Ordeix et al. Parasites Vectors (2019) 12:575 https://doi.org/10.1186/s13071-019-3827-7 Parasites & Vectors RESEARCH Open Access Toll‑like receptors 2, 4 and 7, interferon‑gamma and interleukin 10, and programmed death ligand 1 transcripts in skin from dogs of different clinical stages of leishmaniosis Laura Ordeix1,2, Sara Montserrat‑Sangrà1, Pamela Martínez‑Orellana1, Marta Baxarias1 and Laia Solano‑Gallego1* Abstract Background: Canine leishmaniosis (CanL) caused by Leishmania infantum can have several dermatological manifes‑ tations. The type of immune response elicited against the parasite appears to be at the basis for such clinical variabil‑ ity. Much of the work in CanL has focused on adaptive immune response and there are scarce data on the importance of the innate immune responses. Moreover, few studies have evaluated the immunological response in the cutaneous lesions in dogs naturally infected with L. infantum and with different degrees of disease severity, and no study has compared clinically-lesioned with normal-looking skin. Methods: We determined and compared the transcription of toll like receptors (TLRs) 2, 4 and 7, interferon gamma (IFN-γ), interleukin (IL) 10 and programmed cell death protein ligand (PD-L) 1 by real-time PCR in paired clinically- lesioned and normal-looking skin from 25 diseased dogs (mild disease-stage I (n = 11) and moderate to severe disease-stages II and III (n = 14) as well as in normal-looking skin from healthy dogs (n = 10) from a non-endemic area. We also assessed the association between the transcripts in clinically-lesioned and normal-looking skin of dogs with leishmaniosis with clinicopathological, immunological and parasitological findings. Results: Clinically-lesioned skin from mildly affected dogs was characterized by a significant upregulation of TLR2 (P < 0.0001) and IL-10 (P = 0.021) and downregulation of TLR7 (P = 0.004) when compared with more severely affected dogs. Normal-looking skin of mildly affected dogs was characterized by a significant lower expression of TLR7 (P = 0.003), IFN-γ (P < 0.0001) and PD-L1 (P = 0.001) when compared with more severely affected dogs. TLR2, TLR4, IL-10 and IFN-γ upregulation in clinically-lesioned skin was correlated with lower disease severity while TLR7 upregula‑ tion was correlated with markers of disease severity. Upregulation of TLR7, IL-10, IFN-γ and PD-L1 in normal-looking skin was correlated with disease severity. Conclusions: This study demonstrated different expression profiles of immune genes in clinically-lesioned and normal-looking skin among mildly and more severely affected dogs. These immunological conditions might favor the maintenance and replication of the parasite in the skin of more severely affected dogs. Keywords: Leishmaniosis, Canine, Leishmania infantum, Skin, Cytokine, Toll-like receptors, Gene expression *Correspondence: laia.solano@uab.cat 1 Department de Medicina i Cirurgia Animals, Facultat de Veterinària, Universitat Autònoma de Barcelona, Bellaterra, Spain Full list of author information is available at the end of the article © The Author(s) 2019. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat​iveco​mmons​.org/licen​ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Ordeix et al. Parasites Vectors (2019) 12:575 Page 2 of 13 Background data are available regarding normal-looking or lesioned Canine leishmaniosis (CanL) caused by Leishmania skin from infected or diseased dogs [11]. infantum is a zoonotic and an endemic disease in the A mixed Th1/Th2 cytokine profile in the dermis of dogs Mediterranean basin among other areas such as south naturally infected with L. infantum has been described America, Middle East and Asia [1]. The complex immune [14–17]. Recently, there has been a great interest in the response against the parasite is crucial for determining involvement of Toll Like Receptors (TLRs) in the immu- the outcome of infection [2]. In fact, subclinical infec- nopathogenesis of CanL [16, 18–20]. TLRs are one of tion is the result of an effective T helper 1 (Th1) cellular the most important pattern recognition receptor (PRR) immunity, with the activation of macrophages by inter- molecules which recognize molecular structures charac- feron-gamma (IFN-γ) and tumor necrosis factor-alpha teristic of microbial pathogens and induce an inflamma- (TNF-α) and the elimination of intracellular amastigotes tory response [21]. Studies aimed to determine the role via the l-arginine nitric oxide pathway [2, 3]. On the of TLRs in CanL are mainly in vitro studies performed on other hand, disease development and progression are canine macrophages [22] or studies performed in blood often correlated with increased parasite burdens together [23], liver [16], spleen [16, 20, 24], intestine [19], brain with a strong but non-protective humoral immune [20, 24] or lymph node samples [16, 20]. TLR2 is one of response and reduced or absent T cell-mediated immu- the TLRs associated with the pathogenesis of cutaneous nity [1]. lesions in CanL [17, 25]. Canine leishmaniosis is a systemic disease with var- As discussed above, the suppression of cellular immu- ied clinical signs that range from a self-limiting disease nity is the most important aspect in the pathogenesis and to severe illness or even death [1]. Therefore, a clinical progression of CanL [26]. During the last years, several staging system for CanL that classifies the disease into studies have focused on the regulatory mechanisms and four stages (stage I or mild disease, stage II or moderate have demonstrated that programmed cell death pro- disease, stage III or severe disease and stage IV or very tein (PD)-1 and its ligand (PD-L1) present in regulatory severe disease) based on clinical signs, clinicopathologi- ­IgDhi B cells are involved in the induction of T lympho- cal abnormalities, and