Visceral leishmaniasis (VL) is a potentially fatal parasitic disease associated with fever, cachexia and impaired protective T-cell responses against the parasite.
Peripheral blood leukocytes from 105 subjects with VL and healthy control subjects from the endemic region of Muzaffarpur, Bihar, India, were compared using flow cytometry and reverse-transcriptase quantitative polymerase chain reaction. Findings were correlated with clinical data.
An expanded population of low-density neutrophils that expressed HLA-DR, CD80 and CD86 was observed in subjects with VL. This neutrophil population contracted after successful treatment of disease. Plasma from patients with acute VL was able to induce similar high-level HLA-DR expression in neutrophils from healthy subjects. HLA-DR+ neutrophils from subjects with VL did not stimulate T-cell proliferation, but they did express higher programmed cell death ligand-1 (PDL1) than other neutrophils, and lymphocytes of the same subjects expressed high programmed cell death 1 (PD1).
Patients with acute VL have expanded circulating low-density neutrophils expressing markers of antigen presentation, which diminish after treatment. Development of HLA-DR+ neutrophils is stimulated, at least in part, by components of plasma from patients with acute disease. Although we found no evidence that they act as antigen-presenting cells, these neutrophils expressed markers implicating a role in T-cell exhaustion.
Leishmaniasis refers to a group of syndromes ranging in severity from self-healing cutaneous ulcers to fatal visceral infection. All forms of leishmaniasis are overwhelmingly associated with poverty [1]. The causative Leishmania spp. protozoa are transmitted to humans through the bite of a female hematophagous sand fly (genus Phlebotomus in the Old World, Lutzomyia in the New World). Visceral leishmaniasis (VL) is the most severe form of leishmaniasis, caused by either Leishmania donovani in India and Sudan, or by Leishmania infantum in South America and in countries bordering the Mediterranean. Globally, VL is second only to malaria as a cause of death due to parasitic disease [2].
Leishmania spp. are obligate intracellular parasites in their mammalian hosts, which have been observed in neutrophils (polymorphonuclear leukocytes), monocytes, macrophages, and dendritic cells [3]. Although most parasites reside in macrophages throughout chronic infection, recent studies document the fact that neutrophils are rapidly recruited to the inoculation site, where they internalize parasites. The ultimate impact of neutrophils on the outcome of leishmaniasis remains controversial, in part owing to different methods used for neutrophil depletion and the use of different parasite species and mouse strains in studies [4, 5]. Nonetheless, several studies support the observation that neutrophils help the parasite establish infection [6]. This seems to occur at least partly because of their effects on macrophages or dendritic cells that internalize parasite-laden neutrophils [7]. It remains to be documented, however, whether neutrophils play a role in chronic leishmaniasis.
Neutrophils are not homogenous. Subsets of neutrophils with distinct phenotypic or functional characteristics include the polarized N1 (antitumor) or N2 (protumor) tumor-associated neutrophils, which release proinflammatory and angiogenic factors, respectively [8]. Low-density neutrophils purifying on a density gradient with blood mononuclear cells have been observed in the context of autoimmune diseases, particularly systemic lupus erythematosis, in sporadic reports starting in 1986 [9]. Suppressive granulocytic subsets of myeloid-derived suppressor cells (G-MDSCs) are also reported to migrate in low-density leukocyte fractions [10]. Recently, peripheral blood mononuclear cell (PBMC) fractions from Ethiopian patients with active VL were found to contain CD15+ neutrophils with high arginase levels, and arginine depletion was associated with suppressed T-cell activity [11]. In parallel with these reports, neutrophil subsets expressing HLA-DR and other costimulatory molecules with some functional capacities were reported in the setting of autoimmune diseases [12–15]. The spectrum of clinical settings in which neutrophil subsets are abundant and the origin and functional capacities of neutrophil subsets are ill defined.
