobion miscanthi), forming a complex relationship with its aphid hosts that may influence their immune responses and ability to withstand chemical treatments [24]. This resistance mechanism arises from intricate interactions between the bacterial symbiont and its aphid host, potentially involving modifications of metabolic pathways within the host. This allows aphids to better tolerate or detoxify the chemicals present in insecticides.

On the other hand, some research has indicated that, contrary to earlier expectations, certain bacterial symbionts can enhance the sensitivity of aphids to insecticides. For instance, the Arsenophonus symbiont, the S-type, has decreased insecticide resistance in the brown planthopper (Nilaparvata lugens), a significant agricultural pest [25]. Additionally, Rickettsia symbionts have been investigated for their role in increasing susceptibility to thiamethoxam in B. tabaci [26]. These findings underscore the diverse and intricate nature of the interactions between insects and their symbionts, which can vary significantly across different species and environmental contexts.

The B. tabaci species has been reported to harbor primary symbionts, known as obligate primary endosymbionts (P-symbionts), which maintain a direct mutualistic relationship with their hosts. One such symbiont is Portiera, which is present in all individuals and is located in specialized cells called bacteriocytes. Like other primary symbionts of phloem-feeding insects, Portiera provides essential amino acids and carotenoids to its whitefly hosts [22]. Moreover, B. tabaci harbors several secondary symbionts (S-symbionts) such as Hamiltonella, Arsenophonus, Cardinium, Wolbachia, and Fritschea present in the hemolymphs, midgut [27]. These microbes provide diverse advantages in fitness, such as enhancing fertility, longevity, and the sex ratio of females [28]. Secondary symbionts have been demonstrated to improve the resistance of insects against fungal pathogens significantly [29] and in the detoxification of toxins and insecticides [22].

In Tunisia, the absence of data that indicate a less intensive use of insecticides to manage this pest and the increase of resistance cases reported in many Mediterranean countries emphasize the importance of examining the resistance situation of our populations. Therefore, in the present study, we first aimed to elucidate the molecular mechanisms of insecticide resistance in two Tunisian populations of B. tabaci against organophosphates (OPs) and pyrethroids (Pyrs) insecticide classes. Then, we characterized the gut microbiota composition within these B. tabaci populations using high-throughput 16S rRNA gene sequencing to investigate a further correlation between biotypes and resistance profiles against insecticides.

2. Materials and Methods

2.1. Sampling of Bemisia tabaci and Total DNA Extraction

Two populations of B. tabaci were analyzed in this study. The first population of Bt_pepper (n = 30) was collected in May 2017 from the leaves of greenhouse peppers, Capsicum annuum, in Takelsa, northwest of Cap Bon. The second population, Bt_lantana (n = 30), was collected in April 2017 from the lantana tree Lantana camarain in Chott-Meriem in the Central-East of Tunisia. These samples were preserved in 96% alcohol and placed in a freezer at −20 °C. According to the manufacturer’s instructions, total DNA was extracted using the FastDNA Spin Kit for soil (MP Biomedicals, Eschwege, Germany). All DNA samples were analyzed for quantity and quality using a Qubit 3.0 fluorometer (Invitrogen, Carlsbad, CA, USA).

