Introduction

Visceral leishmaniasis (VL) is a prevalent disease in many tropical and subtropical areas of the world, with more than 90% of cases reported from India, Nepal, Bangladesh, Sudan and Brazil. Leishmania infantum is the main etiologic agent of VL in Latin America and Europe [1]. L. infantum infection outcomes vary from asymptomatic to severe disease, but asymptomatic infection occurs in most people. In endemic regions of Brazil, an improvement in nutrition and vaccination rate coverage for children and better sanitary conditions resulted in an increased trend to asymptomatic infection rather than symptomatic VL; in addition, there has been a decrease of disease among children [2–4]. The factors determining the conversion from asymptomatic infection to VL are still not entirely understood.

Studies indicate that high antibody titers are more often seen in subjects with symptomatic VL [2,3]. Subsets of T cells from asymptomatic L. infantum infected subjects are capable of activation after L. infantum antigens stimulation, detected by the marker CD69⁺. The majority of asymptomatic subjects eventually mount a protective Th1-type cellular immune responses with IL-2 and IFN-γ production, that is long-lasting if there is no immunosuppression, possibly indicating protective immunity [4]. In contrast, symptomatic VL subjects have unique T cell responses characterized by exhaustion and activation profiles, with decreased IFN-γ, TNF and IL-6; and increase IL-10 production [5]. A cytokine storm was observed in VL patients, similar to observed in sepsis, with elevated levels of anti-and pro-inflammatory cytokines, and this phenomenon was associated with elevated levels of LPS, sCD14 and I-FABP, and indicative of microbial translocation [6]. A similar immune profile is seen in dogs with VL and usually these animals do not have self-resolving infection and have increased risk of death [7].

The resurgence of VL in Europe occurred concurrently with the HIV epidemic, first spread through sharing of needles, but also through the expansion of canine VL in several countries including Portugal, Spain, France and Italy. In Brazil, HIV was initially an infection of urban settings and VL was a disease predominantly in rural areas. However, in the last 30 years VL has become frequent in peri-metropolitan areas of major cities in the Northeast and Southeast regions of the country, and HIV has spread to rural areas. The increased regional overlap of both infections has led to changes in the epidemiology of VL due to co-infection [1,8]. In the state of Rio Grande do Norte, Brazil, this resulted in an increase in AIDS/VL cases detected after 2010 [8,9].

A high risk of VL development was observed among people living with HIV (PLHIV) [10], and L. infantum is considered an opportunistic pathogen in this population [11]. The impact of VL in people who are immunocompromised by HIV can be devastating, since both diseases suppress the immune system [12]. Asymptomatic co-infection with HIV and L. infantum increases the risk of VL development [13], and in Brazil co-infected patients have had more VL relapses and higher rates of death than non-HIV VL [1]. Relapses of VL have been associated with falling CD4⁺ T cell counts [10].

Leishmania infection increases HIV viral load in serum from co-infected people [14]. Consistently, macrophages co-infected with HIV in vitro have higher L. infantum loads and increased capacity to uptake the mammalian stage, the amastigote [15,16]. Chronic HIV infection is known to lead to CD4⁺ T cell depletion and CD8⁺ T cell activation [12]. Cellular immune responses are essential for control of Leishmania spp. infection, with contributions of CD4⁺ T cells to the outcome of active disease and CD8⁺ cells contributing to immune protection. It is possible that the low CD4⁺ T cells characterizing HIV infection synergize with the suppressed hematopoietic cell functions due to Leishmania spp. infection [17]. Furthermore, AIDS/VL co-infected subjects exhibited increased CD8⁺ T cell activation during active VL and after remission compared to people with HIV alone [18]; the chronic T cell activation phenotype was positively correlated with T cell senescence, leading to accelerated immune cell impairment [17].