Inspired by the potential influence of neutrophils on T-cell function, we examined the hypothesis that there might be subsets of circulating neutrophils with the ability to influence cellular immune responses in subjects with acute VL. We confirmed an enlarged population of low-density neutrophils in subjects with VL, as reported by Abebe et al [11]. We discovered that this population contains another neutrophil subset expressing HLA-DR and other surface markers of antigen-presenting cells (APCs), indicating that the 2 reported neutrophil “subsets” overlap. Studies of these neutrophils suggest that the HLA-DR+ neutrophils arising during VL do not stimulate T-lymphocyte responses, but they may play an alternate role in the pathogenesis of VL.
METHODS
Human Subjects
Patients with VL were recruited from the Kala-Azar Medical Research Center (KAMRC) in Muzaffarpur, Bihar State, India. The diagnosis was based on the presence of typical symptoms and confirmed by detection of amastigotes in splenic aspirates and/or antibodies to recombinant k39 antigen, using a commercially available dipstick test (InBios Kala Azar Detect Rapid Test). Additional inclusion criteria were age >12 years, nonpregnant status, ability to provide informed consent, and response to therapy with amphotericin B. All subjects included in the study were human immunodeficiency virus negative.
Ethical Approvals
Protocols used in this study were approved by institutional review boards at Banaras Hindu University, the University of Iowa, and the National Institutes of Health (NIH). The Banaras Hindu University institutional review board is registered with the NIH. All subjects and/or guardians provided written informed consent, and children aged 12–17 years signed an assent form.
Human Peripheral Blood Cell Preparations
Three methods were used for preparation of peripheral blood cells:
1. Whole blood. First, 100 µL of heparinized whole blood was incubated with desired antibody staining panel for 30 minutes in the dark. Samples were then incubated in 2 mL of RBC Lysis buffer (Becton Dickinson) for 15 minutes, after which they were washed in phosphate-buffered saline (PBS) by centrifugation and resuspended in PBS with 0.1% bovine serum albumin for flow cytometry.
2. Low- and normal-density neutrophils. First, 5 mL of blood was separated by density using Ficoll Hypaque gradients (Ambion; Thermo-Fischer Scientific), as described for isolation of PBMCs [16]. After centrifugation, the low-density layer migrating above the Ficoll layer was collected and suspended in Roswell Park Memorial Institute medium with 10% heat-inactivated fetal calf serum, penicillin (100 U/mL) and streptomycin (50 µg/mL). Normal-density granulocytes (NDGs) below the Ficoll layer and above the erythrocyte “pellet” (buffy coat) were isolated by hypotonic lysis of erythrocytes (15 minutes; RBC Lysis buffer; BD Biosciences) and centrifugation (300g; 10 minutes). Cell preparations were used to stain for flow cytometry.
3. Isolated CD66b+ cells. Erythrocytes in 15 mL of freshly drawn peripheral blood were lysed in 5–10 volumes of lysis buffer (155 mmol/L ammonium chloride, 10 mmol/L potassium bicarbonate, and 0.1 mmol/L ethylenediaminetetraacetic acid) for 10 minutes at room temperature. Cells were washed by centrifugation and resuspended in MACS(R)™ cell separation column buffer (Miltenyi) (0.5% bovine serum albumin and 2 mmol/L ethylenediaminetetraacetic acid in ×1 PBS). Neutrophils were enriched by positive selection using a CD66abce MicroBead Kit with a Magnetic LS column, according to the protocol described by the manufacturer (Miltenyi Biotec).
Flow Cytometry
Peripheral blood cells were stained with fluorescent marker-conjugated antibodies at KAMRC, fixed, and transported to Banaras Hindu University for analysis. Antibodies were as follows (with BD Biosciences catalog numbers) CD66b (555724), CD62L (555544), CD11b/Mac1 (550019), CD86 (555660), CD80 (557227), CD15 (555401). HLA-DR (347364), CD66b (562254), CD3 (555333), CD14 (555397), CD63 (557305). Data were collected on a FACSCalibur model 4CS flow cytometer (BD Biosciences).
RNA Extraction and Gene Expression
Neutrophils purified on a Militenyi CD66abce MicroBead kit were suspended in RNAlater (Qiagen), and total RNA was isolated using the RNeasy Mini Kit and Qiashredder homogenizers (Qiagen). Complementary DNA (cDNA) was synthesized with a High-Capacity cDNA Archive kit (Applied Biosystems), using random primers and MultiScribe MuLV reverse-transcriptase. Quantitative polymerase chain reaction was performed with a ABI Prism 7900HT system, using primers listed in Table 1. Fold changes were calculated using ΔΔCT, with β2microglobulin as an endogenous control. Patient transcript abundance was compared with the mean value from endemic healthy control subjects.