measurement of anti-leishmanial cyte apoptosis via IL-10 production [27]. These studies antibodies was previously proposed [1] and recently have determined an increased PD1/PD-L1 expression updated [4]. in peripheral mononuclear cells as well as an increase in Among the different clinical manifestations of CanL, the expression of PD-L1 in splenic macrophages in dogs dermatological disease is the most frequent [5, 6]. Cuta- with leishmaniosis [27–29]. However, to the best of our neous lesions are very pleomorphic from a clinical and knowledge, PD-L1 expression in the skin of diseased or histopathological point of view [5] and this clinicopatho- infected dogs has not been investigated. logical variation might reflect a different host-parasite Only few published studies [14–17] have investi- relationship and immune interactions [6, 7]. This is the gated the immunological response in the skin in dogs particular case of papular dermatitis [6]. Papular derma- naturally infected with L. infantum and with differ- titis is a typical dermatological manifestation of CanL ent degrees of disease severity. Moreover, these studies in an endemic area [5], which is classified as a stage I or have been mainly performed on normal-looking skin mild disease in the absence of other clinicopathologi- [15–17]. Therefore, the main objective of this study was cal abnormalities [4]. It has been suggested that there is to determine and compare the transcription of TLR2, strong T-cell mediated immunity against L. infantum that TLR4, TLR7, IFN-γ, IL-10 and PD-L1 in paired clini- configures protection in these dogs [7–10]. On the other cally-lesioned and normal-looking skin from dogs with hand, other dermatological signs observed in CanL, such different clinical stages of leishmaniosis. Furthermore, as exfoliative dermatitis, ulcerative dermatitis, onych- we assessed the association between the transcripts in ogryphosis and muco-cutaneous nodular dermatitis are clinically-lesioned and normal-looking skin of dogs with commonly observed in dogs with moderate to severe leishmaniosis with clinicopathological, immunological leishmaniosis [5, 6]. and parasitological findings The immune response in CanL has been the focus of many investigations during the last years. However, Methods much of this work focused on adaptive immune response Study groups and the data on the importance of the innate immune Twenty-five dogs with CanL and dermatological mani- responses are scarce [11]. It is currently accepted that the festation were prospectively selected from different vet- immune response to the parasite is compartmentalized erinary centers in Catalonia and the Balearic Islands and different among organs [12, 13]. While the skin plays (Spain). These dogs were previously described in a pub- a major role in CanL immunopathogenesis, very limited lished study aimed to characterise and compare the Ordeix et al. Parasites Vectors (2019) 12:575 Page 3 of 13 inflammatory pattern and the parasite burden in paired with a rotor-stator homogenizer (T 10 basic ULTRA- dures. Total RNA was then isolated using the RiboPure™ clinically-lesional and normal-appearing skin from the TURRAX 230V IKA 3420000) using standard proce- same dogs with dermatological manifestation due to CanL at different stages of disease [7]. Briefly, diagnosis Kit (Ambion) under strict RNase-free condition accord- was based on the observation of L. infantum on cyto- ing to the manufacturer’s protocol. In order to remove included using TURBO DNA-free™ DNase Treatment logical and/or dermatopathological examination with contaminating DNA, a DNase digestion step was or without Leishmania-specific immunohistochemical examination of cutaneous lesions [7]. Moreover, a com- and Removal Reagents (Ambion) following the manu- plete blood count using the System Siemens Advia 120 facturer’s instructions. RNA concentration was deter- hematology analyzer (Siemens Healthcare GmbH, Erlan- mined by a Nanodrop device (Thermo Fisher Scientific, gen, Germany), a biochemical profile with the use of an Waltham, USA) and RNA integrity and quality were USA), serum protein electrophoresis using Hydrasys® Olympus AU 400 analyzer (CLIAwaived, San Diego, assessed by using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, USA) in some biopsies. Sam- (Sebia Electrophoresis, Norcross, USA), urinalysis with ples had a final concentration of 9.4–881.2 ng/µl. The urinary protein/creatinine ratio calculation and quanti- majority of samples included in this study had an RNA tative serology for the detection of L. infantum-specific integrity number value greater than 7. The recovered antibodies by means of a serial dilution in-house ELISA RNA was stored at – 80 °C until cDNA synthesis. using the whole L. infantum antigens (strain: MHOM/ cDNA was generated using the SuperScript™ VILO™ FR/78/LEM75 zymodeme MON-1) were performed [30]. cDNA synthesis Blood Leishmania kinetoplast quantitative polymerase chain reaction (qPCR) was also performed [30]. Based cDNA Synthesis Kit (Invitrogen, Thermo Fisher Scien- on clinicopathological data, dogs were classified into tific, Carlsbad, USA) according to the manufacturer’s three clinical stages: LeishVet stage I-mild disease char- instructions. cDNA was aliquoted and stored at – 20 °C acterized by persistent papular dermatitis (n = 11); II- until used for qPCR. moderate disease (n = 12); III-severe disease (n = 2) as previously reported [1]. However, for comparative analy- Quantitative PCR sis dogs were divided into two groups: Group A (11 dogs Canine reference and target immune genes used in this with LeishVet stage I and papular dermatitis); Group B study [23, 31, 32] are listed in Table 1. PCR amplifica- (14 dogs with LeishVet stages II and III and exfoliative tion was performed using the QuantStudioTM 12K Flex