Our overreaching hypothesis was that asymptomatic individuals co-infected HIV/Leish would be treated as HIV without positive Leishmania serology individuals, however, would present an altered immunological profile, similar to individuals with AIDS/VL and, therefore, would have a higher risk of developing AIDS/VL. The immunologic profile of PLHIV with positive serology for L. infantum has not previously been described. We analyzed CD4⁺ and CD8 ⁺T cell activation and senescence in HIV, asymptomatic HIV-L. infantum co-infection, AIDS/VL co-infection, as well as, immunocompetent groups, active symptomatic VL, Recovered VL, DTH+ and healthy subjects. Our results described for the first time the immunological profile of PLHIV with positive serology. Moreover, we observed an increased risk of development of AIDS/VL by asymptomatic HIV/L. infantum co-infected subjects (PLHIV with latent L. infantum infection) compared to HIV (PLHIV with negative L. infantum serology) after three years since the first examination. Of importance, this HIV population was studied prior to the era to the initiation of ART in Brazil. The presence of HIV infection was associated with PD1⁺CD8⁺T cells, and HIV viral load was negatively correlated with CD4⁺ counts. Moreover, L. infantum infection influenced activated CD4⁺T, PD1⁺CD8⁺T and senescent CD8⁺T cells. Suboptimal use of ART and previous history of opportunistic infection were associated with an increased risk of AIDS/VL outcome. Thus, PLHIV residing in areas endemic for visceral leishmaniasis area should have measurement of L. infantum serology and close follow-up with an emphasis on the importance of consistent use of ART.

Methods

Study design and population

1. A cross-sectional study of PLHIV to assess co-infection with Leishmania infantum. The study was performed between July 2014 and January 2016, in which 1,372 people living with HIV residing in Rio Grande do Norte State, Brazil, were tested for L. infantum infection by SLA (soluble leishmania antigens) prepared from a local isolate and rK39 antigens using ELISA [18]. Of importance, at the time of recruitment, in 2014, antiretroviral therapy was recommended for those with CD4+ cell counts below 350 cells/mm³ [19].
2. Retrospective analysis of AIDS-VL. A retrospective analysis was performed in 2019 considering the data on symptomatic VL cases which were obtained from the Notifiable Diseases Information System (SINAN) covering the period from July 2014 to December 2018 in order to ascertain the diagnosed AIDS/VL cases for the study period. AIDS/VL reported cases in SINAN database were cross referenced to 1,372 HIV cases evaluated in the cross-sectional study of PLHIV. The hypothesis herein was that since the HIV population recruited between 2014-2016 had not started promptly on antiretroviral therapy, they were more likely to develop AIDS-VL, once they became infected with L. infantum.
3. Immunological studies to assess T cell function. A total of 69 subjects were recruited and their group allocations are shown in Table 1. HIV (n = 16) and Asympt HIV/Leish (n = 10) subjects were drawn from the cross-sectional HIV study, and these subjects had a second blood sample collected for further serological evaluation and for immunophenotyping of T cells for exhaustion, senescence and activation. People with AIDS/VL and symptomatic VL were recruited at Giselda Trigueiro Hospital, Natal, Brazil, during their treatment for VL; whereas Recovered VL and DTH subjects were recruited from a cohort study of families with a history of a symptomatic case. This latter population has been described previously [20,21]. Uninfected Controls were in the same endemic area from which VL and recovered VL were recruited.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Subject groups undergoing immunologic testing.

https://doi.org/10.1371/journal.pntd.0012848.t001

Results

Immunologic and virologic assessments comparing HIV and HIV-asymptomatic L. infantum infected group

The immunologic and virologic status of a subgroup of asymptomatic HIV/Leish subjects was compared to HIV and AIDS/VL subjects. Results are shown in Table 2, as median (min-max). Four of the seven groups had a male predominance while the remainder were more evenly balanced. Most subjects in each group were male (VL (87.5%), AIDS/VL (85%) and asymptomatic HIV/Leish individuals (70%)). There was no difference in the age between groups (Kruskal-Wallis p = 0.56) (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Clinical and laboratory characteristics of patients included in the immunologic study.

