Dogs naturally infected with L. infantum, tested seropositive for anti-Leishmania antibodies by the rapid dual-path platform assay (TR DPP, Biomanguinhos, Fiocruz, Rio de Janeiro, Brazil) and by enzyme immunoassay (ELISA, Biomanguinhos, Fiocruz, Rio de Janeiro, Brazil) were included in the study and scheduled for euthanasia between 2013 and 2014 according to the guidelines of the Brazilian Ministry of Health. All dog owners provided formal written consent for their participation. Sample collection occurred during necropsies performed by veterinarians at the Laboratório de Pesquisa Clínica em Dermatozoonoses em Animais Domésticos-INI-Fiocruz (LAPCLIN-DERMZOO-INI/Fiocruz). This study received approval from the Comitê de Ética em Uso de Animais (CEUA-Fiocruz) under license LW-54/13 and adhered to Brazilian Law 11794/08 and the guidelines of the Sociedade Brasileira de Ciência em Animais de Laboratório (SBCAL). The access to genetic heritage was registered under number AFB4BD9 on the SisGen platform.

Thirty-four dogs from Barra Mansa, Rio de Janeiro, Brazil, diagnosed with L. infantum infection and referred for mandatory euthanasia at the Evandro Chagas National Institute of Infectious Diseases (INI/FIOCRUZ) were included. Euthanasia was carried out by veterinarians according to the following protocol: a) Intravenous administration of 1.0% thiopental (Tiopentax^®, Cristalia) at a dosage of 1.0mL/kg; and b) After confirming the absence of a corneal reflex induced by deep anesthesia, 10mL of 19.1% potassium chloride (Isofarma) was administered intravenously. Spleen and blood samples were collected during necropsy; clinical data were recorded beforehand by two veterinarians.

Clinical signs of CVL were evaluated per established protocols (21), assessing dermatitis, onychogryphosis, ophthalmic changes, body condition loss, alopecia, and lymph adenomegaly. Each sign was graded from 0 (absent) to 3 (severe) (33), and the total clinical score was used to classify animals as having low (0–2), moderate (3–6), or high (7–18) clinical severity.

Spleen tissue fragments were formalin-fixed in a 10% formalin-buffered solution (Merck, Darmstadt, Germany) and embedded in paraffin (Synth, Diadema, Brazil). Histological sections (5 μm thick) were mounted on microscopic slides and stained with hematoxylin and eosin (Leica Biosystems, Newcastle Upon Tyne, UK) for subsequent analysis using an optical microscope (Zeiss, Oberkochen, Germany). The organization of the splenic lymphoid tissue, including the white pulp, marginal zone, and red pulp, was assessed as previously described (14). The white pulp was graded as follows: 1) Organized, characterized by a periarteriolar sheath, germinal centers, and distinct mantle and marginal zones. 2) Slightly disorganized, exhibiting some hyperplastic or hypoplastic changes resulting in a partial loss of definition of certain white pulp regions. 3) Moderately disorganized, with white pulp regions barely individualized or indistinct. 4) Intensely disorganized, where the follicular structure is scarcely distinguishable from the red pulp and the T cell area. Then, the quantification of lymphoid follicles per square millimeter of tissue was conducted. This evaluation considered both the presence or absence of follicles and the number of lymphoid follicles per field.

Total DNA was extracted from approximately 10 mg of spleen using the QIAmp DNA Mini Kit (Qiagen, Santa Clarita, USA), so involving an initial digestion step with proteinase K (20 mg/mL) for 1 hour at 56°C. The DNA was then eluted in TE buffer and quantified using a NanoDrop^® spectrophotometer (Thermo Fisher Scientific, Waltham, USA).

Parasite load in the spleen was quantified using real-time PCR following established protocols (17). HPRT primers (see Supplementary Table S1 ) were used to normalize the canine DNA content in each sample. Quantification of Leishmania DNA was performed using primers targeting the small subunit ribosomal RNA (ssrRNA) gene, a multi-copy gene commonly used for parasite detection (34) (see Supplementary Table S1 ).

qPCR assays were conducted using a Step One instrument (Applied Biosystems, Molecular Probes, Inc., Foster City, USA) with Power SYBR Green Master Mix (Applied Biosystems, Molecular Probes, Inc., Foster City, USA). Two microliters of purified total DNA (100 ng) were added to a final PCR volume of 20 μl, containing Power SYBR Green 1X and 300 nM of each primer for HPRT PCR assays or 500 nM of each primer for ssrRNA PCR assays. The qPCR protocol involved an activation step at 95°C for 10 minutes, followed by 40 cycles of denaturation and annealing/extension (95°C for 15 seconds, 60°C for 1 minute, and 68°C for 30 seconds).