TaqMan® Universal Master Mix II with UNG (Applied or ulcerative dermatitis). Normal-looking skin samples System Real-Time PCR (Thermo Fisher Scientific) using from 10 clinically healthy non-infected Beagle dogs from a non-endemic area (UK) (Group C) were used as control Biosystems, Foster City, USA). Plates (96-well plates) dogs. were filled with 0.35 µl nuclease free water (Sigma- Aldrich, San Luis, USA), 7.50 µl TaqMan Universal Mas- Skin biopsies ter Mix (2×), 0.75 µl TaqMan assay 20 and 6.4 µl 1/5 For all patients two skin fragments ≤ 0.5 cm from clini- cDNA. Plates were closed with an optical film (Applied cally lesioned skin and skin with normal appearance were Biosystems) centrifuged in order to mix the samples and collected. Normal-looking skin was obtained whenever placed into a laboratory pipetting robot (Epmotion 5057 possible from the lateral aspect of the neck. In cases Liquid-handlingrobot, Eppendorf, Hamburg, Germany) where this region was affected, the biopsy was obtained to generate a 384-well plate. Then, the generated 384 well from an area as far as possible from the macroscopically plates were transferred into a real time PCR device. The affected lesions. Each skin sample was then immediately PCR components and the PCR cycler conditions were cut into two halves. One half was fixed in 10% formalin identical for the all target and reference genes. Denatura- for descriptive histopathology and analysis of the dermal tion program (95 °C, 10 min), amplification and quanti- one was submerged in RNA later ­(RNAlater® Stabiliza- parasite density as described previously [7] and the other fication program were repeated 40 times (95 °C for 15 s, 60 °C for 10 s, 72 °C for 60 s) with a single fluorescence tion Solution, Ambion, Inc., Austin, USA), stored at 4 °C measurement. The baseline and threshold were automati- overnight and then keep at − 80 °C until used. cally defined for the program in each run. Each sample was performed in triplicate for all the target and refer- RNA extraction ence genes and a calibrator sample (one sample from Before RNA isolation protocol, skin samples were Group C) was employed as control in each plate. All tar- RiboPure™ Kit, Ambion, Austin, USA) and homogenized thawed on ice and placed in lysis solution (TRI Reagent, get genes per each dog were run on the same day and in the same plate. Data were processed while applying the Ordeix et al. Parasites Vectors (2019) 12:575 Page 4 of 13 Table 1 Canine reference and target immune genes used in the present study Assay ­IDa Gene symbol Gene name GenBank mRNA GenBank reference sequence Amplicon size Cf02625049_s1 TLR2 Toll-like receptor 2 AF328930.1 NM_001005264.2 69 Cf02622203 g1 TLR4 Toll-like receptor 4 AB080363.1 NM_001002950.1 120 Cf02710573_s1 TLR7 Toll-like receptor 7 AB248956.1 NM_001048124.1 124 Cf02624265_m1 IL-10 Interleukin 10 AF328930.1 NM_001003077.1 83 Cf02623316_m1 IFN-γ Interferon gamma AF091130.1 NM_001003174.1 117 APG2FND PD-L1 Programmed dead ligand 1 NM_001291972.1 NM_001291972.1 164 Cf02643820_m1 LOC479750 Similar to CG14980-PB XM_536878.2 XM_536878.2 78 Cf02664981_m1 SDHA Succinate dehydrogenase com‑ XM_535807.2 DQ402985.1 64 plex; subunit A; flavoprotein a All the assays are commercially available from Thermo Fisher Scientific relative quantification method comparable to the delta- Medicago, Uppsala, Sweden) at a concentration of 10 µg/ delta-quantification cycle value (ddCq) method. For ml. IFN-γ was determined in supernatants obtained five normalization of target gene expression, the arithmetic days after stimulation by a commercial sandwich ELISA mean of the two reference genes were taken for the cal- (DuoSet ELISA by Development System R&DTM, Abing- culation of a reference gene index [23]. Quantitative PCR don, UK). Cytokine concentration from supernatants data analyses was done by the Cloudsuite software (Life with ConA and LSA was calculated after subtracting ­technologiesTM, Thermo Fisher Scientific). the IFN-γ concentration obtained from unstimulated supernatants. Skin parasite load DNA was purified from the interphase and organic phase Statistical analysis generated from the RNA purification process by means Statistical analysis was performed using the IBM SPSS of QIAamp DNA Mini Kit (Qiagen, Manchester, UK) fol- Statistics software (version 1.0.0.1032) (SPSS Inc., Chi- lowing the manufacturerʼs instructions with slight modi- cago, USA) and the blorr, generalhoslem and Deducer fications. Briefly, 20 μl of proteinase K solution and 200 packages of the R software i386 version 3.4.2 (R Devel- μl of tissue sample were used in all samples. The other opment Core Team) for Windows software. Categori- steps were performed as per manufacturerʼs protocol. A cal data were expressed as percentages and statistical fragment of skin from a control dog was used as a control analysis was performed using the Fisher’s exact test to for DNA contamination during DNA extraction. qPCR compare independent variables. Quantitative data are was performed with L. infantum specific oligonucleo- expressed as the means ± standard deviation (SD). The tide primers N13A (5ʹ-AAC TTT TCT GGT CCT CCG non-parametric Wilcoxon signed-rank test and Mann- GG-3ʹ) and N13B (5ʹ-CCC CCA GTT TCC CGC CC-3ʹ) Whitney U-test were used to compare related and inde- were used to amplify a 120-bp fragment of the Leishma- pendent variables, respectively. The Spearman’s rank nia kinetoplast DNA minicircle as previously described order correlation between transcripts in skin samples and [7]. The parasite load was measured with the calculation immunological (L. infantum specific antibody levels and of the delta Cq (dCq = mean values of duplicate deter- blood IFN-γ production), clinicopathological and parasi- mination of Leishmania Cq -18S rRNA Cq). Therefore, tological data was also calculated. low or