https://doi.org/10.1371/journal.pntd.0012848.t002

All Asympt HIV/Leish individuals (10 of 10) had positive anti-IgG SLA serology, and 5 of 10 individual had positive anti-IgG rK39 serology. Asympt HIV/Leish individuals had higher anti-SLA IgG antibodies and similar levels of anti-rK39 IgG antibodies compared to AIDS/VL (Tables 2 and S1). VL patients had higher levels of anti- L. infantum antibodies than AIDS/VL group (SLA and rK39) or DTH+ subjects (SLA). Most AIDS/VL subjects had negative serology to L. infantum antigens (SLA: 71.4%/ rK39: 78.5%). DTH+ subjects had similar rOD antibodies levels as Asympt HIV/Leish individuals or recovered VL subjects. SLA rOD correlated with rK39 rDO (r = 0.88; p<0.0001). HIV groups differed with respect to their HIV viral loads and CD4⁺ T cell counts, but not their CD8⁺ T cell counts (Tables 2 and S2). HIV and Asympt HIV/Leish individuals had lower viral loads than individuals with AIDS/VL. At the same time, CD4⁺ T counts were higher in HIV and in Asympt HIV/Leish individuals than in AIDS/VL. Viral load, and CD4⁺ and CD8⁺ T cell counts did not differ between HIV and Asympt HIV/Leish individuals. HIV viral load was negatively correlated with CD4⁺ count (r = 0.53; p = 0.005), while CD4⁺ count was positively correlated with CD8⁺ count (r = 0.45; p = 0.01). No differences in CD8⁺ T cell counts were observed between the HIV, Asympt HIV/Leish and AIDS/VL groups (ANOVA Kruskal-Wallis p = 0.25) (S2 Table).

ART was used by 81.25% (13/16) of the HIV subjects. However, in L. infantum infected groups, the adherence to ART was lower. Sixty percent (6/10) of asympt HIV/Leish subjects and 7.15% (1/14) of AIDS/VL subjects used ART regularly. Subjects were asked about their history of prior infections with Toxoplasma spp., Mycobacterium tuberculosis or Mycobacterium leprae. Before enrollment, 25% (4/16) of HIV subjects had had one of these infections compared with 40% (4/10) of Asympt HIV/Leish subjects and 84% (11/14) of AIDS/VL subjects. Chi-square analysis revealed that both the incorrect use of ART therapy (Fisher exact test p = 0.0008) and a history of previous infection with Toxoplasma, M tuberculosis, or M. leprae (Fisher exact test p = 0.0007) were associated with an increased risk of AIDS/VL outcome.

Asymptomatic HIV/Leishmania infantum infection increased activation and senescence immunological status in CD4+ and CD8⁺T cells

T cell activation and senescence were examined in groups as listed in Table 1. The overall results are shown in S3–S9 Tables and Figs 1 and 2. CD38 and HLA-DR are markers of cell activation and high levels of these markers were observed in subjects with VL, AIDS/VL and Asympt HIV/Leish (Fig 1 and S3 Table). Subjects with AIDS/VL had more CD3+CD4+CD38+HLA-DR+ activated T cells when compared to recovered VL (p = 0.0004), HIV (p = 0.0001), and DTH+ (p<0.0001) (S4 Table). AIDS/VL and VL subjects had similar higher expression of activation markers, in both CD4⁺ and in CD8⁺ T cells. Recovered VL had the proportion of CD8⁺CD38⁺HLA+-DR⁺ activated T cells less than VL subjects. There were no differences between T-cell activation in UC subjects compared to HIV subjects, recovered VL or DTH+ subjects, in CD4⁺ and in CD8⁺ T cells subtypes. VL patients had significantly higher proportions of CD38+HLA-DR+ T cells than UC. AIDS/VL had higher proportions of activated T cells than HIV or UC (Fig 1 and S3 Table). Asympt HIV/Leish subjects had higher percentage of activated CD8+ T cells than UC subjects (Fig 1 and Table S3). Asympt HIV/Leish subjects also showed higher proportions of activated CD4⁺ T cells than HIV or DTH+ (Fig 1 and S3 and S4 Tables). Asympt HIV/Leish and AIDS/VL subjects had similar proportions of T cells expressing activation markers in CD4⁺ or CD8⁺ T cells. However, proportions of activated CD8⁺ T cells were higher in Asympt HIV/Leish subjects compared to HIV or DTH+ (Fig 1 and S1 and S5 Tables).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Immune activation CD38+ HLA-DR+ status in CD3+CD4+ (A) and in CD3+CD8+ T cells (B).

