Refractoriness of Sergentomyia schwetzi to Leishmania spp. is mediated by the peritrophic matrix

Background The peritrophic matrix (PM) is an acellular chitin-containing envelope which in most blood sucking insects encloses the ingested blood meal and protects the midgut epithelium. Type I PM present in sand flies and other blood sucking batch feeders is secreted around the meal by the entire midgut in response to feeding. Here we tested the hypothesis that in Sergentomyia schwetzi the PM creates a physical barrier that prevents escape of Leishmania parasites from the endoperitrophic space. Methodology/Principal findings Morphology and ultrastructure of the PM as well the production of endogenous chitinase in S. schwetzi were compared with three sand fly species, which are natural vectors of Leishmania. Long persistence of the PM in S. schwetzi was not accompanied by different morphology or decreased production of chitinase. To confirm the role of the PM in refractoriness of S. schwetzi to Leishmania parasites, culture supernatant from the fungus Beauveria bassiana containing chitinase was added to the infective bloodmeal to disintegrate the PM artificially. In females treated with B. bassiana culture supernatants the PM was weakened and permeable, lacking multilayered inner structure; Leishmania colonized the midgut and the stomodeal valve and produced metacyclic forms. In control females Leishmania infections were lost during defecation. Conclusions/Significance Persistence of the PM till defecation of the bloodmeal represents an important factor responsible for refractoriness of S. schwetzi to Leishmania development. Leishmania major as well as L. donovani promastigotes survived defecation and developed late-stage infections only in females with PM disintegrated artificially by B. bassiana culture supernatants containing exogenous chitinase.


Introduction
Sand flies (Diptera: Psychodidae) are the vectors of Leishmania species (Kinetoplastida: Trypanosomatidae) parasitizing humans. Over 800 species of sand flies have been described to date but only 98 species are proven or suspected vectors of human leishmaniases [1,2]. Development of Leishmania infections in the sand fly vectors is a complex, often species-specific process (reviewed by [3][4][5]). Some sand fly species can be infected with various Leishmania species (permissive vectors); other species are considered specific (restrictive) vectors which are only able to harbour the Leishmania species that they transmit in nature (e.g. Phlebotomus papatasi and L. major/L. turanica) [4,6,7].
Factors controlling vector competence act during the early phase of infection preceding or accompanying the defecation of the bloodmeal remnants. Blood digestion and early development of parasites occur inside the peritrophic matrix (PM), an acellular chitin-containing envelope which protects the midgut epithelium against damage and compartmentalizes the midgut into endo-and ectoperitrophic spaces [8]; in some hematophagous insects the PM performs also a central role in heme detoxification [9]. In sand flies, the PM was suggested to protect parasites against the action of digestive enzymes during the early hours post blood feeding [10]. Inside the endoperitrophic space surrounded by the PM, Leishmania amastigotes ingested with the bloodmeal transform to procyclic promastigotes. They must overcome the activity of the digestive enzymes, replicate and then escape to the ectoperitrophic space. This escape coincides with the transformation of procyclic forms to elongated nectomonads [11] which attach to the midgut epithelium to avoid expulsion from the midgut during defecation of undigested blood remnants. Failure of either (i) resistance to harmful environment caused by blood digestion, (ii) escape from the endoperitrophic space or (iii) attachment to the midgut epithelium leads to loss of infection in incompatible sand fly-Leishmania species combinations. In competent vectors, a substantial part of the parasite population survives defecation and colonizes the midgut. Finally the infection spreads anteriorly to be transmitted to the next host during subsequent blood meals [12].
Sand flies of the genus Sergentomyia are proven vectors of reptile Leishmania species, nonpathogenic to humans, which exhibit hypopylarian development patterns in their insect vectors (developing in the hindgut) with transmission occurring by predation of the infected flies by lizards. However, Sergentomyia species are not fully restricted to feeding on reptiles and at least some species feed on humans and/or mammalian reservoirs of Leishmania pathogenic to humans. Therefore, they have been suspected as vectors in some visceral leishmaniasis (VL) and cutaneous leishmaniasis (CL) foci where Sergentomyia spp. were abundant and found to harbour Leishmania (reviewed by [13]). Recently, a vectorial role was strongly suggested for three Sergentomyia species including S. schwetzi in the Mont-Rolland area in Senegal [14]. However, experiments showed that three species of Leishmania pathogenic to humans (L. donovani, L. infantum and L. major) did not survive defecation of bloodmeal remnants in S. schwetzi [15][16][17]. It was proposed that the crucial aspect mediating the refractoriness of S. schwetzi was the relative timing of degradation of the PM and defecation [17] i.e. the extremely long persistence of the PM in S. schwetzi, in comparison with three species of the genus Phlebotomus [18].
The aim of the current study was to compare the morphology and ultrastructure of the PM and activity of endogenous chitinase in S. schwetzi with three Phlebotomus species and to demonstrate that persistence of the PM represents an important mechanism of refractoriness of S. schwetzi. We showed that the PM of S. schwetzi shares similar morphological features with Phlebotomus species transmitting Leishmania. The role of the PM in refractoriness of S. schwetzi was demonstrated experimentally; the artificial disintegration of the PM using culture media from B. bassiana, rich in chitinase activity, enabled the development of mature infections of both L. major and L. donovani in S. schwetzi females.