2.2. Determining the Biotypes of Bemisia tabaci

Biotypes of B. tabaci populations were identified based on the COI (Cytochrome Oxidase 1) marker gene sequencing using 4 primers, including the universal primers for B. tabaci and 3 specific primers (Btab-B (745 bp), Btab-Q (303 bp), and Btab-Uni/Btab-NW (405 bp)), targeting the biotypes B, Q, and New World de B. tabaci, respectively. PCR reactions were performed in 25 μL reaction volumes containing 50 ng of DNA extracted, 10 μM of each primer, 5 U/μL of Taq polymerase (GoTaq^® G2 DNA Polymerase), Taq Polymerase buffer (5X), 100 mM dNTPs, and 25 mM MgCl₂. The different steps include an initial denaturation step for one cycle at 95 °C for 2 min, followed by 35 cycles consisting of a denaturation step at 95 °C for 30 s, a primer hybridization step at 46 °C (Btab-Uni) and 64 °C (Btab-B, -Q, -NW) for 1 min, and an elongation step at 72 °C for 1 min. The reaction ends with an elongation step for a 5 min cycle at 72 °C. Purified PCR products were bi-directionally sequenced on an ABI 3500 Genetic Analyser (Applied Biosystems, Waltham, MA, USA) using BigDye Direct Cycle Sequencing Kit v.3.1 (Applied Biosystems, Waltham, MA, USA) and amplification primers. The raw sequencing data were analyzed and edited manually using BioEdit software, obtaining a consensus sequence for each PCR fragment [30]. Consensus sequences were aligned using the MUSCLE program (https://www.ebi.ac.uk/jdispatcher/msa/muscle?stype=dna, accessed on 15 January 2025) and visualized using GeneDoc Editor (V 2.7.000) (Nicholas & Nicholas 1997). The sequences were then BLASTed against a program (http://blast.ncbi.nlm.nih.gov/Blast.cgi, accessed on 15 January 2025) and the Barcode of Life Data System (BOLD) to confirm sequence identity and similarity.

2.3. Diagnostic Assays for the L925I and F331W Resistance Mutations and Diversity of Resistance Alleles

In the two sampling sites, Takelsa and Chott-Meriem, two classes of insecticides, organophosphates (OPs) and pyrethroids (Pyrs), were, respectively, used to control B. tabaci. PCR-based assays were used to detect resistance mutations to OPs (F331W mutation, ace1 gene) and Pyrs (L925 mutation, kdr gene) as described in Tsagkarakou et al. [26]. Two pairs of primers—Bt-kdr-RIntr1/Bt-kdr-F1 (184 pb) and Bt-ace-R/Bt-ace-F (287 pb)—were used targeting the kdr and ace1 genes, respectively. PCR reactions were performed in 25 μL reaction volumes containing 50 ng of DNA extracted, 10 μM of each primer, 5 U/μL of Taq polymerase (GoTaq^® G2 DNA Polymerase), Taq polymerase buffer (5X), 100 mM dNTPs, and 25 mM MgCl2. The different steps include a first denaturation step for a cycle at 94 °C for 5 min, followed by 35 cycles including a 5 min followed by 35 cycles including a denaturation step at 95 °C for 15 s, a hybridization step at 52 °C, and an elongation step at 72 °C for 40 s. The reaction ends with an elongation step for one cycle for 7 min at 72 °C. PCRs were performed in the Biometra TRIO thermocycler. The obtained sequences of ace1 and Kdr genes were checked using the NEBcutter V2.0 program (https://nc3.neb.com/NEBcutter/, accessed on 15 January 2025) to identify discriminatory restriction sites, L925I, and F331W mutations. Finding that the DdeI and BsrI restriction enzymes are specific for L925I and F331W, respectively. The Kdr gene L925 mutation, associated with pyrethroid resistance, deletes a restriction site for the DdeI enzyme. Without mutation, DdeI generates two fragments of 124 et 60 bp. The F331W mutation in the ace1 gene, related to organophosphate resistance, creates three restriction sites for the BsrI enzyme, leading to 4 fragments of size 140, 79, 61, and 7 bp. However, two restriction sites are created for the ace-s sensitive allele (F331), generating three fragments of size 201, 79, and 7 bp. The resistant ace-r allele (W331) carries three sites, and digestion generates four fragments. The digestions were performed in a reaction volume of 20 μL of each PCR product, enzyme buffers (10X), and the appropriate enzyme (10 U/μL). The mixtures were incubated at 37 °C overnight. The digested products were then analyzed using low-melting 3% agarose gel electrophoresis.