Each point in the graphs represents one individual and horizontal bars represent median. ANOVA Kruskal Wallis test, with Dunn´s post-test was performed with p<0.05 of significance. UC = Uninfected Controls; HIV = PLHIV negative serology to Leishmania infantum antigens; Asympt HIV/Leish = PLHIV with positive serology to Leishmania infantum antigens; AIDS/VL = subjects in AIDS and active visceral leishmaniasis; VL = active visceral leishmaniasis patients without HIV; Recovered VL = subjects without HIV after cure of visceral leishmaniasis; DTH+ = subjects without HIV with positive Montenegro skin test.

https://doi.org/10.1371/journal.pntd.0012848.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Proportions of cells positive for markers of senescence (CD57+) and exhaustion (PD1+) in CD3+CD4+ (A and C) or in CD3+CD8+ T cells (B and D).

Each point in the graphs represents one individual and horizontal bars represent median. Kruskal Wallis test, with Dunn´s post-test was performed with p<0.05 of significance. UC = Uninfected Controls; HIV = PLHIV with negative serology to Leishmania infantum antigens; Asympt HIV/Leish = PLHIV with positive serology to Leishmania infantum antigens; AIDS/VL = subjects in AIDS and VL = active visceral leishmaniasis patients without HIV; Recovered VL = subjects without HIV after cure of visceral leishmaniasis; DTH+ = subjects without HIV with positive Montenegro skin test.

https://doi.org/10.1371/journal.pntd.0012848.g002

CD57 is one of the known senescence associated molecules. The HIV group showed significantly lower proportions of senescent CD4⁺T cells than Asympt HIV/Leish or AIDS/VL subjects. Nonetheless there were significantly lower proportions of senescent CD4⁺T cells than in Asympt HIV/Leish or in AIDS/VL subjects (Fig 2 and S3 and S6 Tables). For senescent CD8⁺T cells, Asympt HIV/Leish expressed higher proportions of CD57⁺CD8⁺ T cells than UC, HIV, VL and DTH+ groups (Fig 2 and S1 Table). After cure, recovered VL subjects showed similar levels of senescent CD8⁺ T cells compared to VL patients (Fig 2 and S3 and S7 Tables).

PD-1 (CD279) is associated with activation and exhaustion [32]. Proportions of CD4⁺T expressing PD1⁺ in AIDS/VL subjects were significantly higher than among HIV subjects (Fig 2 and S3 and S8 Table). In contrast to CD4⁺ T cells, the overall proportions of CD8⁺T cells expressing PD1⁺ differed significantly across the groups (Fig 2 and S3 and S9 Tables). Proportions were highest in the Asympt HIV/Leish group, which differed from proportions in HIV, DTH+ or UC subjects. Levels of CD8⁺ PD1⁺ T cells were also high in HIV, AIDS/VL and VL subjects. There was no difference in AIDS/VL in proportions of CD8⁺ PD1⁺ cells compared to VL or Asympt HIV/Leish subjects. However, AIDS/VL had higher proportions than HIV or UC. VL patients had higher expression of PD in CD8⁺T cells than DTH+ or UC. After cure of VL (recovered VL), the levels of CD8⁺PD1⁺ T cells were similar to those in UC and DTH+ subjects (Fig 2 and S3 and S9 Tables).