Primers Used for qPCRa
| Transcript | Forward Primer | Reverse Primer |
|---|---|---|
| Arginase-1 | 5′-GTTTCTCAAGCAGACCAGCC-3′ | GCTCAAGTGCAGCAAAGAGA |
| β2-microglobulin | 5′-CTCCGTGGCCTTAGCTGTG-3′ | 5′-TTTGGAGTACGCTGGATAGCCT-3′ |
| HLA-DR | 5′-TCCGTCCCATTGAAGAAATG-3′ | 5′-AGTGACACTGATGGTGCTGAG-3′ |
| IL-10 | 5′-GGTGATGCCCCAAGCTGA-3′ | 5′-TCCCCCAGGGAGTTCACA-3′ |
| Interferon-γ | 5′-CCAACGCAAAGCAATACATGA-3′ | 5′-CGCTTCCCTGTTTTAGCTGC-3′ |
| TGF-β | 5′-GCAGAAGTTGGCATGGTAGC-3′ | 5′-CCCTGGACACCAACTATTGC-3′ |
| Transcript | Forward Primer | Reverse Primer |
|---|---|---|
| Arginase-1 | 5′-GTTTCTCAAGCAGACCAGCC-3′ | GCTCAAGTGCAGCAAAGAGA |
| β2-microglobulin | 5′-CTCCGTGGCCTTAGCTGTG-3′ | 5′-TTTGGAGTACGCTGGATAGCCT-3′ |
| HLA-DR | 5′-TCCGTCCCATTGAAGAAATG-3′ | 5′-AGTGACACTGATGGTGCTGAG-3′ |
| IL-10 | 5′-GGTGATGCCCCAAGCTGA-3′ | 5′-TCCCCCAGGGAGTTCACA-3′ |
| Interferon-γ | 5′-CCAACGCAAAGCAATACATGA-3′ | 5′-CGCTTCCCTGTTTTAGCTGC-3′ |
| TGF-β | 5′-GCAGAAGTTGGCATGGTAGC-3′ | 5′-CCCTGGACACCAACTATTGC-3′ |
Abbreviations: IL-10, interleukin 10; qPCR, quantitative polymerase chain reaction; TGF, transforming growth factor.
a SYBR green was incorporated into qPCR reactions with the listed primers to quantify transcript abundance in complementary DNA samples.
Primers Used for qPCRa
| Transcript | Forward Primer | Reverse Primer |
|---|---|---|
| Arginase-1 | 5′-GTTTCTCAAGCAGACCAGCC-3′ | GCTCAAGTGCAGCAAAGAGA |
| β2-microglobulin | 5′-CTCCGTGGCCTTAGCTGTG-3′ | 5′-TTTGGAGTACGCTGGATAGCCT-3′ |
| HLA-DR | 5′-TCCGTCCCATTGAAGAAATG-3′ | 5′-AGTGACACTGATGGTGCTGAG-3′ |
| IL-10 | 5′-GGTGATGCCCCAAGCTGA-3′ | 5′-TCCCCCAGGGAGTTCACA-3′ |
| Interferon-γ | 5′-CCAACGCAAAGCAATACATGA-3′ | 5′-CGCTTCCCTGTTTTAGCTGC-3′ |
| TGF-β | 5′-GCAGAAGTTGGCATGGTAGC-3′ | 5′-CCCTGGACACCAACTATTGC-3′ |
| Transcript | Forward Primer | Reverse Primer |
|---|---|---|
| Arginase-1 | 5′-GTTTCTCAAGCAGACCAGCC-3′ | GCTCAAGTGCAGCAAAGAGA |
| β2-microglobulin | 5′-CTCCGTGGCCTTAGCTGTG-3′ | 5′-TTTGGAGTACGCTGGATAGCCT-3′ |
| HLA-DR | 5′-TCCGTCCCATTGAAGAAATG-3′ | 5′-AGTGACACTGATGGTGCTGAG-3′ |
| IL-10 | 5′-GGTGATGCCCCAAGCTGA-3′ | 5′-TCCCCCAGGGAGTTCACA-3′ |
| Interferon-γ | 5′-CCAACGCAAAGCAATACATGA-3′ | 5′-CGCTTCCCTGTTTTAGCTGC-3′ |
| TGF-β | 5′-GCAGAAGTTGGCATGGTAGC-3′ | 5′-CCCTGGACACCAACTATTGC-3′ |
Abbreviations: IL-10, interleukin 10; qPCR, quantitative polymerase chain reaction; TGF, transforming growth factor.