2.4. Bioinformatic Analysis of Microbiota Based on 16 rDNA Gene Sequencing

Gut microbiota of two populations of B. tabaci were characterized by targeted gene amplification of the 16S rDNA V3–V4 region using the forward primer 5′-TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG-3′ and reverse primer 5′-GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAACC-3′ Support Center (illumina.com). Sequencing libraries were prepared using the Nextera XT kit according to the Illumina Supporting Guide (Support Center (illumina.com)). Libraries were then multiplexed and sequenced on the MiSeq paired-end Illumina platform adapted for 300 bp paired-end read (Illumina, San Diego, CA, USA). The reads (Illumina) were analyzed for quality control using FastQC software Version 0.12.0 (Babraham Bioinformatics, Cambridge, UK). Trimming was processed using an expected error threshold of 1 combined with filtering low-quality reads (only reads with a length above 150 bp were retained) using the Trimmomatic 0.32 [31] program and Chimaera detection and removal. The filtered paired-end reads were then merged using the command fastq-join [32] and clustered at a 0.03 cutoff % level of similarity into OTUs. Ribosomal RNA gene reads were classified against the non-redundant version of the SILVA SSU reference taxonomy (release 123; https://www.arb-silva.de/documentation/release-123/, accessed on 15 January 2025). For bacterial diversity estimation in the samples, the number of operational taxonomic units (OTUs) at 97% sequence identity was determined, and rarefaction analyses were carried out.

3. Results and Discussions

3.1. Identification of Bemisia tabaci Biotypes

Results showed that up to four primers were used for the amplification of the mtCOI gene region, and only two primers successfully amplified the MEAM1-B biotype-specific gene (478 bp) in the Bt_pepper population sampled from greenhouse peppers in Takelsa. Meanwhile, for the Bt_lantana population collected from lantana trees in Chott-Meriem, only the universal primer Btab-Uni revealed a band of size 745 bp. These results agree with previous works examining the structure of B. tabaci populations in Tunisia, which show the prevalence of MED-Q and MEAM1-B biotypes of B. tabaci [11]. B. tabaci is a species complex of more than 40 species, among which are the closely related invasive B and Q biotypes [7,33]. These biotypes are morphologically identical but differ in traits like insecticide resistance profiles, where the Q biotype is known to develop higher resistance to insecticides than the B biotype. At the same time, B is characterized by high fecundity and a broad host range [34].

3.2. Detection of L925i (kdr) and F331W (ace) Mutations

Gene amplification and sequencing of kdr and ace1 gene sequences of B. tabacii (Bt) populations showed that the Bt_lantana-Q biotype has an identity of 100% and 99.31% within ace1 (accession no. XM_019040190) and kdr (accession no. AY094601) gene sequences of B. tabaci species, respectively. Indeed, for the Bt_pepper-B biotype, the population showed an identity of 99.6% and 99.31% sequences of ace1 (LC199311) and kdr (DQ205212) genes, respectively. The kdr and ace1 gene sequences of B. tabaci populations were deposited in the Genbank database under the accession numbers as follows: Bt_pepper-B-ace (ON980571); Bt_lantana-Q-ace (ON980572); Bt_pepper-B-kdr (ON980573); and Bt_lantana-Q-kdr (ON980574). Comparisons of kdr and ace1 gene sequences allowed the detection of polymorphism between the two populations in the regions flanking the L925I and F331W mutations. Concerning the L925I mutation, the kdr gene sequences from both populations were aligned, and the results revealed the substitutions of two nucleotides at positions 20 and 123 (Figure 1A). Only the TTA mutation to ATA at position 20 is known to generate the resistant allele I925. This mutation is found only at the sequence level of the Bt_pepper-B biotype population. This mutation consists of a change from leucine to isoleucine at position 925 [35] (Figure 1B). This result was specifically identified in a B strain of the MEAM1-B group from Arizona and was associated with pyrethroid resistance. An additional kdr mutation, known as T929V, has also been identified to confer resistance to pyrethroids in B. tabaci. This mutation is specific to the MED-Q biotype [16]. Although the Bt_lantana-Q biotype population analyzed does not have this mutation, it was detected in El-Hamma (Tunisia) by Gauthier et al. [36] with high prevalence. Regarding the F331W mutation, sequence analysis of the ace1 gene sequences of the two populations shows a double nucleotide substitution of TTC to TTG at positions 99 and 100 (Figure 2A). This mutation, only found within the Bt_pepper-B biotype population, consists of a change from phenylalanine to tryptophan at position 331 of the acetylcholinesterase enzyme protein sequence (Figure 2B) to make it insensitive to organophosphate binding [15]. This mutation has been identified in the MEAM1-B populations of B. tabaci [15] and the Culex tritaeniorchynchus mosquito [37]. The L925I and F331W mutations are not unique to the MEAM-B biotype. Indeed, these mutations have been detected in several countries in MED-Q populations, including Greece [31,36,37], Spain, France [15], and China [38].

