Bacterial diversity of the American sand fly Lutzomyia intermedia using high-throughput metagenomic sequencing

Parasites of the genus Leishmania cause a broad spectrum of diseases, collectively known as leishmaniasis, in humans worldwide. American cutaneous leishmaniasis is a neglected disease transmitted by sand fly vectors including Lutzomyia intermedia, a proven vector. The female sand fly can acquire or deliver Leishmania spp. parasites while feeding on a blood meal, which is required for nutrition, egg development and survival. The microbiota composition and abundance varies by food source, life stages and physiological conditions. The sand fly microbiota can affect parasite life-cycle in the vector. We performed a metagenomic analysis for microbiota composition and abundance in Lu. intermedia, from an endemic area in Brazil. The adult insects were collected using CDC light traps, morphologically identified, carefully sterilized, dissected under a microscope and the females separated into groups according to their physiological condition: (i) absence of blood meal (unfed = UN); (ii) presence of blood meal (blood-fed = BF); and (iii) presence of developed ovaries (gravid = GR). Then, they were processed for metagenomics with Illumina Hiseq Sequencing in order to be sequence analyzed and to obtain the taxonomic profiles of the microbiota. Bacterial metagenomic analysis revealed differences in microbiota composition based upon the distinct physiological stages of the adult insect. Sequence identification revealed two phyla (Proteobacteria and Actinobacteria), 11 families and 15 genera; 87 % of the bacteria were Gram-negative, while only one family and two genera were identified as Gram-positive. The genera Ochrobactrum, Bradyrhizobium and Pseudomonas were found across all of the groups. The metagenomic analysis revealed that the microbiota of the Lu. intermedia female sand flies are distinct under specific physiological conditions and consist of 15 bacterial genera. The Ochrobactrum, Bradyrhizobium and Pseudomonas were the common genera. Our results detailing the constituents of Lu. intermedia native microbiota contribute to the knowledge regarding the bacterial community in an important sand fly vector and allow for further studies to better understand how the microbiota interacts with vectors of human parasites and to develop tools for biological control.


Background
Parasites of the genus Leishmania cause a broad spectrum of diseases in humans worldwide, collectively known as leishmaniasis. American cutaneous leishmaniasis (ACL) is a neglected disease found in 18 countries and approximately 60,000 new cases are diagnosed every year [1]. In Brazil, it is present in all states [2] and is prevalent mainly in the northern and northeastern regions. Even though several Leishmania spp. have been described, Leishmania braziliensis is the main species causative of ACL and can be transmitted by different sand fly species in distinct endemic geographical regions. In fact, there are several sand fly species recognized as proven vectors of L. braziliensis, including Lutzomyia intermedia, Lutzomyia whitmani, Lutzomyia neivai, Lutzomyia migonei, Lutzomyia wellcomei and Lutzomyia complexa [3].
In the southeastern region of Brazil, Lutzomyia intermedia is the main vector of L. braziliensis [4,5]. This opportunistic species is associated with forests, nature reserves and modified domestic and peridomestic environments, such as animal shelters [6][7][8].
The female sand fly can acquire or deliver Leishmania spp. parasites while feeding on a blood meal, which is required for nutrition, egg development and survival. All events involved in parasite transformation, development, survival and establishment of vector infection occur in the digestive tract. A variety of microorganisms, such as fungi and bacteria, can interact with Leishmania spp. in the vector and during their transmission to a vertebrate host. Bacterial communities have been studied to understand how bacteria interact with parasites. The insect microbiota varies by food source and life stage; during the immature stage, sand fly larvae ingest microorganisms from the soil environment. Furthermore, adult sand flies have the opportunity to ingest microorganisms, either from plant sap or sticky honeydew from aphids. Alternatively, adult females also ingest yeasts and bacteria from their vertebrate hosts used as blood source [9][10][11][12][13]. All these microbes will contribute to form their midgut microbiota.
In 2002, Volf et al. [10] studying P. duboscqi showed microbiota modulation due to blood meal. Two days after the ingestion of a blood meal, there was an increase in bacterial number in the midgut followed by a decrease after blood meal digestion. This fact is in line with other studies performed with other nematoceran insects (i.e. Aedes aegypti and Anopheles gambiae) showing an increase in their midgut microbiota after a blood meal [14,15]. More recently, in vitro effect of microbiota on Leishmania spp. survival was demonstrated showing that Serratia, Bacillus and Haemophilus parainfluenzae induced lysis of promastigote forms [12,16,17]. Moreover, sand flies pre-fed with Pseudozyma, Asaia or Ochrobactrum showed a reduction in parasite survival [18].
Here, we performed a metagenomic analysis for microbiota composition and abundance in female adult sand flies, specifically in Lu. intermedia originated from an endemic area in Brazil. This study establishes basic knowledge for future studies to understand the role of microbiota during the course of L. braziliensis infection in this important vector.