a SYBR green was incorporated into qPCR reactions with the listed primers to quantify transcript abundance in complementary DNA samples.
RESULTS
Subjects
VL caused by L. donovani has been endemic in Bihar and adjacent states since the late 1800s [17]. Splenic aspiration with microscopic examination for Leishmania amastigotes is the standard diagnostic procedure for VL in India. In addition to diagnosis, microscopic analysis of splenic aspirates allow a rough scoring of disease severity, ranging from highest (5) to lowest (1) parasite loads. Individuals with a confirmed diagnosis of VL were invited to participate in the study at the time of admission to KAMRC. A total of 105 subjects (36 female and 69 male) were entered between July 2013 and August 2016. Blood samples were drawn on day 0 before treatment but after confirmation of the VL diagnosis. All patients were treated in the hospital with amphotericin B in either its deoxycholate or liposomal form for 30 days [18], with resolution of symptomatic disease. A second blood sample was obtained after completion of treatment, before discharge from the hospital. Healthy endemic control subjects were selected from members of patient households who accompanied patients with VL to KAMRC (n = 31). Blood from healthy controls was collected and processed with patient samples. No serious complications or deaths occurred in the patients included in this study.
Clinical Characteristics
Patients undergo a complete blood cell count on admission to KAMRC. Data in Supplementary Table 1A shows aggregate blood cell counts and clinical chemistry results from 573 subjects patients with active VL admitted to KAMRC. Data exclusively from the first 83 subjects in the current study are listed in Supplementary Table 1B. Both sets of data document leukopenia during acute disease, which resolved after therapy. As expected, liver transaminase levels were not elevated. Subjects did not develop significant renal impairment (measured by serum creatinine) at the end of therapy. In the 573 patients admitted with VL to the medical center, there was no discernable correlation between the duration of fever (mean [standard deviation [SD], 43.8 [1.7] days) and neutrophil counts on day 0 before treatment (R2 = 0.00302; P = .19; Pearson correlation coefficient). There was, however, a weak inverse correlation between disease severity according to splenic score and peripheral neutrophil counts in 573 subjects with VL (mean [SD], 1612 [889] or 1187 [620] neutrophils/mL, with splenic scores of 1 or 5, respectively; P < .05).
Leukocytes in Whole Blood Preparations
Flow cytometry of whole blood was sufficient to distinguish neutrophils from monocytes and lymphocytes. Neutrophils were gated in whole blood samples according to their forward and side scatter properties, as indicators of size and granularity, and according to their staining for the neutrophil marker CD66b. The identity of these gated cells were confirmed by staining for an additional neutrophil marker (CD15), monocyte marker (CD14), and lymphocyte marker (CD3). There was a clear separation between CD66b+CD15+ neutrophils and either CD14+ monocytes or CD3+ lymphocytes (Supplementary Figure 1). Similar separation was shown between neutrophils and CD4+ or CD8+ lymphocytes (not shown).
