Leishmania amazonensis from distinct clinical forms/hosts has polymorphisms in Lipophosphoglycans, displays variations in immunomodulatory properties and, susceptibility to antileishmanial drugs

Abstract Lipophosphoglycan (LPG), the major Leishmania glycoconjugate, induces pro‐inflammatory/immunosuppressive innate immune responses. Here, we evaluated functional/biochemical LPG properties from six Leishmania amazonensis strains from different hosts/clinical forms. LPGs from three strains (GV02, BA276, and LV79) had higher pro‐inflammatory profiles for most of the mediators, including tumor necrosis factor alpha and interleukin 6. For this reason, glycoconjugates from all strains were biochemically characterized and had polymorphisms in their repeat units. They consisted of three types: type I, repeat units devoid of side chains; type II, containing galactosylated side chains; and type III, containing glucosylated side chains. No relationship was observed between LPG type and the pro‐inflammatory properties. Finally, to evaluate the susceptibility against antileishmanial agents, two strains with high (GV02, BA276) and one with low (BA336) pro‐inflammatory activity were selected for chemotherapeutic tests in THP‐1 cells. All analyzed strains were susceptible to amphotericin B (AmB) but displayed various responses against miltefosine (MIL) and glucantime (GLU). The GV02 strain (canine visceral leishmaniasis) had the highest IC50 for MIL (3.34 μM), whereas diffuse leishmaniasis strains (BA276 and BA336) had a higher IC50 for GLU (6.87–12.19 mM). The highest IC50 against MIL shown by the GV02 strain has an impact on clinical management. Miltefosine is the only drug approved for dog treatment in Brazil. Further studies into drug susceptibility of L. amazonensis strains are warranted, especially in areas where dog infection by this species overlaps with those caused by Leishmania infantum.

Brazil (Lainson et al., 1994;F. T. Silveira et al., 2004). Among the cutaneous forms, ADCL is the most severe as therapeutic failures are common. A distinguished feature of this form is an impairment in cellular responses, causing T cells anergy and lack of delayedtype hypersensitivity (Convit et al., 1972;Desjeux, 2004; F. T. Silveira et al., 2005). The anergic nature of L. amazonensis remains obscure, although several mechanisms have been suggested (Real et al., 2013).
The reduced incidence of the number of L. amazonensis LCL cases in the north and northeast of Brazil, where this parasite is distributed, may be a result of its zoonotic and occupational patterns (Camara Coelho et al., 2011;J. P. de Oliveira et al., 2007). This species is transmitted by Bichromomyia flaviscutellata, a sand fly species found in the ground of forested areas, having wild rodents as hosts (Lainson & Shaw, 1968). Bats have also been suggested as opportunistically involved in wild cycles outside the Amazon region (E. F. de Oliveira et al., 2015;Savani et al., 2010). B.
flaviscutellata is widely distributed in the Amazon region and other Brazilian states (Carvalho et al., 2015). In Minas Gerais, southeastern Brazil, L. amazonensis was found in wild-caught sand flies (M. S. Cardoso et al., 2019;Rêgo et al., 2015) and in dogs, causing CVL (Dias et al., 2011;Valdivia et al., 2017). This finding is of particular concern since in 2017, the treatment of dogs with miltefosine (MIL) was approved in Brazil. Most CVL cases are caused by Leishmania infantum and the natural resistance of L. amazonensis to antileishmanial drugs (Bittencourt et al., 1989;Convit et al., 1989) may lead to therapeutic failure. However, the drug susceptibility profile from a viscerotropic L. amazonensis causing CVL is unknown.
The exuberant growth of L. amazonensis in culture facilitated its use as a model species for immunology and chemotherapy (Rocha et al., 2013;Rodrigues et al., 2010). The classical TH1/TH2 phenotype, observed for Leishmania major in C57BL/6 and BALB/c mice, is not followed by L. amazonensis. It causes severe cutaneous lesions in both mice with a mixed cytokine profile (Pereira & Alves, 2008). Several reports attempted to elucidate Leishmania virulence factors during infection, especially those involving the parasite glycoconjugates. Lipophosphoglycan (LPG), the major cell surface glycoconjugate of Leishmania, has been implicated in a wide range of functions (de Assis et al., 2012). Regarding dermotropic species, functional studies of L. amazonensis LPGs have shown their role in macrophages and neutrophils. Those include induction of neutrophil extracellular traps (Guimarães-Costa et al., 2009), double-stranded RNA-dependent protein kinase (PKR) ( de Carvalho Vivarini et al., 2011), LTB 4 (Tavares et al., 2014), NO/cytokines via TLR4 (Nogueira et al., 2016), caspase-11 via NLRP3 (de Carvalho et al., 2019, and IL-32 via TLR2/NOD2 (Silveira et al., 2022 (Mahoney et al., 1999;Nogueira et al., 2017;Paranaíba et al., 2015;Soares et al., 2004). In the past few years, some studies have increased the panel of strains from different clinical manifestations/hosts. LPGs from viscerotropic/dermotropic L. infantum possess three types of LPG: (I) without side chains, (II) with one β-glucose linked to the repeat units, and (III) with two to three β-glucoses as side chains. Those polymorphisms did not affect sand fly development but affected NO/cytokine production by murine macrophages (Cardoso et al., 2020;Coelho-Finamore et al., 2011;Soares et al., 2002). Like L. infantum, Leishmania braziliensis LPGs from different clinical forms also displayed unbranched and branched sugars in their repeat units. These polymorphisms did not correlate with NO and cytokine production by murine macrophages (Vieira et al., 2019). Finally, preliminary reports on L. amazonensis LPG showed glycosylated and galactosylated side chains in the strains PH8 and Josefa, isolated from sand flies and humans, respectively. These LPGs were potent TLR4 agonists and induced NO and cytokine production by murine macrophages. However, they did not affect sand fly interaction with Migonemyia migonei and Lutzomyia longipalpis (Nogueira et al., 2016(Nogueira et al., , 2017.
As part of a wider project on the glycobiology of Leishmania parasites, we evaluated the role of L. amazonensis LPGs from distinct clinical forms/hosts during interaction with murine macrophages. Since we have a valuable panel of strains, we also evaluated their susceptibility profile against antileishmanial drugs.