To determine which factors were associated with activation and senescence in CD4⁺ and CD8⁺T cells a multivariable analysis was performed considering L. infantum or HIV presence or absence, viral load, CD4⁺ and CD8⁺ counts, rOD SLA and rODrK39. The presence of L. infantum infection was associated with activation of CD4⁺T (p = 0.04), CD8⁺PD1⁺ (p = 0.01), and senescence CD8⁺T cells (p = 0.006). HIV infection was associated CD8⁺PD1+ T cells (p = 0.04). CD8⁺PD1⁺T cells (p = 0.02) or senescent CD8⁺T cells (p = 0.0008) had weak associations with CD8 counts. Spearman correlation test showed negative correlation of CD4⁺ counts with activated CD4⁺T cells (r = -0.45; p = 0.02) and with CD4⁺PD1⁺ cells (r = -0.42; p = 0.02). SLA rOD was positively correlated with senescent CD8⁺T cells and (r = 0.42; p = 0.03) and CD8⁺PD1 cells (r = 0.51; p = 0.008). All the correlations are shown in S1 Fig.

Principal component analysis (PCA) was performed to tease out how the concomitant infections by L. infantum and HIV influenced the immune response and to narrow down the variables potentially involved with the phenotypes. All 69 subjects included with the immunological parameters were considered. The parameters are the ones presented in Figs 1 and 2 and Tables 2 and S3.

The PCA analysis suggested presence of two groups: one with active L. infantum infection (Asymptomatic HIV/VL, AIDS/VL and VL) and a second one which included the control group (UC), HIV or people who had prior L. infantum infection (DTH+, Recovered VL), as shown in Fig 3. Asymptomatic HIV/VL, AIDS/VL and VL had similar patterns, as shown previously in this study. The groups with anti-Leishmania antibodies (positive serology to SLA and rK39) are people with current L. infantum infection by virtue of presenting with anti-Leishmania antibodies, in L. infantum infection antibodies may be transient and they tend to inverse correlate with a positive cellular response and may be a sign of presence of parasite. Subjects who were DTH+ or Recovered VL followed apart from groups with active Leishmania infection. HIV and UC groups, without antibody production, were grouped together with DTH+ and Recovered VL in the PCA analysis. Thus, PCA analysis suggested division of active infection in Asymptomatic HIV/VL, AIDS/VL and VL and past and/or no Leishmania infection in DTH+, Recovered VL, HIV and UC groups. As the PC scores were used to evaluate the presence of clusters in groups variables, using K-means method, the silhouette plot showed the existence of clustering (S2 Fig), corroborating with the PCA. The K-means method suggested the existence of four clusters’ groups, mainly formed by VL (group 1); AIDS/VL (group 2); DTH+, HIV, REC VL and UC (group 3) and Asympt HIV/Leish (group 4). AIDS/VL and VL cases did not discriminate perfectly and it could be related to a low number of cases (S3 Table and S2 Fig).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Score plot of the principal component analysis results for immunological data for all analyzed groups.

Each dot in the graphs represents one individual. Each color in the PCA graph represents one groups of subjects: Blue dot = Asymptomatic HIV/Leish; Red dot = AIDS/VL; Pink dot = VL; Orange dot = HIV; Purple dot = Recovered VL; Green dot = DTH+; Gray dot = UC.

https://doi.org/10.1371/journal.pntd.0012848.g003

Discussion

The urbanization of L. infantum infections in northeastern Brazil and the coincident spread of HIV have contributed to the likelihood of concurrent infections. Since 2010, an increase in AIDS/VL cases as well as in asymptomatic L. infantum infection have been observed in the outskirts of Natal, as shown in this study, and in other Brazilian cities [33,34]. As it is well recognized, L. infantum infections may yield several outcomes varying from asymptomatic to severe systemic disease, which can be fatal even with treatment. However, the majority of individuals control infection if they have intact immune system [35]. Asymptomatic infection can be detected either by a Montenegro skin test (DTH+ subjects) or more recently by a Quantiferon test, that measures IFN-Ɣ production by Leishmania-specific cells in a research setting [36]. Peripheral blood mononuclear cells (PBMCs) from subjects with symptomatic VL have suppressed Leishmania antigen-specific immune responses which recover after cure, as shown by the magnitude of antigen-specific IFN- Ɣ, TNF, IL-2 and IL-6. In contrast, PBMCs from patients with VL release high amounts of IL-10 [5,37]. AIDS/VL subjects have been shown to have worse prognoses for VL, with higher rate of relapses [12]. People with HIV, particularly AIDS, have lower antibodies production by new infection, thus, AIDS/VL cases produce lower levels of SLA or rK39 antibodies [34,38]. Asymptomatic HIV/L. infantum infection may be detected by screening HIV subjects with no symptoms of VL with different diagnostic methods to detect L. infantum [34].