Origin of exogenous chitinases
Chitinases from Streptomyces griseus and Trichoderma viride (both from Sigma, Cat. No C6137 and C8241, respectively) were diluted in PBS buffer and stored at -20 o C. Beauveria bassiana CCF 4422, obtained from the Culture Collection of Fungi (CCF, Dept. of Botany, Faculty of Science, Charles University, Prague, Czech Republic) was grown at 26 o C in a medium composed of (w/v) glucose 5%, Neopeptone (BD) 5%, Yeast extract (Sigma) 1%, NaCl (0.25%) in distilled water (modified after [25]). After three weeks of cultivation, the whole volume was centrifuged at 250 g for 10 min, the supernatant was sterilized using 0.22 μm millipore filter units (Millex-GP), concentrated by centrifugation through 30 kDa filters (Amicon) and stored at -80 o C.

Sand fly infections
Leishmania promastigotes from log-phase cultures (day 3-4 post inoculation) were resuspended in heat-inactivated rabbit blood (MVDr. Zdenek Petr, Czech Republic) and mixed 1:1 with a supernatant from B. bassiana culture containing chitinase (preparation described above) to obtain final concentration 1x10 6 promastigotes/mL and 0.07 U of exochitinase activity/mL. Control flies were fed on the same Leishmania suspension in blood mixed with the fresh culture medium or heat-inactivated culture supernatant from B. bassiana. Sand fly females (5-9 days old) were infected by feeding through a chick-skin membrane (BIO-PHARM, Czech Republic) on the suspension. Engorged sand flies were maintained in the same conditions as the colony. Females were dissected in several time intervals post bloodmeal (PBM), the density and location of Leishmania infections in their digestive tract was examined by fluorescence microscopy. Parasite loads were graded according to [20] as light (< 100 parasites per gut), moderate (100-1000 parasites per gut) and heavy (> 1000 parasites per gut). Experiments with each Leishmania-sand fly combination were repeated four to six times.

Morphometry of parasites
Midgut smears of sand flies infected with Leishmania parasites were fixed with methanol, stained with Giemsa, examined under the light microscope Olympus BX51 and photographed with an Olympus D70 camera. Body length, flagellar length and body width of parasites were measured using Image-J software. Four morphological forms were distinguished as described in [21]: procyclic promastigotes (PP), elongated nectomonads (EN), metacyclic promastigotes (MP) and short promastigotes (SP). Haptomonads were not distinguished as these forms remain attached on the gut and their numbers are often underestimated on gut smears. In total, 200 promastigotes from four females/smears were measured for each Leishmania species.

Light microscopy of the PM
Bloodfed females were dissected at 10 intervals after feeding on anesthetized BALB/c mice, starting immediately post bloodmeal (PBM) and at each of the following times: 1, 3, 6, 12, 24, 48, 72, 96 and 120 h PBM. Dissections were carried out in the isotonic saline solution with brief washing of the gut in distilled water in order to better separate the PM [22]. For each sand fly species and time interval, at least 20 females were analysed. Slides were observed under an Olympus BX51 microscope with Nomarski contrast and photographed with an Olympus D70 camera and software. The colour of the PM and presence of the anterior plug (AP) and posterior ending of the PM were checked.