A standard curve was generated for each target. All reactions were performed in duplicate for each target, and both targets were run on the same plate for the same sample. For quantification, peripheral blood mononuclear cells (PBMCs) from uninfected dogs and bulk cultures of L. infantum promastigotes (strain MCAN/BR/2007/CG-1) were used. PBMCs and promastigotes were quantified using a cell counter (Z1™ COULTER COUNTER^®, Beckman Coulter, Fullerton, USA), and total DNA was extracted from 1x10⁶ PBMCs and 1x10⁷ promastigotes. Standard curves for the HPRT and ssrRNA genes were prepared using 10-fold serial dilutions. The animals were then categorized into high or low-parasite load groups based on previously established criteria (17).

Spleen tissue fragments were embedded in Tissue-Tek OCT resin (Sakura, Alphen aan den Rijn, The Netherlands) and sectioned at a thickness of 5 μm. Sections were mounted on silanized slides (Dako, Carpinteria, CA, USA) using acetone P.A. (Merck), then hydrated in phosphate-buffered saline (PBS) for 10 minutes. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide (Dako) for 1 minute at room temperature. After two 5-minute washes in PBS, non-specific binding was blocked with 0.4% bovine serum albumin (BSA) for 20 minutes at room temperature.

Excess blocking solution was removed, and primary antibodies targeting CD68 (macrophages), CD206 (mannose receptor), NOS2 (nitric oxide synthase 2), TGF-β (transforming growth factor beta), and Leishmania amastigotes were applied. Sections were incubated with primary antibodies for 18 hours at 4°C. Control sections were prepared by omitting the primary antibody.

After incubation, sections were washed twice in PBS (5 minutes each) and incubated with the appropriate biotinylated secondary antibody (Dako) for 25 minutes, followed by two more PBS washes. Streptavidin-peroxidase (Genetex) was then applied for 25 minutes. After additional PBS washes, the immunoreactivity was visualized using an AEC substrate kit (Invitrogen, Carlsbad, USA). The reaction was stopped with type II water upon microscopic visualization. Slides were counterstained with Meyer’s hematoxylin (Sigma-Aldrich, Saint Louis, USA) and mounted in Faramount medium (Dako).

Quantification of NOS2, CD206, and TGF-β positive cells was performed in five fields per spleen section at 40× magnification. Amastigote quantification was performed in 20 fields at 1000× magnification. Results are expressed as the median percentage and range (minimum to maximum) of positive cells or positive cells per square millimeter. Qualitative analyses were performed in both the red and white pulp regions.

Immunofluorescence was conducted as previously described (21). Briefly, splenic tissue sections were fixed in acetone P.A. (Merck) and blocked to reduce non-specific antibody binding. The sections were then incubated with primary antibodies targeting macrophages (Bio-Rad, USA), pSTAT-3, arginase-1 (Cell Signaling Technology, USA), NOS2, PD-L1, CD68 (Abcam, USA), and L. infantum antigen. Secondary anti-mouse IgG antibodies conjugated with phycoerythrin (PE-red) and fluorescein isothiocyanate (FITC) were used, along with anti-mouse IgG conjugated with Dylight 633 and anti-mouse IgG conjugated with Dylight 488 or Alexa fluor 488. Slides were mounted with Fluoromount-G medium containing DAPI (4′, 6-diamino-2-phenylindole) (Thermo Fisher Scientific) and examined using a fluorescence microscope. Image processing and overlay analyses were performed using ImageJ software (NIH, USA). The brightness and contrast were adjusted.

Splenocytes were isolated from 1 cm splenic fragments through mechanical disruption and successive washings, then cryopreserved in a solution of 90% fetal bovine serum (FBS) and 10% dimethyl sulfoxide (DMSO; Sigma) and stored in liquid nitrogen until further analysis. Thawing was performed at 37°C, followed by the gradual addition of RPMI medium supplemented with 20% FBS (1 mL to 4 mL increments) at one-minute intervals. Cells were subsequently centrifuged at 1800 rpm for 10 minutes at 7°C to remove the supernatant and resuspended in 5 mL of RPMI with 20% FBS. Following homogenization, an additional 5 mL of RPMI with 20% FBS was added. The samples were centrifugated for 10 minutes at 1800 rpm at 7°C. Cells were incubated on ice for 20 minutes in a blocking solution containing 10% goat serum and 10% FBS to block Fc receptors. After centrifugation, cell viability, and concentration were assessed using a Neubauer chamber and Trypan blue exclusion method and adjusted to 1×10⁷ cells/mL. Only samples exhibiting a viability greater than 60% proceeded to the labeling protocol.