negative values of dCq represented higher parasite Multivariable logistic regression was constructed to load than elevated dCq. assess the relationships between all transcripts studied and skin Leishmania qPCR in clinically-lesioned skin Whole blood IFN‑γ release assay when compared with normal-looking skin. The model An IFN-γ release whole blood cultured assay was per- was performed with blorr, generalhoslem and Deducer formed as described previously [33]. Briefly, 500 µl of hep- packages of the R software. Only quantitative and quali- arinized whole blood was separately mixed with 4.5 ml of tative variables with a P-value of equal or less than 0.2 three different conditions: (i) unstimulated medium; (ii) based on the univariate analysis were included in the medium with soluble L. infantum antigen (LSA, 5 mg/ model. Models were performed with data for all dogs ml, Facultat de Farmacia, Universitat Autònoma de Bar- and with data obtained from dogs of Group A and B, celona) at a concentration of 10 µg/ml; and (iii) medium separately. The response variable was the type of skin with the mitogen concanavalin A (ConA, 100 mg, (clinically lesioned or normal-looking skin) and the Ordeix et al. Parasites Vectors (2019) 12:575 Page 5 of 13 explanatory variables were the results of the skin Leish- Z = − 2.773, P = 0.006). All Beagle healthy dogs (Group mania qPCR (numeric result and interpretation) and C) were male and between three and six years of age. the transcripts (TLR2, TLR4, TLR7, IL-10, IFN-γ and PD-L1). Multivariate models were constructed in a step- Histology wise fashion, beginning with a full model and removing Histological features of the skin of diseased dogs (Group variables one-by-one. Those variables included in the A and B) were previously described elsewhere [7]. Briefly, model were assessed for collinearity by Spearman’s cor- normal-looking skin of dogs from Group B was more fre- relation test, a result of ≥ 0.7 between two variables was quently inflamed than normal-looking skin of dogs from assessed as two non-independent correlated variables Group A (78.6% and 27.3%, respectively; Fisher’s exact and the variable with the weakest relationship with the test, P = 0.017). The perivascular to interstitial inflam- response variable (higher P-value) was excluded. Good- matory pattern was more common in clinically-lesioned ness-of-fit was assessed by deviance of the residuals, rul- skin from dogs from Group B than in clinically-lesioned ing out overdispersion. A ROC curve was used to validate skin from dogs from Group A (71.4% and 18.2%, respec- the model. Differences were considered significant with a tively; Fisher’s exact test, P = 0.015). On the other hand, 5% significance level (P < 0.05). nodular to diffuse pattern with granuloma formation was more common in clinically-lesioned skin from dogs from Group A than in clinically-lesioned skin from dogs from Results Group B (36.4% and 0%, respectively; Fisher’s exact test, Dogs P = 0.017). Eleven purebred dogs belonging to ten breeds and 14 mixed-breed dogs were included. Both sexes were rep- Immunological, parasitological and clinicopathological resented by 11 females and 14 males. The median age data was 2.5 years with a range from five months to 10 years. Control dogs from a non-endemic area (Group C) were Specifically, dogs from Group A were six females and deemed clinically healthy seronegative non-infected dogs five males with a median age of 10 months, whereas dogs and were not included in comparisons reported in this from Group B were five females and nine males with a section. The most relevant evaluated parameters studied median age of 54 months. Age difference was statisti- in diseased dogs are listed in Table 2. As expected, dogs cally significant among groups (Mann-Whitney U-test, classified in Group A were in a less severe disease status Table 2 Clinicopathological data, antibody levels, IFN-γ production in stimulated blood and skin parasite load of sick dogs (groups A and B) Parameter (reference intervals and units) Group A Group B P-value Z-value (n = 11) (n = 14) UPC (< 0.5) 0.2 ± 0.1 0.7 ± 0.7 Creatinine (0.5–1.5 mg/dl) 0.8 ± 0.1 0.9 ± 0.2 Urea (21.4–59.9 mg/dl) 43 ± 14 31.3 ± 7.8 0.037 − 2.083 Total protein (5.4–7.1 g/dl) 5.7 ± 0.6 9.1 ± 2.4 0.0001 − 3.686 Albumin (2.6–3.3 g/dl) 3 ± 0.4 2.3 ± 0.8 Beta globulin (0.9–1.6 g/dl) 1.3 ± 0.3 1.9 ± 0.6 0.01 − 2.572 Gamma globulin (0.3–0.8 g/dl) 0.7 ± 0.9 3 ± 2.6 0.002 − 3.079 Abumin/globulin ratio 1.1 ± 0.2 0.5 ± 0.4 0.004 − 2.899 Hematocrit (37–55 %) 45.6 ± 8.5 33 ± 9.8 0.024 − 2.251 Hemoglobin (12–18 g/dl) 15.4 ± 2.4 11.1 ± 3.8 0.032 − 2.148 Leishmania infantum specific antibody levels (cut-off 136.8 ± 196.1 8892.7 ± 17807.7 < 0.0001 − 3.887 value 35 EU units) Blood L. infantum specific IFN-γ (pg/ml) 2046 ± 2746.6 713.7 ± 832.8 Blood parasite load (parasites/ml) 7.8 ± 10.5 6.7 ± 11.1 Skin parasite load (dCq) Clinically-lesioned 3.6 ± 4.4 − 0.4 ± 4.7 0.043 − 2.026 Normal-looking 6.1 ± 4 1.7 ± 4.5 0.004 − 2.869 Notes: Low or negative values of dCq represent higher parasite load than elevated dCq. Differences were assessed via Mann–Whitney test Abbreviations: UPC, urinary protein/creatinine ratio; EU, enzyme-linked immunosorbent assay units; dCq, delta-cycle threshold Ordeix et al. Parasites Vectors (2019) 12:575 Page 6 of 13 than dogs classified in Group B as they had significantly Transcripts in normal‑looking skin in sick dogs lower values for total proteins, beta and gamma globulins with different clinical staging and higher values for albumin/globulin ratio, hemato- Normal-looking skin from dogs of Group A showed crit and hemoglobin. Moreover, dogs from Group B had significant downregulation of TLR7, IFN-γ and PD-L1 significantly higher levels of specific antibodies and skin in comparison with normal-looking skin from dogs of parasite