Histology
Females at 24 h PBM were fixed at 4 o C for 48 h in AFA solution (formaldehyde: ethanol: acetic acid: distilled water, 1.5:12.5:1:10). After washing in phosphate-buffered saline (pH 7.6) and dehydration in 70% to 96% ethanol, the samples were embedded in JB-4 resin following the manufacturer's instructions (Polysciences). Histological sections (2-6 μm thick) were stained with Ehrlich's acid haematoxylin and 0.2% eosin, mounted on glass slides with Euparal Mounting Medium (BioQuipProducts) and observed and photographed. Ten females of each sand fly species were used.

Electron microscopy
Females were dissected at 24 h (P. argentipes) or 48 h PBM (P. orientalis, P. papatasi, S. schwetzi), fixed in modified Karnovsky´s fixative [23], and post-fixed with a 2% osmium tetroxide solution (both at 4 o C). The samples were dehydrated in ascending concentration of ethanol (35-100%) then acetone (100%) and embedded in SPURR resin (SPI-chemo). Semithin sections (500 nm) and ultrathin sections (80-90 nm) were obtained using a Reichert-Jung Ultracut E ultramicrotome. Semithin sections were stained with toluidine blue for light microscopy. Ultrathin sections stained with uranyl acetate and lead citrate [24] were observed and photographed with a Jeol 1011 transmission electron microscope with iTEM 5.1 software (Olympus). Thickness of the PM was measured on representative images using the Image-J software. Together 400 measurements per each sand fly species were obtained (four females per species, 10 images per female and 10 measurements per image).

Assessment of chitinase activity
The exochitinase activity of supernatants from B. bassiana culture (preparation described above) or dissected sand fly midguts was quantified using the Chitinase Assay Kit (Sigma, Cat.

Effect of chitinases from S. griseus and T. viride on Leishmania growth in vitro
L. major promastigotes (at concentration of 1x10 6 /mL) were exposed to chitinases from S. griseus and T. viride serially diluted in PBS in 96-well flat bottomed microtitre plates. Five μL of each dilution of chitinase were added to 200 μL of the Leishmania suspension in culture medium or defibrinated and inactivated rabbit blood. In controls, five μL of PBS were added. Numbers of parasites were counted after 24 h of cultivation at 26 o C using the Burker cell counter.

Effect of exogenous chitinases from S. griseus and culture supernatant from B. bassiana on Leishmania survival in vitro
L. major and L. donovani promastigotes at concentration of 1x10 6 /mL in a volume of 100 μL of culture medium were mixed with 100 μL of the supernatant from B. bassiana culture (prepared as described above) in 96-well flat-bottomed microtitre plates. The resultant concentration of chitinase was 0.07 U/mL (an identical concentration to that used for experimental infections of sand flies). As a control, 100 μL of the fresh culture medium for B. bassiana instead of the supernatant were added. For testing of the effect of S. griseus chitinase, 5 μL of chitinase solution in PBS was added to promastigotes at concentration of 1x10 6 /mL in a volume of 200 μL of culture medium to obtain chitinase concentration of 1 U/mL, 0.5 U/mL or 0.07 U/mL. As a control, 5 μL of PBS was added to the culture medium. Numbers of parasites were counted after 24 h of cultivation at 26 o C using the flow cytometer CytoFLEX S (Beckman Coulter, Inc., Brea, California, USA) equipped with 4 lasers (405 nm, 488 nm, 561 nm, 638 nm) and 13 fluorescence detectors. Dead cells were marked with DAPI (4 0 , 6-Diamidine-2 0 -phenylindole dihydrochloride, 0.005 mg/mL; Thermo Fisher Scientific). The mortality was assessed as a ratio of the number of dead cells showing both green and blue fluorescence (GFP and DAPI) to the number of live cells showing green fluorescence (GFP). In samples where dead cells were completely disintegrated and total numbers of detected cells were significantly lower in comparison with control samples, mortality was counted using the total numbers of parasites in control wells. Leishmania promastigotes killed by solution of 1% formaldehyde in PBS and permeabilised with 0.5% Triton X-100 (Sigma) were used as a control for dead cells. GFP was excited using 488 nm laser and its fluorescence emission was detected using 525/40 filter, DAPI was excited by 405 nm laser and detected using 450/50 filter. Analysis of cytometry data was performed using CytExpert software (Beckman Coulter). The experiments were conducted in duplets and repeated 2-3 times.