Cell viability was assessed using the Zombie Aqua dye (Invitrogen) for flow cytometry. A total of 1×10⁶ cells per tube were washed with 200 µL PBS and incubated with 100 µL of Zombie dye for 20 minutes at room temperature. After washing with PBS, surface staining was conducted. The primary anti-PD-L1 antibody (Biolegend, USA) was incubated for 20 minutes at 4°C. After washing with PBS/BSA (PBS containing 0.1% BSA, 0.05% sodium azide), the secondary anti-mouse PE antibody (Invitrogen) was added and incubated for 20 minutes at 4°C. After another wash with PBS/BSA, the anti-CD14-APC antibody (BD Biosciences, USA) was added and incubated for 20 minutes at 4°C. Cells were then fixed and permeabilized using Cytofix/Cytoperm solution (BD), followed by washing with a permeabilization buffer (PBS/BSA containing 0.5% saponin). Intracellular staining was performed by resuspending cells in 30 µL of the permeabilization buffer and incubating with anti-arginase-1 PE-Cyanine 7 antibody (Invitrogen) for 30 minutes at 4°C. After a final wash, samples were washed and resuspended in PBS for acquisition in the flow cytometer.

Flow cytometric analysis was performed on a 12-color CytoFLEX flow cytometer equipped with 405 nm, 488 nm, 565 nm, and 638 nm lasers (Beckman Coulter, Brea, CA, USA) at the Multi-user Research Facility of Flow Cytometry – Multiparametric Analysis, Instituto Oswaldo Cruz, FIOCRUZ. Data acquisition was managed using CytExpert software (Beckman Coulter). Fluorochromes were analyzed as follows: PE and PE-Cy7 (excited by 565 nm laser) with filters 585/42 and 780/60, respectively; APC (excited by 638 nm laser) with filter 660/10; and Zombie Aqua (excited by 405 nm laser) with filter 525/40.

Data analysis was conducted using FlowJo software version 7.6.5 (Becton Dickinson, Franklin Lakes, NJ, USA). Initial gating was based on forward scatter (FSC) versus side scatter (SSC) to delineate the cell population, followed by the exclusion of doublets via FSC-A vs. FSC-H gating ( Supplementary Figure 1 ). Viability was determined using a histogram of Zombie Aqua staining, and only samples greater than 50% viable cells were included. From the live cell population (Zombie-negative, live gate), expression levels of CD14, arginase-1, and PD-L1 were quantified. Single, dual, and triple expression profiles were assessed using the combination gating tool in FlowJo. The combination gating tool is an automatically combination of gates defined in simple analyses which permits all possible combinations (positive/negative), creating ‘AND’, ‘OR’ and ‘NOT’ populations for each antibody, resulting in different cell phenotypes’.

A non-probabilistic sampling approach was employed in this study, with animals classified into three groups based on clinical presentation: (1) low clinical score, (2) moderate clinical score, and (3) high clinical score. Furthermore, animals were stratified according to the degree of organization in splenic lymphoid tissue: (1) organized (ranging from well-organized to mildly disorganized splenic white pulp) and (2) disorganized (ranging from moderately to severely disrupted splenic white pulp). Additionally, animals were categorized based on parasite burden: (1) low and (2) high.

The Shapiro–Wilk test was applied to assess the normality of data distribution. For nonparametric data, comparisons between independent groups were conducted using the Mann–Whitney U test or Kruskal–Wallis test followed by Dunn’s post hoc test. For parametric data, analyses were performed using the unpaired Student’s t-test or one-way ANOVA followed by Tukey’s multiple comparisons test. Correlations were assessed using either Pearson’s or Spearman’s rank correlation coefficient, as appropriate. Statistical analyses were conducted using GraphPad Prism version 8.0 (GraphPad Software, San Diego, CA, USA). A p-value less than 0.05 was considered statistically significant. Results are expressed as median and range (minimum–maximum).

Thirty-four dogs naturally infected with L. infantum were enrolled in this study. The groups were homogeneously distributed according to age and sex, as presented in the Table 1 .

The most frequently observed clinical manifestations of Canine Visceral Leishmaniasis (CVL) – including dermatitis, alopecia, onycogryphosis, keratoconjunctivitis, lymphadenomegaly, and poor body condition – were evaluated.

Based on the severity of these clinical signs, the animals were categorized into three groups: low (n = 10), medium (n = 13), and high clinical score (n = 11), as previously described (21). Additionally, splenic parasite burden was quantified, and animals were categorized into low (n = 18) and high (n = 16) parasite load groups.

Disorganization of the splenic white pulp was assessed and defined by reducing the number and compartmentalization of lymphoid follicles and the periarteriolar lymphoid sheath ( Figure 1 ). A significant proportion of the animals (30/34; 75%) exhibited varying degrees of white pulp disorganization, ranging from mild (n = 9), moderate (n = 10), to severe (n = 11). This histopathological alteration appears to be a common feature and potentially associated with the progression of CVL (21).