Herein we report a high prevalence (24.2%) of L. infantum positive serology among residents of the state of Rio Grande do Norte, Brazil, living with HIV. The prevalence of asymptomatic Leishmania spp. infections in people with HIV has been reported in other VL-endemic areas at rates between 3% and 25% [39–41]. Differences in the sensitivity and specificity of L. infantum detection tests can affect the calculated prevalence of asymptomatic Leishmania spp. co-infection [42]. In the past, cured L. infantum infection was estimated by the Montenegro skin test, which was generated from L. amazonensis antigen in Brazil. Despite the difference between species, the Montenegro test was a reliable method to detect cellular T cell responses to prior L. infantum infection in endemic regions [20,22]. Regulatory considerations have necessitated replacing the Montenegro with Quantiferon; the overlap between the two has yet to be documented but likely they are reasonably close. However, there is no Leishmania Quantiferon clinical test available, only in research settings. In the current study, rK39 and SLA ELISA were used to screen the PLHIV for current or prior infection with L. infantum. Although rK39 has been reported as more sensitive for acute as opposed to prior leishmaniasis, we observed a strong correlation between SLA rOD and rK39 rOD, indicating similar humoral responses in the HIV+ population. In immunocompetent subjects, a higher rK39 ELISA was associated with VL [18]. In the population of PLHIV, either SLA or rK39 could be used to identify PLHIV with asymptomatic L. infantum infection. AIDS/VL patients reportedly had lower titers or no response to anti- L. infantum antibodies prior to the availability of ART [38]. Of note, asymptomatic HIV/Leish (PLHIV with no clinical signs of L. infantum infection but positive serology) have an increased risk of developing AIDS/VL. The lower rOD results in ELISA tests observed here for recovered VL and for DTH+ subjects in comparison to VL cases are similar to our prior report [18].

Asymptomatic HIV/Leish and AIDS/VL subjects in this study were predominantly males. The prevalence of HIV is higher among males than females in the state of Grande do Norte [38]. Furthermore, symptomatic VL has also been reported to be more frequent in males [8,43]; despite the fact that both sexes were shown to be equally exposed to L. infantum infection in the region [20]. Demographic data suggest that sex hormones may play a role in the pathogenesis and progression to VL, once males reach puberty. Consistent with this hypothesis, higher levels of IL10 and TNF are observed in male than female mice infected with L. infantum [44]. Although we cannot dissect the contribution of higher male burden of HIV versus higher male susceptibility to VL in our population, it is possible that both contribute to the higher incidence of coinfection amongst males.

The consequence of HIV and L. infantum co-infection is an increased risk of progression to AIDS/VL according to the current and to published studies [1]. Specific T cell immune responses to Leishmania spp. antigens as detected by a positive DTH+ responses are a marker of protective immunity against symptomatic VL or relapsing from the disease. After cure, individuals who have recovered from VL had similar humoral and cellular immune responses as asymptomatic DTH+ subjects [5]. In the current study we observed an inverse correlation between CD4⁺ T cells counts and HIV viral load. Furthermore, viral loads were higher and CD4+ T cell counts were lower in AIDS/VL versus asymptomatic coinfection groups. This leads us to hypothesize there is a spectrum from asymptomatic HIV/Leish to AIDS/VL. We cannot tell from these data whether the two infections are synergistic, or whether VL symptoms are a consequence of worsening HIV infection status. To better understand that, we would need an additional group with Leishmania positive serology without HIV infection. Either way, the lack of adherence to ART therapy or a history of prior opportunistic infections were both associated with an increased risk of progressing to AIDS/VL. One assumes that adherence to ART could maintain protective Th1 responses by CD4⁺T cells [45].