HLA-DR–Expressing Neutrophils in Whole Blood of Subjects With VL
Detection of neutrophils expressing HLA-DR in whole blood of subjects with visceral leishmaniasis (VL). Peripheral venous blood was collected from subjects with active VL before treatment (VL) or after 28 days of treatment and before discharge (Tx). Blood was collected from healthy endemic control subjects (Cont) during the same period. Whole blood leukocytes were stained for flow cytometry and gated on neutrophils as in Supplementary Figure 1. A, Representative plot of CD66b+ neutrophils stained for HLA-DR. B, Proportions of neutrophils expressing HLA-DR in 84 subjects with active VL, 20 of these same subjects after completion of therapy, and 33 controls. *P < .001, †P < .0001. C, Neutrophils were separated from whole blood leukocytes of subjects with VL before or after completion of therapy, using the Miltenyi CD66abce MicroBead kit. Complementary DNA was used to analyze gene expression by SYBR green with primers in Table 1. Data are shown for HLA-DR expression relative to β2-microglobulin as an endogenous reference and normalized to the mean cycle threshold values for controls. Statistical analyses were done with paired t tests (n = 7 patients with VL).
HLA-DR–expressing neutrophils are enriched in low-density neutrophil fractions in subjects with acute visceral leishmaniasis (VL). Blood was collected from subjects with VL before (VL) or after treatment (Tx) and from endemic healthy controls (ECs). Blood was separated by Ficoll-Hypaque density sedimentation and the low-density fraction migrating above the Ficoll layer, which also contains peripheral blood mononuclear cells, and stained for flow cytometry. A, B, Representative scatterplot (A) and histogram (B) of CD66b+ neutrophils in low-density (LD) leukocyte fraction. C, Proportions of neutrophils found in low-density fractions separated from blood of 56 subjects with VL, 16 with treated VL, and 19 controls. *P < .01; †P < .001. D, Microscopic images of neutrophils enriched from the low-density fraction or the normal-density fraction of patients with VL before treatment, using a CD66abce MicroDead kit (Militenyi) and stained with Diff-Quik stain kit (Wright Giemsa equivalent; Fisher Scientific). Abbreviations: LDGs, low-density granulocytes; NDGs, normal-density granulocytes.
We calculated the absolute number of LDGs in the blood of patients with VL based on clinical laboratory complete blood cell and differential counts, in concert with flow cytometry data. Twenty-two patients had all information available before and after treatment. The mean (SD) absolute LDG number was 1.79 (2.38)–fold higher in patients with VL before therapy than it was after completion of therapy. Although the SD is high because of variability between patients, every subject had fewer LDGs after than before therapy. The difference was statistically significant (P < .003; paired t test). Thus both the proportion and the number of LDGs were significantly higher in patients with VL before as opposed to after treatment for VL.
![HLA-DR–expressing neutrophils are enriched in low-density neutrophil fractions in subjects with acute visceral leishmaniasis (VL). Blood was collected from patients with VL before or after successful treatment. Blood was also collected from healthy endemic control subjects. Leukocytes were studied in whole blood, or separated by Ficoll density sedimentation into low-density and high-density fractions, containing low-density granulocyte (LDG) and normal-density granulocyte (NDG) populations, respectively. Samples were stained with antibodies to human CD66b and HLA-DR and gated for neutrophils. A, Representative plots of HLA-DR against CD66b in the LDG fractions of 2 subjects with the indicated phenotypes. B, Proportions of neutrophils in each fraction (whole blood [WB], LDG, and NDG) that stain positive for HLA-DR in the indicated subjects. Statistical analyses were done with GraphPad Prism 7 software, using 1-way analysis of variance and Tukey posttest (75 subjects with VL and 33 endemic control subjects). *P < .001.](https://oup.silverchair-cdn.com/oup/backfile/Content_public/Journal/jid/214/10/10.1093_infdis_jiw394/3/m_jiw39403.jpeg?Expires=1748032543&Signature=TB8J45Q8YlvtvMHoqx120o38YfFj3EZf2mspC1bVEP91R0hgeFQToox9DncBOJEWSWVsyNcy0hsAxr9Hx~mqOrjlNUawqC1WVwEmFSDz5NDPGbEP90Df-L9qp92bFuKFNMYjIRU90By3hhD-qQE1v60u6yFjiXI-9lxVrTAikHynG8kxbSAIGhvpZvWLX3BmMSp~0oWKKQCYAjYS0pVBu9jEWQ8zaAOn8usaGW0RwrL-njxg8mPj4qWsJ8vbCJd89qPw5dwhjrBTIHj4AuEZ8dfwZ6ER7ZmLqkXQEtHGMKM7AiUbe8H3Rdowjtko-77KkaE--0PGCuZJwMLL2bo~eQ__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
HLA-DR–expressing neutrophils are enriched in low-density neutrophil fractions in subjects with acute visceral leishmaniasis (VL). Blood was collected from patients with VL before or after successful treatment. Blood was also collected from healthy endemic control subjects. Leukocytes were studied in whole blood, or separated by Ficoll density sedimentation into low-density and high-density fractions, containing low-density granulocyte (LDG) and normal-density granulocyte (NDG) populations, respectively. Samples were stained with antibodies to human CD66b and HLA-DR and gated for neutrophils. A, Representative plots of HLA-DR against CD66b in the LDG fractions of 2 subjects with the indicated phenotypes. B, Proportions of neutrophils in each fraction (whole blood [WB], LDG, and NDG) that stain positive for HLA-DR in the indicated subjects. Statistical analyses were done with GraphPad Prism 7 software, using 1-way analysis of variance and Tukey posttest (75 subjects with VL and 33 endemic control subjects). *P < .001.