On the other hand, screening of the L925I (kdr) mutation using a specific restriction enzyme (DdeI) showed that the Bt_pepper-B biotype revealed a single fragment of 184 bp. In comparison, the population collected on the lantana Bt_lantana-Q biotype revealed two fragments of 124 and 60 bp (Figure S1). As a result, Bt_pepper-B carries the resistant kdr-r allele (I925) (BT-B-R), and the Bt_lantana-Q biotype carries the sensitive allele kdr-s (L925). Concerning F331W (ace) mutation analysis, the BsrI-digested PCR products showed that the Bt_pepper-B biotype carries the ace-r resistant allele (W331), revealing three fragments of size 140, 79, and 61 bp. In contrast, the Bt_lantana-Q biotype carries the susceptible ace-s allele (F331), revealing two fragments of sizes 201 and 79 bp (Figure S2). Our results showed that only the population collected on pepper (Bt_pepper-B) has the resistance alleles kdr-r and ace-r. In contrast, the population collected on lantana (Bt_lantana-Q) is susceptible to both genes.

Our results collectively agree with the work of Gauthier et al. [36], which inspected the resistance status of MED-Q populations in many countries of the Mediterranean basin, including Tunisia. These researchers revealed that the Tunisian MED-Q population was the only one among the MED-Q populations in the other inspected countries that showed very low frequencies or even the absence of kdr and ace resistance alleles. Regarding the Tunisian MEAM1-B biotype, no study has been conducted on its resistance to insecticides. Thus, the present study provides the first evidence that B. tabaci MEAM1-B harbors resistance alleles to organophosphates and pyrethroids. On the other hand, our results contradict observations generally reported in the literature. Typically, the MED-Q biotype is known for its high insecticide resistance [10].

In contrast, our study showed that the MED-Q population analyzed exhibits no resistance alleles. Such absence could be explained by the lower selection pressure on MED-Q populations in Tunisia than on MEAM1-B populations. Indeed, in Tunisia, this biotype is mainly found in vegetable crops, while MED-Q is found on ornamental plants, including Lantana camara [39]. This plant is considered to have no significant economic value, consequently receiving fewer intentional insecticide treatments. In contrast, resistance alleles are more likely to be selected in economically important vegetable crops. This argument can be supported by the case of the MED-Q population collected from eggplant plants in El Hamma, one of the main producing regions of vegetable crops. According to Gauthier et al. [36], this is the only Tunisian MED-Q population with high resistance allele frequencies. This finding reflects the role of insecticide treatments in selecting resistance mutations.

3.3. Gut Microbiota Associated with Bemisia tabacii Biotypes

Eighty-four thousand eight hundred twenty-one reads were obtained from samples of MED-Q and MEAM1-B with an average length of 419 bp. 4712 OTUs were obtained, of which 70% and 97.05% could be taxonomically clustered for MED-Q and MEAM1-B (

Table 1. Summary of the characteristics of Bt_lantana-Q and Bt_pepper-B biotypes sequences analyzed and alpha diversity indices based on the number and pattern of OTUs (0.03 cutoff) observed in each sample.