load in dogs with both clinically-lesioned and Group B (Fig. 1). Although non-statistically significant, a normal-looking skin than in dogs from Group A. trend for a downregulation of TLR2 and IL-10 was also observed in Group A when compared with Group B. Transcripts in clinically‑lesioned skin in comparison Correlations with transcripts in clinically‑lesioned with healthy skin from control dogs skin and clinicopathological, immunological Relative quantifications of the expression of the immune and parasitological findings response genes analyzed in the present study are shown Correlations between transcripts in clinically-lesioned in Fig 1. All transcripts except TLR7 were significantly skin from sick dogs and the different parameters are increased in clinically-lesioned skin from dogs of Group illustrated in the heatmap shown in Fig 2. A significant A when compared with Group C (Fig. 1). In Group A, positive correlation was noted between TLR2, TLR4 and TLR7 was significantly downregulated. On the other IL-10 transcripts, whereas a significant negative cor- hand, although all the transcripts were higher in clini- relation was observed between TLR2 and total protein cally-lesioned skin from dogs of Group B when compared and specific L. infantum antibodies. TLR4 in addition with Group C, only TLR2, IFN-γ, IL-10 and PD-L1 were was negatively correlated with gamma globulin concen- significantly upregulated (Fig. 1). tration. TLR7 was the transcript with more significant correlations obtained. In fact, it was associated with Transcripts in clinically‑lesioned skin in comparison clinicopathological parameters suggestive of disease with paired normal‑looking skin from sick dogs severity. Furthermore, IL-10 was positively correlated Dogs from Group A showed significant upregulation of to skin IFN-γ transcript, hematocrit, hemoglobin and TLR2, IL-10, IFN-γ and PD-L1 in clinically-lesioned skin blood IFN-γ production, whereas a negative correlation when compared with paired normal-looking skin. In con- between IL-10 and specific antibodies was observed. Skin trast, in Group B only TLR2 transcript was significantly IFN-γ transcript, in addition to the aforementioned cor- higher in clinically-lesioned skin when compared with relations, was positively correlated with PD-L1. paired normal-looking skin (Fig. 1). Correlations with transcripts in normal‑looking skin and clinicopathological, immunological Transcripts in clinically‑lesioned skin in sick dogs and parasitological findings with different clinical staging More significant correlations were determined between Clinically-lesioned skin from dogs of Group A showed transcripts and clinical, immunological and parasito- significant upregulation of TLR2 and IL-10 and downreg- logical findings in normal-looking skin than in clini- ulation of TLR7 in comparison with clinically-lesioned cally-lesioned skin (Fig. 3). The TLR2 transcript was skin from dogs of Group B (Fig. 1). Although non-sta- positively correlated with TLR4, IL-10 and IFN-γ. A tistically significant, a trend for an upregulation of TLR4 significant positive correlation was obtained between and IFN-γ was also observed in Group A. TLR7 and IL-10, IFN-γ, PD-L1, total protein, beta and gamma globulins and specific L. infantum anti- bodies. TLR7 was negatively correlated with albumin, Transcripts in normal‑looking skin from sick dogs albumin/globulin ratio, hematocrit, hemoglobin, and in comparison with healthy skin from control dogs Leishmania dCq in qPCR. IL-10 showed similar cor- Relative quantification of TLR7 and IL-10 was signifi- relations than TLR7, except for a positive correlation cantly downregulated and upregulated, respectively, in with TLR2 and UPC ratio and a negative correlation the skin from dogs of Group A when compared with the with blood IFN-γ production. Skin IFN-γ was positively skin from dogs of Group C (Fig. 1). Although all the tran- correlated with TLR2, TLR7, IL-10, PD-L1, UPC ratio, scripts were higher in normal-looking skin from dogs of total protein, beta and gamma globulins and specific L. Group B when compared with Group C, only TLR2, IFN- infantum antibodies. In addition, there was a negative γ, IL-10 and PD-L1 were significantly upregulated. correlation between skin IFN-γ and albumin, albumin/ globulin ratio, hematocrit, hemoglobin and Leishmania Ordeix et al. Parasites Vectors (2019) 12:575 Page 7 of 13 Fig. 1 Relative quantification of the immune genes studied. a TLR2 transcripts. b TLR4 transcripts. c TLR7 transcripts. d IL-10 transcripts. e IFN-γ transcripts. f PD-L1 transcripts. Circles and squares represent individual data of each dog. Horizontal and vertical lines represent mean and standard deviation, respectively. Solid lines with P-values: Mann-Whitney U-test; dashed lines with P-values: Wilcoxon signed-rank test. Abbreviations: A, clinically-lesioned skin and normal-looking skin from Group A (stage I-mildly affected dogs); B, skin from Group B (stage II-III-severely affected dogs); C, skin from Group C (healthy non-infected dogs). Rq, normalized relative quantification Ordeix et al. Parasites Vectors (2019) 12:575 Page 8 of 13 Fig. 2 Heatmap illustrating the positive (red) and negative (blue) correlations between immune gene transcripts in clinically-lesioned skin from Group A (stage I- mildly affected dogs) and Group B (stage II-III-severely affected dogs) and clinicopathological, immunological and parasitological findings (correlations with P < 0.05 are indicated by an asterisk) dCq in qPCR. Finally, PD-L1 was positively correlated Discussion with TLR7, IL-10, IFN-γ, UPC ratio, total protein, beta This study aimed to investigate, to our knowledge for and gamma globulins and antibody levels, whereas a the first time, the transcription of TLR2, TLR4, TLR7, negative correlation was observed between this