Sand flies
Lutzomyia intermedia adults were collected using CDC light traps [19] from a locality in Corte de Pedra, State of Bahia, Brazil (longitude 59°30'W; latitude 13°26'S); a hilly region (600 m above sea level) covered by different stages of secondary Atlantic forests. Corte de Pedra is a village with human interference involved with cocoa and clove plantations. The light traps were placed around animal shelter in April 2014.

Specimen preparation
The collected sand flies were anesthetized at -20°C for a few seconds and had their surfaces carefully sterilized to prepare for serial passage as follows: 10 s in 70 % ethanol, 1 min in 1 % hypochlorite and 1-min wash in PBS three times [20]. The specimens were morphologically identified according to Young & Duncan [21]. Sand fly midguts were dissected under a stereoscope and separated into groups according to their physiological condition: (i) absence of blood meal (unfed = UN); (ii) presence of blood meal (blood-fed = BF); and (iii) presence of developed ovaries (gravid = GR). The pooled samples (300 of each group) were stored in micro centrifuge tubes with (100 μl) of RNAlater® (Qiagen, Hilden, GER) at 4°C for later RNA extraction and metagenomic analysis.

Metagenomic process RNA extraction
Samples previously stored in RNAlater were washed in distilled water using a 35 μm nylon filter prior to RNA extraction. RNA extraction was performed in two steps. First, the pooled samples were ground in TRI Reagent® RT (Molecular Research Center, Inc., Cincinnati, USA) and chloroform with Teflon pestle and then incubated at room temperature for 15 min. The samples were centrifuged at 15,294 rcf at 4°C for 15 min followed by the addition of 70 % ethanol to the supernatant. The supernatant was transferred to a column (RNeasy Mini Kit, Qiagen, Hilden, GER) and processed according to the manufacturer's protocol. RNA quality was assessed using a NanoDrop spectrophotometer (Thermo Fisher Scientific, Massachusetts, USA), and 50 ng/μl of RNA was used to synthetize cDNA using the Reverse Transcription Kit (Qiagen, Hilden, GER). cDNA was diluted 10×, stored at -20°C and submitted for high-throughput Illumina sequencing.

Sequence analysis and taxonomic profiling
The reads corresponding to the targeted 16S rRNA V1-V3 region were QC processed using the tools implemented within the Quantitative Insights Into Microbial Ecology (QIIME) software v1.7.0 pipeline [22]. Briefly, primer sequences were removed, and reads were trimmed at the extremities using a Phred value Q20. High quality reads (≥ 250 bp) were further processed to remove chimera sequences using ChimeraSlayer. Sequences within an identity threshold of 97 % were clustered into Operational Taxonomic Units (OTUs) using the UCLUST [23] pipeline. The most abundant sequence within each of the clusters was selected as representative, and its taxonomic affiliation was assigned using the Ribosomal Database Project (RDP) classifier [24] via QIIME considering a minimum confidence value of 80 %.
The bacterial community composition of each experimental group was compiled considering only OTUs encompassing at least a 0.5 % relative abundance. The cutoff value was set to address rare sequences that most likely corresponded to random sequencing errors (eg, previously reported thresholds for community composition tables in other insect models vary from 1 to 0.04 % [25,26].

High-throughput Illumina sequencing
Sequencing revealed a total of 587,144 valid reads, which were grouped as follows: UN: 117,492; GR: 44,318; and BF 425,334. After the sequences were quality checked and assembled, the UN group was clustered into 66 OTUs, the GR group into 30 OTUs, and the BF group into 40 OTUs; only 12, 9 and 8 OTUs in the UN, GR and BF groups respectively, were considered valid bacterial OTUs according to the cut-off value. An average of 2 % reads were unassigned and 2.9, 1.4, and 1.4 % of OTUs fell below the cut-off line in the UN and GR and BF groups, respectively.

Diversity of bacterial microbiota
Bacterial metagenomic analysis revealed differences in microbiota composition based upon the distinct physiological stages of the adult insect. Sequence identification revealed two phyla (Proteobacteria and Actinobacteria), 11 families and 15 genera; 87 % of the bacteria were Gram-negative, while only one family and two genera were identified as Gram-positive (Table 1).