Statistical analysis
Measurements of parasites and mortality of parasites in vitro were compared using Analysis of Variance (ANOVA) including Tukey Post Hoc Test, measurements of the thickness of the PM were tested with Nested ANOVA and Post Hoc Test (t-test with Bonferroni-Holm correction). Differences in defecation between chitinase treated flies and control group were analysed by proportional test with Holm-Bonferroni correction. All the statistical evaluations were performed with statistical software SPSS version 23 or R software (http://cran.r-project.org).

Morphology and ultrastructure of the PM in four sand fly species
Morphology and ultrastructure of the PM of S. schwetzi was compared with three species of the genus Phlebotomus: P. orientalis, P. argentipes and P. papatasi to ascertain if there are specific morphological traits which may be connected with the long persistence of the PM observed in S. schwetzi [18]. Light microscopy and analysis of sections from specimens embedded in JB-4 resin showed that these four sand fly species did not differ substantially in the gross morphology of the PM. The anterior plug, i.e. the most anterior part of the PM secreted by the thoracic midgut, was formed in all the four sand fly species (S1 Fig). The PM was evenly closed on the posterior end, usually forming a short funnel-shaped closed structure called the posterior tail (Fig 1). In all the four species, the PM was originally transparent and then darkened as it became encrusted by heme, this process was apparent beginning 24 h PBM. The only parts of the PM which remained permanently transparent were the AP and the posterior tail.
Thickness and ultrastructure of the mature PM was assessed using images from electron microscopy. Based on knowledge of the PM kinetics [18], the optimal time for evaluation of the mature PM was 48 h PBM in P. orientalis, P. papatasi and S. schwetzi and 24 h PBM in P. argentipes (degradation of the PM proceeds faster in P. argentipes, before 48 h PBM [18]). The mature PM of S. schwetzi was as thick as those of P. orientalis and P. papatasi, while the PM of P. argentipes was significantly thinner than the PM of these three sand fly species (Table 1). This result corresponds with observations made during dissections of fed females: the PM of P. argentipes was extremely fragile and its extraction from the midgut was difficult. The ultrastructure of the mature PM also differed between P. argentipes and the remaining three sand fly species. While the mature PM consisted of thin solid outer layer and thick granular inner layer in S. schwetzi, P. orientalis and P. papatasi, the PM of P. argentipes appeared fairly  homogeneous throughout the entire cross-section (Fig 2). Thus, the PM of S. schwetzi did not show any unique morphological features in comparison with the other two vectors of the genus Phlebotomus. Midgut chitinase activity Next we tested chitinase activity in midguts of the four sand fly species to determine whether the long persistence of the PM in S. schwetzi might be caused by lower activity of midgut chitinase in this species. Interestingly, the dynamics and levels of exochitinase activity of S. schwetzi were similar to those in P. papatasi and P. orientalis: activity was detectable at 24 h PBM, peaked at 48 h PBM and then decreased at 72 h PBM (Fig 3). Moreover, the activity levels observed in S. schwetzi by 72 h PBM were the highest among all four sand fly species studied. A distinct course of exochitinase activity was observed in P. argentipes: the activity peaked at 24 h PBM, decreased by 48 h and declined almost to zero by 72 h PBM. Hence, the long persistence of the PM in S. schwetzi cannot be explained by low exochitinase activity in the midgut.

Effect of exogenous chitinases on the survival of Leishmania promastigotes in vitro
Before feeding exogenous chitinase to Leishmania-infected sand flies, we had to prove that chitinase has no lethal effect on the parasites. However, preliminary experiments showed that both commercially available chitinases from S. griseus and T. viride significantly decreased growth or even killed Leishmania promastigotes in vitro at all concentrations tested. (S1 Table). Therefore,  Table)-it killed L. major and L. donovani at all tested concentrations in the same rate as did the negative control (parasites killed by 1% solution of formaldehyde) (Fig 4, P > 0.05 for all chitinase concentrations and both Leishmania species, ANOVA and Tukey Post Hoc Test).