Expression of Costimulatory Molecules on HLA-DR+ Neutrophils
Costimulatory molecules in HLA-DR+ versus HLA-DR− neutrophils. Peripheral blood from subjects with visceral leishmaniasis (VL) before (VL; n = 11) or at the end of treatment (VL Tx; n = 11) or blood from endemic healthy controls (ECs; n = 15) was studied as whole blood or separated into low- and normal-density fractions. Samples were stained for CD66b, HLA-DR, and either CD80 or CD86. Neutrophils were gated based on scatter properties and CD66b stain as in Supplementary Figure 1, and the amount of CD80 or CD86 stain on HLA-DR+ or HLA-DR− neutrophils was assessed. Data show the percentage of HLA-DR+ or HLA-DR− neutrophils in whole blood (WB) or low-density granulocyte (LDG) fractions that stain positive for CD80 or CD86. Results expressed as geometric mean fluorescence index of the CD80 or CD86 stain were similar (not shown). *P < .001, †P < .0001 (1-way analysis of variance).
Priming and Activation of Neutrophils in Whole Blood of Subjects With VL
Circulating neutrophils from patients with visceral leishmaniasis (VL) show a primed phenotype. A, B, Whole blood samples from subjects with acute VL (n = 38) or healthy controls (n = 17) were stained for markers of priming/activation. CD66b+ neutrophils were selected by gating and analyzed by flow cytometry for their expression of surface CD62L, CD11b, and CD63. Data are presented as mean fluorescence intensity (MFI) for CD62L and CD11b (A), and for the percentage of HLA-DR+ or HLA-DR− neutrophils with high CD11b or low CD62L expression (B). *P < .05. C, Percentage of neutrophils expressing CD63 is compared between whole blood of individuals with acute VL and endemic healthy controls. Whole blood of individuals with VL was separated into normal-density granulocyte (NDG) and low-density granulocyte (LDG) fractions, and all fractions were stained for surface CD63. †P < .01; ‡P < .0001. D, Peripheral blood neutrophils from 12 subjects with acute VL were stained for CD62L and CD33, either as whole blood or after separation of LDGs from NDGs. The percentage of neutrophils in each fraction staining positive for CD33 is indicated. Statistical analyses were done using analysis of variance for 3 fractions or paired t test when comparing VL versus control groups or HLA-DR+ versus HLA-DR− neutrophils. *P < .05.