Biotypes	Raw Reads	Valid Reads	Average Reads Length	OTUs	Classified OTUs	Chao-1	ACE	Simpson	Shannon	Good’s Coverage of Library (%)
Bt_lantana-Q	35.057	34.999	419	2154	70%	857.91	1028.00	0.40	1.25	99.05
Bt_pepper-B	49.754	49.439	419	2558	97%	844.16	1455.33	0.40	1.16	99.38

). Valid reads (99% for both populations) were rarefied, and results showed that the rarefaction curve for each sample tended to saturation, indicating that the OTUs identified were representative of the 16S amplicons in each sample (Figure S3). Alpha diversity assessment (ACE, Simpson, chao richness) (Table 1) of OTUs of the two biotypes showed the same profile (Table 1).

After filtering raw sequences from the V3_V4 regions of 16S rRNA, 84,821 reads were obtained from MED-Q and MEAM1-B biotypes samples with an average length of 419 bp. 4,712 OTUs were obtained, of which 70% and 97.05% could be taxonomically clustered for MED-Q and MEAM1-B, respectively (Table 1). Valid reads (99% for both populations) were rarefied, and the rarefaction curve for each sample tended to saturation, indicating that the OTUs detected were representative of the 16S amplicons (Figure S3). Alpha diversity metrics (ACE, Simpson, chao richness) showed no significant differences among susceptible or resistance profiles of BT (Table 1).

After filtering raw sequences obtained from the V3_V4 regions of 16S rRNA, 84,821 reads were obtained from samples of Bt_lantana-Q and Bt_pepper-B biotypes with an average length of 419 pb. A total of 4712 OTUs were obtained, of which 70% and 97.05% could be taxonomically clustered for Bt_lantana-Q and Bt_pepper-B (Table 1). Valid reads (99% for both populations) were rarefied to reads, and the rarefaction curve for each sample tended to saturation, indicating that the OTUs detected were representative of the 16S amplicons in each sample (Figure S3). Comparisons of the OTUs using alpha diversity metrics (ACE, Simpson, chao richness) showed no significant differences among susceptible or resistance profiles of Bt (Table 1).

Taxonomic assignment of OTUs revealed two phyla; the first is Proteobacteria, with a relative abundance of 66% and 95% for Bt_lantana-Q and Bt_pepper-B biotypes, respectively (Figure 3 and Figure 4). These results are consistent with previous studies reporting the dominance of the Proteobacteria phylum in the gut of B. tabaci [28], in the gut of mosquitoes [40], and in the gut of the insect Helcoverpa armigera [41]. Proteobacteria have been shown to assist insects in the degradation of carbohydrates [42], the detoxification of pesticides [43], and the synthesis of essential amino acids [44]. The second phylum is Cyanobacteria, identified only within the Bt_lantana-Q biotype with a relative abundance of 1% (Figure 3). Other phyla were also detected within the two biotypes, with less than 1%.

Within the Proteobacteria phylum, two classes were observed with uneven distribution between Bt_lantana-Q and Bt_pepper-B biotypes. The first class is γ-Proteobacteria, the most prevalent with relative abundance of 36% and 61% for Bt_lantana-Q and Bt_pepper-B biotypes, respectively (Figure 3 and Figure 4). Other findings reported the prevalence of the γ-Proteobacteria group since it encompasses mainly primary symbionts (P-endosymbionts)—bacteria known to form a mutualistic relationship within the host, supplying nutrients to its hosts [27,45], and enhancing fruitfulness and longevity [28]. The second class is α-Proteobacteria, represented by a lower relative abundance of 30% and 34% for Bt_lantana-Q and Bt_pepper-B biotypes, respectively. The alpha-proteobacteria group is considered a secondary and non-obligate symbiont (S-endosymbionts). It forms a less stable partner with their host and offers adaptive benefits to its partner [37].