| Extraction, purification, and quantitation of LPG
LPGs were extracted with organic solvents and purified using phenyl-Sepharose from late log-phase cells (Nogueira et al., 2017). Organic eluates were dried through N 2 evaporation and purified LPGs were resuspended in endotoxin-free water (Sanobiol) and quantitated using the phenol-sulfuric method (Dubois et al., 1956). The LPG concentrations were adjusted to 10 μg/mL in RPMI before functional experiments (Nogueira et al., 2016) (Figure 1a).

| Cytokine, chemokine, and nitrite measurements
To evaluate the production of different mediators in response to LPG and LPS, macrophage culture supernatants were collected after 48 h of incubation. Tumor necrosis factor alpha (TNF-α), interleukin 6 (IL-6), IL-10, IL-12p70, and MCP-1 concentrations were determined using BD cytometric bead array (CBA) Mouse Inflammation Kit (BD Biosciences) according to the manufacturer's specifications. Flow cytometry measurements were performed on FACSVerse flow cytometry (BD Biosciences).
The Cell-Quest software package provided by the manufacturer was used for data acquisition and the FlowJo v. 7.6.4 (Tree Star Inc.) was used for data analysis. A total of 2500 events were acquired for each analysis. The results are representative of three experiments in duplicate. Nitrite (NO) concentrations were determined by the Griess reaction (Drapier et al., 1988).

| Fluorophore-assisted carbohydrate electrophoresis (FACE)
As LPGs have differently activated murine macrophages, our next step was to evaluate the existence of sugar polymorphisms in their repeat units. LPGs were depolymerized after mild acid hydrolysis and phosphorylated repeat units were recovered after butanol:water partition (Soares et al., 2002). Then, they were treated with alkaline phosphatase (1 U) in 15 mM Tris buffer (pH 9.0, 16 h, 37°C). Samples were desalted by passage through a two-layered column of AG50W-X12 (H+) over AG1-X8 (acetate) and fluorescently labeled with 0.05 N ANTS (8-aminonaphthalene-1,3,6-trisulfate) and 1 M cyanoborohydride for 16 h, 37°C (Soares et al., 2004). After this step, samples were subjected to FACE analysis, including the labeled oligoglucose ladders (G 1 -G 7 ) used as standards. For monosaccharide analysis, the repeat units were subjected to strong acid hydrolysis (2 N trifluoroacetic acid, 3 h, 100°C). Samples were desalted as

| Macrophage experimental infection and drug susceptibility assay
To evaluate the susceptibility of L. amazonensis strains to current antileishmanial drugs, chemotherapeutic assays were performed as previously reported (Rugani et al., 2018). Briefly, human monocytederived macrophages from THP-1 cell line (ATCC#TIB-202) were infected with 2 × 10 5 stationary-phase promastigotes (MOI 10:1) for 3 h. Noninternalized parasites were removed, and cells were incubated for 72 h in the presence/absence (untreated control) of amphotericin B (AmB), MIL, and glucantime (GLU). Assays were performed twice in three replicates. The infection index was obtained by dividing the total number F I G U R E 1 Strategies employed for Leishmania amazonensis strains characterization. (a) Extraction, purification, dot-blots, and interaction with murine macrophages. Parasite cell pellets were subject to extraction with organic solvents as described elsewhere. For purification, the solvent E extract was dried under N 2 evaporation and applied into a phenyl-Sepharose column. Purified lipophosphoglycan (LPGs) were used for biological and immunological assays. (b) Biochemical characterization of LPG repeats units by fluorophore-assisted carbohydrate electrophoresis (FACE). LPG was depolymerized, subjected to butanol:water partition and treated with alkaline phosphatase. After desalting, neutral repeat units were subjected to FACE. (c) Chemotherapeutic assays. THP-1 cells were exposed to parasites (MOI 10:1) before drug exposure and IC 50 determination.