tran- IFN-γ, IL-10 and PD-L1 in paired clinically-lesioned and script and albumin, albumin/globulin ratio, hematocrit, normal-looking skin from the same dogs with differ- hemoglobin and Leishmania dCq in qPCR. ent clinical stages and degrees of disease severity. More A low dCq means higher skin parasite load. Therefore, significant differences among immune gene transcripts negative correlations determined between dCq values in normal-looking skin and clinically-lesioned skin were and TLR7, IL-10, IFN-γ and PD-L1 mean an association determined in mildly affected dogs than in more severely between an upregulation of these transcripts and a high affected dogs. This was not surprising considering that parasite load in normal-looking skin. normal-looking skin of more severely affected dogs pre- sent an increased frequency of microscopic inflammatory lesions and higher parasite load than in normal-looking Multivariable logistic regression models skin of mildly affected dogs [7]. A multivariable logistic regression model was con- TLR2 was significantly upregulated in clinically- structed with data of all dogs. The model showed that lesioned skin of sick dogs when compared with healthy TLR2 and IFN-γ transcripts are upregulated in clinically- skin of non-infected dogs as previously documented in lesioned skin from dogs with CanL (odds ratio > 1). The other tissues including the intestine [19], brain [20, 24], model developed with data of Group A did not show peripheral lymphoid organs [16, 20], liver [16], blood [23] any statistical significance, whereas for Group B showed and skin [16, 17], and this was associated with disease TLR2 upregulation (odds ratio > 1). The results are sum- severity and progression. Although a lower expression marized in Table 3. of TLR2 in clinically-lesioned skin of dogs with papular Ordeix et al. Parasites Vectors (2019) 12:575 Page 9 of 13 Fig. 3 Heatmap illustrating the positive (red) and negative (blue) correlation values between immune gene transcripts in normal-looking skin from Group A (stage I-mildly affected dogs) and Group B (stage II-III-severely affected dogs) and clinicopathological, immunological and parasitological findings (correlations with P < 0.05 are indicated by an asterisk) Table 3 Multiple logistic regression models performed with all data available that presented statistical significance Variables Coefficient (β) SE χ2 P-value Odds ratio 95% CI All dogs (clinically-lesioned Intercept − 1.99 0.52 vs normal-looking skin) TLR2 transcript 0.240 0.060 6.790 0.009* 1.29 0.62–2.73 (n = 35) IFN-γ transcript 0.003 0.001 4.570 0.033* 1.07 0.52–2.23 Group B (clinically- Intercept − 1.150 0.910 lesioned vs normal- TLR2 transcript 0.300 0.150 4.120 0.042* 1.34 0.47–3.82 looking skin) (n = 14) IL-10 transcript − 0.260 0.160 2.790 0.095 0.78 0.26–2.15 IFN-γ transcript 0.003 0.002 3.230 0.072 1.00 0.35–2.85 PD-L1 transcript 0.160 0.110 2.190 0.139 1.16 0.40–3.29 Notes: Coefficient is the mathematical weightings of the explanatory variables in the equation. Intercept is a mathematical constant of the model (no clinical interpretation) Abbreviations: CI, confidence interval; SE, standard error; group B, stage II-III-severely affected dogs *P < 0.05 dermatitis than in the skin of more severely affected be related to the retrospective design of the study by dogs was initially suggested [25], the present results Esteve et al. [25], the lower number of cases included and contradict the results of [25]. This discrepancy might because TLR2 expression has been measured by means Ordeix et al. Parasites Vectors (2019) 12:575 Page 10 of 13 of immunohistochemistry, a technique that is less accu- organ compartmentalization of TLR4 gene expression rate and sensitive than qPCR. Moreover, the discrepan- could be possible as no differences were found in TLR4 cies might be because the two techniques are evaluating relative quantification in non-stimulated blood between different parameters (protein versus mRNA expression). mildly and more severely affected dogs [35]. Nonethe- Similar to the present study, TLR2 gene expression has less, and similar to TLR2, milder forms of human cuta- been documented in different clinical presentations of neous leishmaniasis due to L. braziliensis are associated tegumentary leishmaniasis in humans [34]. Mild forms with higher expression of TLR4 [34]. Moreover, TLR4 of the disease (i.e. localized cutaneous leishmaniasis and polymorphisms have been associated with susceptibility borderline disseminated cutaneous leishmaniasis) caused to cutaneous leishmaniasis in humans [36, 37]. There- by L. braziliensis has been reported to present higher fore, TLR4 might induce an adequate proinflammatory TLR2 expression than the severe form mucosal leishma- response to control L. infantum infection in the skin of niasis [34]. In addition, TLR2 transcript was negatively dogs as previously suggested in humans [34]. correlated with total protein and specific antibody levels There are limited studies that determine TLR7 tran- in clinically-lesioned skin, rendering our finding, those scripts in CanL and TLR7 transcription appears to be dogs with stage I presented higher TLR2 transcript in unchanged in brain and spleen [24] as well as in canine clinically-lesioned skin, reasonable. TLR2 might induce monocyte-derived macrophages [22]. This TLR has been an adequate proinflammatory response to control L. rarely studied in canine skin [38] and it has never been infantum infection in the skin of dogs as previously sug- studied in the skin of dogs infected with L. infantum. gested in humans [34]. Interestingly, in the present study, TLR7 gene expres- Upregulation of TLR2 gene was evident in normal- sion was significantly lower in dogs with both clini- looking skin of dogs naturally affected by leishmaniosis cally-lesioned and normal-looking