L. infantum and HIV infections, alone or in presence of co-infections, cause several immunological alterations contributing to immune suppression either temporary as in the case of L. infantum, or long-term as in the case of HIV. In HIV infection, the co-expression of CD38 and HLA-DR on T cells was shown to predict HIV progression and correlated with high levels of viral load [1]. Persons with symptomatic VL have previously been shown to have higher percentages of ex vivo activation, but after L. infantum specific stimulation, T cell activation did not upregulate activation markers [5]. Our study showed lower ex vivo activation of both CD4+ and CD8+ T cells in DTH+ or recovered-VL subjects than those with Asympt HIV/Leish or AIDS/VL, although the former groups have been shown to mount specific T cell response after stimulation [5]. The results shown herein indicated high activation levels in AIDS/VL and Asympt HIV/Leish co-infected groups. The presence of L. infantum infection was associated with the activated status of both CD4⁺ and CD8+ T cells, as has been shown previously [23,46]. The chronic stimulation of cells by Leishmania spp. antigens could increase the levels of activation in comparison to healthy subjects. The data showed in this study suggest the presence of L. infantum infection may increase the levels of activation markers. In addition to direct effects of the parasite, patients with symptomatic VL have high circulating levels of LPS, sCD14 and intestinal fatty acid binding protein (-IFABP), indicating bacterial translocation across the gut mucosa may occur [6,23]. This is similar to HIV alone [47]. Nonetheless in our subjects, co-infection with L. infantum and HIV, in both the HIV/Leish and AIDS/VL groups, was associated with higher T cell activation markers in CD4⁺T and CD8⁺ T cells. The activation status of CD4⁺T and CD8⁺T cells was similar in both coinfected groups.

PD-1 expression is induced by T cell activation, and in a response to a strong TCR signal, it can suppress T cell functions. Cells with exhausted phenotype express the markers PD-1, TIM-3, LGA-3 [48,49]. In active symptomatic canine VL, exhausted CD4+ and CD8 T cells, increase PD-1 expression, decrease cell proliferation and IFN-γ production [46]. Chronic T activation, such as occurs in HIV infection or VL, can lead to dysfunctional T cells with reduced responses to antigen stimulation. Higher percentages of CD4⁺ or CD8⁺T cells expressing markers CD57⁺ (senescence) or PD1⁺ (marker of activation or exhaustion) were observed in asymptomatic HIV/Leish and AIDS/VL subjects compared to other groups. Senescent CD8⁺T (CD57⁺) or CD8⁺T PD1⁺ were similar and higher between asymptomatic HIV/Leish, AIDS/VL and VL subjects. A multivariate analysis indicated that L. infantum infection was associated with the levels of senescent CD8⁺ or CD8⁺ PD1⁺ cells. Senescent cells have shortened telomeres that can affect cellular function and proliferation [32]. Exhausted or senescent cells lost the capacity to perform their functions [50], as showed by reduced T cell proliferation and cytokine production after Leishmania antigen stimulation [5,37]. Exhausted or anergic CD8⁺T cells from active VL patients did not produce IFN-γ in response to specific stimulation. However, this suppression was transient, since after treatment, CD8⁺T cells recover the capacity of specific response [51]. Senescence of T cells was more accentuated in AIDS/VL than in HIV subjects in our study, although CD4⁺ T cell counts were significantly higher in the HIV group. Asymptomatic HIV/Leish T cells displayed similar proportions of senescence markers compared to the AIDS/VL group, raising the possibility that asymptomatic HIV/Leish group could be in the beginning stages of progression to functional impairment. Moreover, this similarity between Asymptomatic HIV/Leish and HIV/Leish group may be due to chronic HIV and/or Leishmania infantum antigen exposure or to increase of gut barrier permeability.