Effect of culture supernatant from B. bassiana on the PM of S. schwetzi and development of Leishmania parasites in vivo
Females of S. schwetzi were membrane-fed on a mixture of blood with culture supernatant from B. bassiana containing chitinase to disrupt the PM and mimic early disintegration of the  PM occurring in most Phlebotomus species. Light microscopy 24 h PBM revealed that the PM was disintegrated in 94% of chitinase-treated females (N = 49) while no disintegration was observed in control females (N = 35). By day 3 PBM the PM in control flies was consistently dark and opaque due to incrustation with heme ( Fig 5A) while the PM in chitinase-treated flies remained mostly transparent with patches of dark coloration (Fig 5B). Presence of red blood cells in the ectoperitrophic space was indicative of partial disintegration and increased permeability of the PM in chitinase-treated females. Electron micrographs of cross sections of the AMG on day 2 PBM confirmed the differences at the structural level. While the PM in control flies was thick and multilayered (Fig 5C), addition of B. bassiana culture supernatant to the bloodmeal resulted in thin PMs without apparent inner structure, only scarcely distinguishable from the background (Fig 5D). Surprisingly, although the PM in flies treated with the culture supernatant was weakened and permeable, it maintained its structure in midguts longer than controls (even till day10 PBM) as the defecation in this group was significantly delayed (S2 Fig). The same culture supernatant from B. bassiana with 0.07 U/mL of chitinase was used for experimental infections of S. schwetzi with Leishmania parasites. Addition of culture supernatants significantly enhanced development of L. donovani and L. major in S. schwetzi. In control females fed on heat-inactivated culture supernatants or fresh culture media mixed with rabbit blood, neither species of Leishmania were able to develop mature infections, only very exceptionally survived defecation and then the infections were lost. On the other hand, in groups treated with B. bassiana culture supernatants, both Leishmania species escaped from the PM (Fig 6A and 6B), developed heavy infections in most females (Fig 7A and 7B) and by day 10 PBM colonized the stomodeal valve in 12% and 36% of females infected with L. donovani and L. major, respectively (Fig 6C and 6D and Fig 7C and 7D). Moreover, analysis of morphological forms revealed the presence of metacyclic promastigotes in late infections of both Leishmania species (Table 2).
Despite colonization of the ectoperitrophic space, the infection rates in females treated with culture supernatants gradually decreased from more than 90% on day 1 PMB to 46% and 47% in L. donovani and L. major, respectively, on day 10 PBM. In sand flies post defecation, infection rates ranged usually between 20% and 30% in both Leishmania species (Fig 7E and 7F) and attachment of parasites to the midgut epithelium was never observed.

Glycosylation of the midgut epithelium
Midgut mucins bearing O-linked glycosylation with terminal N-acetyl-galactosamine have been previously demonstrated to participate in Leishmania attachment in so called permissive sand fly vectors [6,26,27]. Therefore, midgut lysate of S. schwetzi was subjected to SDS-PAGE followed by blotting with HPA lectin specific for N-acetyl-galactosamine. Midgut lysates of P. argentipes and P. papatasi served as a positive and negative control, respectively [26]. In S. schwetzi midgut lysate, lectin HPA showed specific reactivity with a molecule of apparent molecular mass around 35-40 kDa; however, the reaction was weak in comparison with P. argentipes. Control strips with inhibitory sugar added were negative, confirming the specificity of HPA binding (S3 Fig).