skin with papular [17]. In the present study, only more severely affected dermatitis than in more severely diseased dogs. Moreo- dogs showed a significant overexpression of TLR2 in ver, TLR7 overexpression in either in clinically-lesioned normal-looking skin when compared with healthy skin of and normal-looking skin was associated with altered non-infected dogs, and, although not statistically signifi- clinicopathological parameters suggestive of disease cant, dogs with papular dermatitis showed lower TLR2 severity. Based on these results, a pathogenic role of this gene expression than more severely diseased dogs. This innate receptor in CanL is likely. In fact, recent evidence difference may be related to the increased frequency of associated TLR7 activation with disease exacerbation of microscopic inflammatory lesions and higher parasite visceral leishmaniasis due to L. donovani in mice [39, 40]. load in normal-looking skin of more severely than in Endosomal TLR7 activation in B cells by L. donovani has mildly affected dogs [7]. Taken all these findings together, been suggested to be responsible for disease exacerba- it would seem that in more severely affected dogs there tion through IL-10 and IFN-type I production and for the is a progressive TLR2 downregulation from earlier stages promotion of hypergammaglobulinemia [39]. Moreover, of inflammation to more chronic dermatitis. This reflec- local tissue damage mediated by persistent inflammation tion agrees with the observation of Hosein et al. [16] who has been reported to lead to suppression of protective T described an upregulation of TLR2 in the skin only in the cell responses during chronic visceral leishmaniosis due earlier stages of an experimental infection when com- to L. donovani in mice via signaling of TLR7 by apoptotic pared with the controls [16]. cell material [40]. TLR4 has been scarcely studied up to now in CanL Cytokine studies on clinically-lesioned skin are very in several tissues [16, 20, 23, 24], but, to the best of our limited [14]. Noteworthy, in the present study, IL-10 knowledge, never in lesioned skin of dogs with leish- gene expression was studied in clinically-lesioned skin, to maniosis. Most of previously published data indicate an our knowledge, for the first time. A significantly higher upregulation of TLR4 in several tissues such as spleen, IL-10 gene expression in papular dermatitis than in the lymph nodes [20] or brain [24] from sick dogs. In the skin lesions of more severely affected dogs was found present study the TLR4 transcript was significantly and a positive relationship of this cytokine expres- increased in clinically-lesioned skin of dogs with papu- sion with parameters associated to disease control was lar dermatitis compared with control skin. In addition, observed. IL-10 is an immunoregulatory cytokine with a higher TLR4 transcription level was observed in these multiple roles in immunopathology [41] but the roles dogs than in more severely affected dogs, although the of IL-10 in CanL remain uncertain. However, it seems difference was not significant. Moreover, a negative cor- that IL-10 is not a marker of disease severity at least in relation among this transcript and gamma globulins is, clinically-lesioned skin as previously observed in IFN-γ to our knowledge, demonstrated for the first time, sug- whole blood release assays (WBA) [13, 33] opposite gesting an association with less disease severity. Also, an to reports in mice and humans [42, 43]. In contrast, Ordeix et al. Parasites Vectors (2019) 12:575 Page 11 of 13 polysymptomatic-diseased, naturally infected dogs have skin was observed among different disease stages. This presented an increased IL-10 production by T lympho- was an unexpected finding as higher T-cell apopto- cytes from blood along with increased blood parasite sis was hypothesized in clinically-lesioned skin of more burden [44]. As previously described [13, 15, 17] and in severely affected dogs. It would be interesting to evaluate agreement with the present findings, an upregulation of if further increase of the number of studied dogs would IL-10 in normal-looking skin of dogs with leishmaniosis change this finding. On the other hand, it is possible that was detected and associated with parameters of disease factors other than PD-L1 exist as a cause of suppression severity such as parasite density. In summary, higher lev- of Th1 cell effector function as previously suggested [46, els of IL-10 gene expression would be an immunological 47]. parameter marker of disease severity in normal-looking A lower PD-L1 gene expression was determined in nor- skin but not in clinically-lesioned skin. mal-looking skin from dogs with papular dermatitis than As expected, an upregulation of the IFN-γ transcript in more severely affected dogs in agreement with the was observed in clinically-lesioned skin from dogs with lower inflammatory process observed in normal-looking leishmaniosis, both in mildly and more severely affected skin of mildly affected dogs [7]. In accordance, positive dogs, when compared with healthy skin of non-infected correlations of PD-L1 transcript with clinicopathological dogs. This result agrees with those previously published parameters associated with disease severity, antibody lev- [14]. Although not statistically significant, a higher IFN-γ els and parasite density were detected in normal-looking gene expression was observed in clinically-lesioned skin. Therefore, PD-L1 is suggested as an immunological skin of mildly affected dogs when compared with more marker for disease severity only in normal-looking skin. severely affected dogs. IFN-γ is a protective Th-1 associ- ated cytokine, which increases the leishmanicidal activ- ity of macrophages [2, 11]. Therefore, it is plausible that Conclusions overexpression of this