Effect of Heterologous Plasma From Patients With VL on Neutrophils
Neutrophil interactions with other immune elements in vitro. A, Neutrophil interactions with plasma. Blood samples from healthy subjects were incubated in plasma samples from subjects with acute visceral leishmaniasis (VL; n= 21), subjects with treated VL (Tx; n = 12), heterologous (Hetero) healthy subject plasma (n = 8), or autologous plasma (Auto) (n = 17) at 37°C and 5% carbon dioxide. After 24 hours cells were stained for flow cytometry. Graph shows percentages of gated neutrophils that stained positive for HLA-DR. B, Low-density and high-density leukocyte fractions were separated over a Ficoll-Hypaque gradient from whole blood of subjects with VL or endemic healthy controls. CD15+ low-density neutrophils (LDG) and CD3+ T-cells were separated from low-density fractions on Miltenyi beads, and the remaining low-density cells were enriched in antigen presenting cells (APC). Normal-density neutrophils were separated from low density leukocytes on CD15+ beads (NDG). T-cells were labeled with CFSE and incubated for 48 hours with each of the purified leukocyte fractions at 37°C, 5% CO2 in the presence of soluble Leishmania antigen. The proportion of proliferating (CFSE-low) T-cells in each condition, detected by flow cytometry, is shown on the graph. C and D, Whole blood or purified low-density and normal-density neutrophils (WB, LDG, NDG, respectively) were prepared from blood of subjects with VL or from endemic healthy controls. These leukocytes were stained with antibodies to CD66b, PD1 and PDL1 and analyzed by flow cytometry. Gates for CD66b+ neutrophils and CD66b- lymphocytes were set from forward/side-scatter and CD66b plots. Proportions of lymphocytes from VL or control subjects staining positive for PD1 are shown in panel C, and panel D shows fractions of neutrophils staining positive for PDL1 in blood of VL subjects. Abbreviations: APC, antigen-presenting cell; CFSE, carboxyfluorescein succinimidyl ester; LDG, low-density granulocyte; NDG, normal-high density granulocytes; PD1, programmed cell death protein-1; PDL1, programmed cell death ligand-1; PMNs, polymorphonuclear leukocytes; WB, whole blood. A, *P < .05; †P < .0001. B, *P < .01. C, P < .003; †P < .0001.
LDG-T-Cell Interactions
Findings of prior studies suggest that HLA-DR–expressing neutrophils can present antigen to T cells [21]. In contrast, our studies addressing this possibility, using HLA-DR+ neutrophils from subjects with VL, suggested that neutrophils that arise in this setting may not stimulate T-cell responses (Figure 6B). For this study, we used bead affinity purification to fractionate the low-density leukocyte layer into CD3+ T cells, CD15+ LDGs, and remaining cells containing classic APCs. Normal-density neutrophils (NDGs) were isolated with anti-CD15 beads from below the Ficoll fraction. Carboxyfluorescein succinimidyl ester (CFSE)–labeled T cells were incubated in total soluble Leishmania antigen plus buffer (negative control), NDGs, purified LDGs, or APCs, and proliferation was detected by CFSE dilution into daughter cells, measured by means of flow cytometry. As shown in Figure 6B, the only condition in which T cells proliferated significantly above background levels was in the presence of antigen and classic APCs. The apparent differences between T cells without or with NDGs or LDGs were not statistically significantly.
Complementing the above evidence for a lack of T-cell stimulation, surface proteins mediating T-cell exhaustion were up-regulated in LDGs. Lymphoid cells in either the whole blood or the low-density PBMC fraction in subjects with VL express high levels of programmed cell death ligand-1 (PDL1) (Figure 6C), consistent with our prior observation of high PD1 expression on CD8+ T cells from patients with [22]. Furthermore, there was a significant increase in PDL1 expression on the surface of on LDGs from patients with VL (Figure 6D).
DISCUSSION
VL leads to a dysregulated peripheral inflammatory state with defective cutaneous delayed-type hypersensitivity responses to Leishmania antigen and failure of cellular responses to control parasite growth [23, 24]. The long-standing belief that failure to clear VL is due to a defect in CD4+ T-cell interferon (IFN) γ release has required modification with recent evidence that lymphocytes behave quite differently in whole blood versus mononuclear cell preparations [25–28]. These findings underscore a need to dissect further the cellular responses underlying the inflammatory state during human VL.
Studies in murine models highlight a role for neutrophils in leishmaniasis, despite the fact that these short-lived cells were previously thought incapable of maintaining infection during a chronic protozoal infection [6, 29, 30]. Based on the hypothesis that human neutrophils would also contribute to pathogenesis, we examined neutrophils in human subjects with active VL. We made the surprising observation that a subset of circulating neutrophils express markers of antigen presentation, including major histocompatibility complexclass II and costimulatory molecules CD80 and CD86 in subjects with VL. These cells migrated in the low-density fraction of peripheral blood leukocytes, a fraction that harbors mononuclear leukocytes and a recently described LDG population, which is best studied in patients with malignancy or autoimmune disease [31]. We showed these cells are neutrophils according to surface markers, a characteristic microscopic appearance, and the presence of a neutrophil-specific transcript. Even though subjects with VL had low circulating neutrophil counts, they had more LDGs before than after treatment, and a proportion of these exhibited some degree of activation. Primed/activated and HLA-DR+ neutrophils migrated to the low-density fraction of blood leukocytes.