Immunohistochemical analysis was performed on splenic sections to detect and quantify the expression of NOS2, TGF-β, and CD206 (mannose receptor), along with the presence of Leishmania amastigotes ( Figure 2 , Supplementary Table S2 ). Positive cells for NOS2, CD206, and TGF-β were observed in all animals, indicating the involvement of distinct macrophage profiles in the response to L. infantum infection. Overall, NOS2⁺ cells were the most prevalent, compared to CD206⁺ and TGF-β⁺ markers ( Supplementary Table S2 ).

We observed two types of CD206 labeling: cellular labeling predominantly in macrophages, but also diffuse labeling in the extracellular medium. Notably, a significantly greater number of CD206⁺ cells was detected in spleens with moderate to severe white pulp disorganization compared to those with preserved or slightly altered architecture (p = 0.048; unpaired t-test) ( Figure 2E ). No statistically significant differences were found in CD206⁺ cell counts when stratified by clinical score or parasite load. However, there was a correlation between CD206/mm² and amastigotes/mm² (p < 0.000; r = 0.795) ( Figure 2H ).

Macrophage polarization was further assessed via immunofluorescence. During this analysis, macrophages (CD68⁺ cells) co-expressing phosphorylated STAT-3, arginase-1, or NOS2 were identified ( Figure 3 ). No significant differences were detected in the frequencies of CD68⁺pSTAT-3⁺, CD68⁺arginase-1⁺, or CD68⁺NOS2⁺ cells across groups categorized by splenic white pulp organization, clinical score, or parasite burden ( Supplementary Table S2 ). However, a positive correlation was observed between CD68⁺pSTAT-3⁺ and CD68⁺arginase-1⁺ cell frequencies (p = 0.046; r = 0.361; Spearman’s correlation).

By using immunofluorescence analysis of spleen sections at 100× magnification, the co-localization of arginase-1⁺ macrophages and Leishmania infantum amastigotes was confirmed ( Figure 4 ). This observation indicates that arginase-1⁺ macrophages are phagocytic and capable of harboring intracellular amastigotes ( Figure 4C ). Notably, amastigotes were also detected in arginase-1⁻ cells.

No statistically significant differences in amastigote density (amastigotes/mm²) were observed across groups stratified by clinical score or splenic white pulp organization ( Figures 4D, E ). However, animals exhibiting a high parasite load demonstrated a significantly reduced CD68⁺NOS2⁺/CD68⁺Arginase-1⁺ cell ratio ( Table 2 , Figure 4F ).

Immunofluorescence analysis of spleen sections at 40× magnification revealed in situ expression of the exhaustion marker PD-L1 on macrophages (CD68⁺) ( Figure 5 ). A trend towards elevated PD-L1 expression was observed in animals exhibiting disorganized splenic white pulp (p = 0.053; Mann–Whitney test) ( Figure 5F ).

To assess monocyte recruitment in the spleen, CD14⁺ cells were quantified via flow cytometry, alongside the co-expression of arginase-1 and PD-L1 ( Supplementary Table S3 , Figure 6 ). A decrease in the frequency of CD14⁺ cells (p = 0.055; one-way ANOVA) and CD14⁺arginase-1⁻PD-L1⁻ cells (p = 0.048; Kruskal–Wallis test) was observed in dogs with a high clinical score ( Figures 6A, B ). Furthermore, significant correlations were found between CD14⁺ and arginase-1⁺ cells (p = 0.001; r = 0.623; Spearman correlation), as well as between CD14⁺arginase-1⁺PD-L1⁻ and CD14⁻arginase-1⁺PD-L1⁺ cell populations (p = 0.034; r = 0.443; Spearman correlation) ( Figures 6C, D ). In Panel C, the correlation occurs between CD14⁺ cells and total cells expressing arginase-1. Note that the percentage values (X-axis of Panel C) are much higher than those in Panel D because it includes not only M2 cells but also various leukocytes expressing the enzyme, including lymphocytes and neutrophils. In Panel D, the correlation occurs between CD14⁺Arginase-1⁺PD-L1^- cells and CD14^-Arginase-1⁺PD-L1⁺ cells. Notice that the percentages are much lower, as it represents only a fraction of the population demonstrated in Panel C.

Flow cytometric analysis revealed the presence of arginase-1^high and arginase-1⁺PD-L1⁺ co-expressing cells in the spleens of naturally infected dogs ( Supplementary Table S3 ). Although some of them were CD14⁺ cells, it was not possible to define the other arginase-1^high cells due to the absence of additional profile markers in our experiments. No statistically significant differences were observed according to clinical score, parasite load and splenic white pulp organization groups ( Figure 7 ). However, a correlation between CD14^-arginase-1^-PD-L1⁺ and arginase-1^high cells was observed (p = 0.002; r = 0.605, Spearman correlation) ( Figure 7D ).