| Statistical analysis
For nitrite, cytokine, and chemokine measurements, the Shapiro-Wilk normality test was conducted to test the null hypothesis that data were sampled from a Gaussian distribution. 3 | RESULTS

| Functional analysis
LPGs were able to differentially stimulate the production of different mediators by peritoneal murine macrophages (Figure 1). LPGs from BA276, LV78, and GV02 strains induced higher levels of NO, IL-6, and TNF-α than the others (p < .05) (Figure 2a-c). Those levels were even higher than that for LPS (positive control). The heterogeneous F I G U R E 2 Nitrite (a) and cytokines/chemokine (b-f) production by interferon-gamma primed murine macrophages stimulated with lipophosphoglycan (LPGs) from distinct Leishmania amazonensis strains. Nitrite concentration was measured by Griess reaction and cytokine concentrations were determined by flow cytometry. Negative control and positive control were medium and lipopolysaccharide (100 ng/ml), respectively. Asterisks indicate statistical differences (p < .05).

C Cell ell B Biology iology I International nternational
| 1951 production of IL-12p70 and MCP-1 was observed among strains being statistically higher than the BA336 strain (Figure 2d,e). Finally, IL-10 production was higher for GV02 and LV78 LPGs only ( Figure 2f).

| LPG polymorphisms in L. amazonensis strains
To investigate if functional variations could be due explained by the LPG polymorphisms, these molecules were biochemically characterized. They displayed qualitative differences in their repeat units by (types I-III) from different clinical forms/hosts are summarized in Table 2.

| Susceptibility to antileishmanial drugs
As L. amazonensis strains elicited distinct immunomodulatory responses on macrophage experimental infection, we investigated the drug susceptibility status of three strains: one isolated from a dog (GV02) and two (BA276 and BA336) isolated from ADCL patients that are often resistant to available antileishmanial protocols (Zauli-Nascimento et al., 2010). All strains were sensitive to GLU, AmB, and MIL in a dose-dependent manner ( Figure 5). However, intraspecies variations in IC 50 values showed different susceptibility profiles against antileishmanial drugs (Table 3, Figure 6).     (Nogueira et al., 2016(Nogueira et al., , 2017. Although LPG polymorphisms were evident, they did not trigger differential immune responses in macrophages. Here, to better address this subject, the number of L. amazonensis strains was expanded. The functional properties of their LPGs and susceptibility to current antileishmanial drugs were evaluated. Functionally, L. amazonensis LPGs were able to trigger distinct pro-inflammatory innate immune responses. The Leishmania LPGs from dermotropic species/strains can usually induce higher NO/ cytokines production than viscerotropic ones (Cardoso et al., 2020;Ibraim et al., 2013;Nogueira et al., 2016;Paranaíba et al., 2015;Vieira et al., 2019). When an increased number of strains is used, variations in their pro-inflammatory/immunosuppressive LPG properties were documented (Cardoso et al., 2020;Coelho-Finamore et al., 2011;Vieira et al., 2019). Consistent with these observations, L. amazonensis also followed this pattern. With exception of IL-10 (BA276), higher induction of NO, IL-6, TNF-α, and IL-10 were observed for BA276 (ADCL), LV78 (rodent), and GV02 (CVL) strains. In some cases, this induction was even higher than that caused by LPS. Overall, IL-10 is a pleiotropic immunomodulatory cytokine suppressing Th1-dependent cell-mediated immunity and increasing TH2 immune responses (De Waal Malefyt et al., 1991;Fernandez-Botran et al., 1988). High IL-10 levels in the initial phase of VL lead to susceptibility of infection by decreasing the frequency of multifunctional CD4 T cells (Mesquita et al., 2018). Here, highest IL-10 levels for rodent (LV78) and canine (GV02) were detected. VL pathogenesis appears at least in part due to a shift in the balance of effector/regulatory mechanisms. In this specific case, the higher IL-10 level triggered by L. amazonensis LPG from a canine strain may contribute to an inefficient TH1 response.
On the other hand, BA336 (causing ADCL) LPG, was a poor inducer of IL-12 and MCP-1. IL-12 is a key cytokine promoting cellular activation during Leishmania infection, leading to a TH1-type response (Lohoff et al., 1999). Although this finding was interesting, it could not be correlated to anergy since the LPG from another ADCL strain (BA276) induced higher levels of this cytokine. Based on the proinflammatory properties of the LPGs, a clear correlation between clinical form/host was not noticed. Our next step was to check for LPG polymorphisms affecting macrophage activation.
Consistent with our previous reports, polymorphisms in repeat units were detected and considered to be of three types: Some were devoid of side chains (type I), others had 1-2 β-galactoses (type II) or 1-2 β-glucoses (type III) as side chains (summarized in Table 2). Type II and III structures were already reported for L. amazonensis (Nogueira 1954 (Passero et al., 2015). Type II repeat units were detected in dermotropic strains BA125 (LCL) and BA276 (ADCL). Poly-galactosylated LPGs were already reported in L.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.

ETHICS STATEMENT
Animals were kept in the Animal Facility of the Instituto René