Discussion
We demonstrated that the structure and thickness of the PM of S. schwetzi is very similar to those of P. orientalis and P. papatasi. It consists of a thin solid outer layer and an amorphous thick inner layer which correspond with data on the ultrastructure of PMs of several other species of the genera Phlebotomus and Lutzomyia [11,[28][29][30]. On the other hand, the PM of P. argentipes was thin and homogeneous without distinct outer and inner layers which may be related to the exceptionally fast blood digestion described for this sand fly species [18]. The gross morphology of the PM of S. schwetzi also resembled other tested sand fly species: all four species had an anterior plug secreted by the thoracic midgut repeatedly detected in sand flies [11,[29][30][31]. The posterior end of the PM was closed, with posterior tail similar to that observed previously in P. duboscqi and P. papatasi [11,32].
As we did not find significant differences between the PM of S. schwetzi and PMs of other sand fly species we tested the hypothesis that the persistence of the PM of S. schwetzi could be caused by lower activity of midgut chitinases. Surprisingly, experimental data suggest the opposite; the exochitinase activity in S. schwetzi midguts was the highest of the four sand fly species tested. Chitin degradation in insects is catalysed by various chitinolytic enzymes, for example in mosquitoes, chitinolysis is controlled by at least two distinct chitinases (endo-and exochitinase) and β-N-acetylglucosaminidase [33]. Therefore, we cannot exclude the possibility that low activity of enzymes other than exochitinase might be responsible for delayed  degradation of PM in S. schwetzi. High exochitinase activity in S. schwetzi midguts may be connected with the fact that chitinolytic activities of chitinases are related not only to degradation, but are involved also in synthesis and modulation of the PM [33].
Experiments described here and our previous studies [17,18] were performed with S. schwetzi from the colony originating in north-western Ethiopia and infected with L. donovani parasites isolated in the same region. Previously, S. schwetzi originating from Kenya were showed to be refractory to local L. major and L. donovani [15,16]. The origin of sand flies and parasites is important mainly in context of recent findings from the Mont-Rolland region in Senegal, where L. infantum DNA was found in 4.19% of captured S. schwetzi females and living parasites were observed in anterior midgut of a single female without a bloodmeal [34]. This fact evokes the question as to what extent is the vector competence of S. schwetzi populationspecific. This species shows a substantial morphological variability and two morphological forms of S. schwetzi ("typical" and "atypical") were found in Senegal [35,36], a large area from Sudan to West Africa [37], Uganda [37] and Kenya [38]. Therefore, revision of the taxonomy of S. schwetzi and further experimental studies with sand flies and Leishmania parasites originating from different areas would be very interesting. In addition, it is not clear if the prolonged persistence of the PM is the common feature of the genus Sergentomyia. Previously the PM was studied in a single member of this genus; S. arpaklensis (corresponds to S. sintoni based on recent nomenclature). Its PM did not break down with the end of digestion and was defecated intact [39][40][41], presumably causing refractoriness for the reptile pathogen L. gymnodactyli [40].
Importance of the PM in vector competence was supported experimentally by several authors. Addition of a chitinase inhibitor allosamidin to the infective bloodmeal led to entrapment of L. major within the PM of P. papatasi [10]. Similarly, silencing of the gene for PpChit 1, a sand fly-derived chitinase involved in degradation of the PM [42], reduced L. major load in midguts of P. papatasi [43] while knockdown of the PpPer1 gene (peritrophin involved in the formation and scaffolding of the PM [42]) led to increase in the parasite load [44].
Here, the addition of the supernatant from B. bassiana culture with chitinase activity to the bloodmeals of S. schwetzi led to disintegration of the PM which enabled survival of Leishmania   [10], the absence of PM caused by addition of exogenous commercial chitinase to the bloodmeal of P. papatasi was associated with the total loss or 20% reduction of amastigoteand promastigote-initiated L. major infections, respectively. Authors suggested that the lack of the PM exacerbated lethal conditions as a result of proteinase activity in the blood-fed midgut. Partial degradation of PM due to culture supernatants of B. bassiana, together with delayed defecation, enabled Leishmania parasites to establish the infection in the midgut of S. schwetzi. However, we did not observe the attachment of promastigotes to the midgut epithelium and infection rates decreased with time post bloodmeal. This could be explained by the shortage of ligand molecules in the midgut as O-glycosylated epitopes in S. schwetzi midgut were present in very low amounts, significantly lower than in permissive species P. argentipes [26]. The presence of O-glycosylated mucins in a permissive vector P. argentipes as well as their absence in P. papatasi, a specific vector of L. major, is in accordance with previously described data [26]. Other factors, like antimicrobial peptides [45,46] may also negatively influence the establishment of Leishmania infections in the gut of S. schwetzi.
The results of the present study clearly confirm that the prolonged persistence of the PM till defecation contributes significantly to the refractoriness of Ethiopian S. schwetzi to Leishmania parasites.