Asymptomatic HIV/Leish co-infected subjects seemed to be controlling both infections: this group did not have any VL symptoms and had similar HIV viral loads, CD4⁺ and CD8⁺ T cell counts in comparison to HIV subjects. These similarities reflect other published reports [39]. Based on our observations, we hypothesize that progression of asymptomatic to symptomatic VL in PLHIV, could be result of lack immune control of HIV infection, cause by poor ART adherence. The use of ART decreased the symptomatic VL in PLHIV as observed before [22]. As observed in HIV/Leish co-infection, the active L. infantum infection stimulates immune activation [23], leading to increased HIV viral load and accelerating progression to AIDS/VL [12,52]. Our results showed a greater risk of developing AIDS/VL in asympt HIV/Leish subjects who had low adherence to treatment and had previous infections. Together the above results suggest that L. infantum infection may be able to worsen caused by HIV infection, elevating the levels of activated and senescent T cells and amplifying disease progression.

This work had some limitations: 1) all PLHIV were studied in an era when ART was deferred until a relatively lower CD4 cell count, and some individuals had uncontrolled HIV and histories of several infectious complications. Moreover, we did not know when the subjects acquired HIV infection. 2) another limitation of this work was that we did not do PCR or DTH testing for L. infantum to assess the asymptomatic HIV/Leish group, and only positive serology was considered. 3) We did not include an additional group of subjects with positive Leishmania serology without HIV; instead, we evaluated DTH+ subjects and this group was considered immunocompetent asymptomatic VL; 4) Other markers as Tim-3, Lag-3 were not tested with PD-1, to determine exhaustion in T cells. In immunocompetent subjects, active VL and asymptomatic L. infantum infection present high levels of anti-Leishmania antibodies [53], but most AIDS/VL subjects present low levels of Leishmania antibodies and 30% are anti-Leishmania antibody negative [38]. To assess Leishmania infection in PLHIV, positive Leishmania serology was used to classify persons as asymptomatic HIV/Leishmania co-infected. We are not able to infer causality due to the cross-sectional design; thus, most of our results should be hypothesis-generating only. Furthermore, there may be significant confounding variables, such as gender, age, and ART adherence. Most HIV subjects were regularly on ART (13/16; 81.25%), while only 6 of 10 (60%) of Asympt HIV/Leish adhered to ART appropriately. ART adherence can influence the frequencies of T cell activation, senescence and exhaustion, resulting in decreased risk of co-morbidity [17].

The immunological study showed that L. infantum itself is associated with T cell activation and T cell senescence. However, in immunocompetent subjects this is temporary; after cure the frequencies of activated and senescent T cells decrease, as we have shown here in the Recovered VL group. In AIDS/VL subjects, the chronic activation induces loss of peripheral T cells and immune reconstitution could be achieved after HIV treatment, if the subject did not have a new episode of AIDS/VL (relapsing) [54].

Finally, the PCA showed that there is discrimination between subjects with anti-Leishmania antibodies and the remaining groups either controls or subjects with a previous history of L. infantum infection (DTH+). It is likely that active L. infantum infection results in greater activation and senescence of T cells. However, the major caveat is that those parameters decrease after resolution of the L. infantum infection either by treatment or spontaneously. A prior study with AIDS/VL in another population showed that there is gut bacterial translocation, which could increase the amount of LPS and induce greater inflammatory response [23].

The management of asymptomatic HIV/Leishmania co-infected subjects starts with optimizing ART therapy and close follow-up of clinical status, viral load and CD4⁺ T count. The role of anti-leishmanial therapy in the asymptomatic co-infected person is uncertain, although our data raise the possibility that it may be very important to maintain the infection in remission. Longitudinal trials are needed to determine whether, in the setting of universal ART and sustained viral load suppression, T cell activation and exhaustion are seen in PLHIV with positive serology for Leishmania and, if so, whether these immune parameters predict active VL despite suppressed virus.

Supporting information

Acknowledgments

We are thankful to the health personnel from Giselda Trigueiro Hospital, Central Laboratory (Laboratório Central de Saúde Pública do Rio Grande do Norte Dr. Almino Fernandes) for helping with the volunteers’ recruitment.