At the genus level, three main shared genera were identified in this study within the Bt_lantana-Q and Bt_pepper-B (Figure 3 and Figure 4):

(i) P-endosymbiont Portiera (species Portiera aleyrodidarum) presents a relative abundance of 22% and 49% for Bt_lantana-Q and Bt_pepper-B biotypes, respectively. Candidatus Portiera (Figure 3B and Figure 4B), a Gram-negative bacterium lacking the outer cell wall membrane [27,46], was identified in the gut of B. tabacii populations. It usually resides in specialized cells within eukaryotes, such as bacteriocytes, forms a mutualistic relationship with insects, and is known to furnish hosts with essential amino acids and carotenoids [47,48]. On the other hand, we noticed in the current work the prevalence of the Portiera in the population resistant to organophosphates and pyrethroids, which suggests the involvement of Portiera in such resistance. Other studies have reported similar findings, indicating an increase in the abundance of Portiera in populations treated with thiamethoxam, whereas it declines in the presence of imidacloprid [47]. This demonstrates the variability or the specificity of the responses of the endosymbionts to different insecticide groups.

(ii) The S-endosymbiont Rickettsia genus (alpha-Proteobacteria phylum), like Portiera, was found to be more represented within and Bt_pepper-B (34%) than Bt_lantana-Q (30%) biotype (Figure 3B and Figure 4B). Rickettsia, a Gram-negative α-Proteobacterium [40], is a common insect facultative symbiont detected in seven B. tabaci species (MED, MEAM1, Asia II 1, Asia II 3, Asia II 7, China 1, and Sub-Saharan Africa) [8,49]. The S-endosymbiont bacteria form a less stable partner with their host and offer adaptive benefits to their partner, with either positive or negative effects or neutral in vital traits [27]. Limited findings have been reported on the role of specific symbionts in the insecticide resistance profile of the hosts. Some studies suggest insecticide resistance in these hosts may be associated with Wolbachia and Rickettsia infections. In contrast, other studies have shown that whiteflies (B. tabaci) infected with Rickettsia were more susceptible to five typical insecticides (thiamethoxam, imidacloprid, acetamiprid, pyriproxyfen, and spiromesifen) than uninfected whiteflies [50].

(iii) The S-endosymbiont Hamiltonella genus (γ-Proteobacteria phylum) represented 14% and 12% for MED-Q and MEAM1-B, respectively. Hamiltonella is a facultative endosymbiont found only in the B. tabaci Q and B biotypes [50,51], which agrees with our findings. It is a Gram-negative bacterium belonging to the Enterobacteriaceae family of γ-Proteobacteria and has so far been detected as an endosymbiont of sap-sucking insects, including whiteflies [52]. In whiteflies, B. tabaci has only been detected in the Mediterranean (MED), Middle East–Asia Minor 1 (MEAM1), and New World 2 species [8]. Global phylogenetic studies of Hamiltonella in whiteflies reveal only one group, suggesting a single ancestral acquisition of this endosymbiont [8,44]. It is noteworthy that in the current study, only two S-endosymbionts were found, Rickettsia and Hamiltonella, out of seven known within the B. tabaci group, including Wolbachia [53], Arsenophonus [54], Cardinium [55], Megaira [56], and Fritschea [57].

4. Discussion

Taken together, it can be stated that a relationship exists between the prevalence of the P-endosymbiont Portiera and the S-endosymbiont Rickettsia, and the resistance profile of organophosphates and pyrethroids, where both could be implicated in the resistance profile. In another study, higher insecticide susceptibility was correlated with a greater relative abundance of Rickettsia within B. tabaci [34]. Another study found that the thiamethoxam-resistant whitefly population contained a higher abundance of Rickettsia than the susceptible population while hosting fewer Portiera and Hamiltonella symbionts [48]. The different symbiont responses may result from symbiotic species or even strains from host genotypes. The interaction mechanism between insect resistance and symbionts is still poorly understood in insects. Noteworthy, the detoxification of organophosphates by a bacterial symbiont of the apple maggot was first studied by Boush and Matsumura [21], who reported the degradation of six different insecticidal active ingredients by an insect-associated bacterium. Further, Kikuchi et al. [22] showed that Burkholderia bacteria colonizing the guts of the bean bug confer resistance to the organophosphorus pesticide fenitrothion on its insect hosts.