Low-density neutrophils purifying on a density gradient with blood mononuclear cells were observed as early as in 1986 in patients with systemic lupus erythematosus, rheumatoid arthritis, or acute rheumatic fever. Neutrophil subsets expressing HLA-DR and other costimulatory molecules have also been reported in the synovial fluid or circulation of subjects with rheumatoid arthritis or Wegener granulomatosis, respectively [12, 13], although these were not shown to coincide with LDGs before the current report. Neutrophil specific transcripts have been detected in PBMC microarrays of patients with systemic lupus erythematosis, and their presence was confirmed by means of immunohistochemistry and microscopy. These LDGs are activated and secrete type I IFNs, tumor necrosis factor α, and IFN-γ but have diminished capability to phagocytose [9]. LDGs in patients with septic shock reportedly express arginase and the zeta chain of the T-cell receptor (CD247), and the numbers of these neutrophils in the PBMC fractions correlates with the severity of disease [31]. Finally, PBMC fractions from Ethiopian patients with active VL were found to contain CD15+ neutrophils with high levels of arginase. In these same subjects depletion of arginine was suggested to be associated with suppressed T-cell activity [11, 32].
It can be debated whether LDGs are the same as the so-called G-MDSCs, suppressive myeloid cells with properties of immature granulocytic lineage that also migrate in the low-density leukocyte fraction [31, 33]. Such cells can be induced by transforming growth factor-beta, which is amply present in the circulation of individuals with VL [26, 33]. Nonetheless, we have no evidence that the HLA-DR+ neutrophils observed in acute VL have an immature cell type, and they do not up-regulate CD33, as is described for G-MDSCs. We do show evidence for a potential suppressive effect through expression of PDL1 in a host with high levels of PD1+ lymphocytes. Rather than finding these the same as G-MDSCs, we believe that these cells add another level of complexity to the possible neutrophil subsets observed in a disease state.
Factors that induce or enhance the development of functionally distinct neutrophil subsets are not well defined. One report suggests that unconventional T cells induce human neutrophil differentiation into cells capable of antigen cross-presentation [21], and others suggest that HLA-DR and costimulatory molecules are induced by cytokines, including IFN-γ and/or granulocyte-macrophage colony-stimulating factor [34, 35]. Finally, our own studies of subjects with cutaneous leishmaniasis in northeast Brazil have documented the presence of HLA-DR–expressing LDGs in this state (R. E. D. et al, unpublished data).
The functional role of neutrophil subsets, and low-density HLA-DR+ neutrophils in particular, is not known. Studies of other conditions by other investigators suggest that these cells will be capable of presenting antigen to T cells in some capacity. Our data suggest that HLA-DR–expressing neutrophils do not act as APCs. We verified the data of others that lymphocytes from subjects with VL express high levels of PD1 and the very new finding that high levels of PDL1 are expressed on HLA-DR+ low-density, but not on normal-density neutrophils. This leads us to the hypothesize that HLA-DR–expressing neutrophils that become prominent during VL may be involved in promoting T-cell exhaustion and not in presenting antigen.
Notes
Financial support. This work was supported in part by a Tropical Medicine Research Center grant from the NIH (P50 AI-074321), the NIH (grants R01 AI076233 and R01 AI045540 to M. E. W.), the Department of Veterans Affairs (grants 5I01BX001983 and 2I01BX000536), and the Council of Scientific and Industrial Research.
Potential conflicts of interest. All authors: No reported conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References
Author notes
Presented in part: Annual Meeting of the American Society of Tropical Medicine and Hygiene, Washington DC, 13–17 November 2013.
S. S. and M. E. W. contributed equally to the work.