Correlation in diversity and abundance
Lutzomyia intermedia microbiota diversity and abundance varied based on the studied groups. In the UN and GR groups, the family Enterobacteriaceae was the most abundant, representing 54 % and 47 % of each group, respectively. Interestingly, in the BF group, the family Rickettsiaceae was the most prevalent, with 46.7 % of the group represented by Wolbachia and only 4.3 % of the group represented by Enterobacteriaceae. The genera Ochrobactrum, Bradyrhizobium and Pseudomonas were found across all of the groups.
The abundance of the genera Bradyrhizobium and Pseudomonas increased after the blood meal (6.0-fold and 3.8-fold, respectively), but abundance increased more in gravid sand flies (8.2-fold and 18.0-fold, respectively). Particularly, the abundance of Ochrobactrum bacteria decreased 2.2-fold and 3.7-fold in the BF and GR sand fly groups, respectively (Fig. 1).

Discussion
Knowledge of the microbiota in the vectors of humanborne diseases is essential to comprehend the relationships with the pathogens they transmit. To our knowledge, this is the first investigation of the native microbiota of Lu. intermedia, a proven vector of L. braziliensis. Sand flies were collected near Corte de  Pedra, a small village and cutaneous leishmaniasis endemic area where 2,384 new cases of ACL were registered in 2015 [27]. Metagenomic analysis revealed only two phyla: (i) Proteobacteria, a dominant phylum with ten families and 13 genera, and (ii) Actinobacteria, with only one family and two genera. The majority of these families originate from the environment [28][29][30]. Sand flies can acquire several species of bacteria from the soil during larva nurturing and in plant sap during adult feeding [10][11][12][13]. The most abundant family in the UN and GR groups was the Enterobacteriaceae, while the Rickettsiaceae was the most prevalent in the BF group. The Enterobacteriaceae is described as the most common family present in insects. Although the majority of the data was obtained from the UN and GR groups (64,917 and 20,809 reads, respectively), it was not possible to identify the Enterobacteriaceae members into genera taxon due to the similarity of the 16S rRNA variable regions [31]. In fact, these data are supported by other studies that detected members of Enterobacteriaceae in New and Old World sand flies using cultivation-dependent techniques [9,[32][33][34].
We assumed that the microbiota from the midgut of unfed Lu. intermedia sand flies is a quiescent community of bacteria. These sand flies had no previous contact Fig. 1 The circular web represents the relative abundance of the bacterial components of Lutzomyia intermedia microbiota identified by New Generation Sequencing. Generated with Circos online tool [39] with blood meal; as such, this group was considered a control group and was used to proportionally compare the abundance of bacterial genera present in the other groups. Fed sand flies showed a 33.4-fold and 2.45-fold increase in the abundance of Kluyvera and Wolbachia, respectively, compared to controls due to the presence of blood meal in the midgut. Moreover, the abundance of the genera Bradyrhizobium and Pseudomonas also increased after the blood meal (6.0-fold and 3.8-fold, respectively), but the abundance increased more in gravid sand flies (8.2-fold and 18.0-fold, respectively).
Particularly, the abundance of Ochrobactrum bacteria decreased 2.2-fold and 3.7-fold in the blood-fed and gravid sand flies, respectively. Curiously, the Enterobacter present in the unfed sand flies were not detected in the blood-fed sand flies but were present at the same abundance in the gravid sand flies. It is interesting to observe that Kluyvera and Pseudomonas, both found in soil and plants, and Wolbachia, a maternally transmitted endosymbiont, are the microbiota genera that were most enhanced by the presence of blood meal in the midgut.
The family Enterobacteriacea remained present as the second most abundant after the blood meal represented by the genus Kluyvera. In our opinion, this is expected since this family is favored in hematophagous insects after the blood meal. In 2011, Wang et al. [15] showed that members of Enterobacteriacea proliferate under oxidative stress due to their capacity to cope with the oxidative stress. Interestingly, Wolbachia spp. were most predominant. Wolbachia spp. are intracellular organisms and for this reason are protected from the highly oxidative environment in the gut lumen. Until today, the nature of the relation between Wolbachia spp. and sand flies is still obscure.

Conclusion
Our metagenomic analysis revealed that the microbiota of female Lu. intermedia sand flies are distinct under specific physiological conditions and consist of 15 bacterial genera with Ochrobactrum, Bradyrhizobium and Pseudomonas being the common genera. We demonstrated that the presence of a blood meal influences bacterial abundance in the midgut, in a more exacerbated way than the gravid condition. Our results detailing the constituents of Lu. intermedia native microbiota contribute to the knowledge of bacterial communities in an important sand fly vector and allow for further studies to better understand how the microbiota interacts with vectors of human parasites and to develop tools for biological control.