pro-inflammatory cytokine in This study demonstrated, to our knowledge for the first mildly affected cases may be the result of granuloma for- time, different expression profiles of immune genes in mation in papular lesions with consequent lower parasite clinically-lesioned and normal-looking skin from dogs density as previously demonstrated [7, 25]. with leishmaniosis. Moreover, differences among mildly The relative levels of IFN-γ in normal-looking skin and more severely affected dogs were revealed. Clini- from mildly affected dogs were significantly lower than cally-lesioned skin from mildly affected dogs was charac- in normal-looking skin from more diseased dogs and terized by a significant upregulation of TLR2 and IL-10 were associated with disease severity (high specific anti- and downregulation of TLR7 when compared with skin body levels and high parasite density) [13]. The lower from more severely affected dogs. On the other hand, inflammation observed microscopically in normal- normal-looking skin of mildly affected dogs was charac- looking skin from mildly affected dogs may account for terized by a downregulation of TLR7, IFN-γ and PD-L1 this finding [7]. This result is in line with the results of when compared with skin from more severely affected a previous study on normal-looking skin from naturally dogs. Therefore, these immunological conditions might infected dogs demonstrating increased IFN-γ expression favor the maintenance and replication of the parasite in in symptomatic dogs in comparison with asymptomatic the skin of more severely affected dogs leading to disease dogs [15]. Therefore, this pro-inflammatory environ- progression. ment observed in normal-looking skin of more severely affected dogs is not enough to confer protection, as pre- Abbreviations viously suggested [13]. CD4: cluster of differentiation 4; CanL: canine leishmaniosis; ConA: concanava‑ To the best of our knowledge, PD-L1 expression has lin A; Cq: quantification cycle; DNA: deoxyribonucleic acid; ELISA: enzyme- linked immunosorbent assay; EU: ELISA units; IFN-γ: interferon-gamma; IL-10: never been investigated in the skin of dogs with leish- interleukin-10; LSA: Leishmania infantum soluble antigen; qPCR: quantitative maniosis. An increase of PD-L1 in clinically-lesioned and polymerase chain reaction; PD-L1: programmed cell death protein ligand 1; normal-looking skin of dogs with leishmaniosis was dem- PRR: pattern recognition receptor; RNA: ribonucleic acid; Th1: type 1 T helper cells; Th2: type 2 T helper cells; TLR: toll like receptor. onstrated. Therefore, this overexpression may suggest a role of PD-L1 in the immunopathogenesis of CanL. This Acknowledgements protein is related to a decreased T-cell mediated immu- The authors thank Dr. Shazia Hosein for providing the skin samples from the control group (Royal Veterinary College, University of London, UK) and Lorena nity due to T-cell exhaustion via its union with PD-1 on Serrano and Dr. Laura Altet from Vetgenomics and Dr. Anna Mercader from T-cells surface [26]. As suggested in human leishmania- Servei de Genetica, Veterinary Faculty (UAB) for their help and advices on sis, expression of PD-L1 might represent a mechanism molecular work. We are also grateful to Betlem Cabrera from the Departament de Genètica Animal del Centre de Recerca Agrigenòmica (UAB) for her advice that parasites exploit to avoid the host immune response on technical processing of skin samples for RNA extraction. The publication [45]. However, similar expression in clinically-lesioned Ordeix et al. Parasites Vectors (2019) 12:575 Page 12 of 13 of this paper has been sponsored by Bayer Animal Health in the framework of 8. Ordeix L, Solano-Gallego L, Fondevila D, Ferrer L, Fondati A. Papular the 14th CVBD World Forum Symposium. dermatitis due to Leishmania spp. infection in dogs with parasite-specific cellular immune responses. Vet Dermatol. 2005;16:187–91. Authors’ contributions 9. Bottero E, Poggi M, Viglione M. Lesioni papulari indotte da Leishmania LO and LSG designed the study. LO performed RNA and DNA extractions. LO spp. in 8 cani giovani. Veterinaria. 2006;20:33–6. and SM performed the molecular work. PMO performed whole blood assay, 10. Lombardo G, Pennisi M, Lupo T, Chicharro C, Solano-Gallego L. Papular collection of supernatants and serological testing. MB performed logistic dermatitis due to Leishmania infantum infection in seventeen dogs: diag‑ regression analysis and prepared the figures. LO and LSG contributed with nostic features, extent of the infection and treatment outcome. Parasites data analysis and interpretation. LO wrote the manuscript. LSG revised the Vectors. 2014;24:120. manuscript. All authors read and approved the final manuscript. 11. Hosein S, Blake DP, Solano-Gallego L. Insights on adaptive and innate immunity in canine leishmaniosis. Parasitology. 2017;144:95–115. Funding 12. Maia C, Campino L. Cytokine and phenotypic cell profiles of Leishmania This study was supported by Grants AGL2012-32498 and AGL2015-68477 of infantum infection in the dog. J Trop Med. 2012;2012:541571. the Ministerio de Economía y Competitividad (Spain) and the ESVD Research 13. Rodríguez-Cortés A, Carrillo E, Martorell S, Todolí F, Ojeda A, Martínez- grant 2012. Flórez A, et al. Compartmentalized immune response in leishmaniasis: changing patterns throughout the disease. PLoS ONE. 2016;11:e0155224. Availability of data and materials 14. Brachelente C, Muller N, Doherr M, Sattle U, Welle M. Cutaneous leishma‑ Data supporting the conclusions of this article are provided within the article. niasis in naturally infected dogs is associated with a T Helper-2-biased The datasets used and/or analysed during the present study are available from immune response. Vet Pathol. 2005;42:166–75. the corresponding author upon reasonable request. 15. Menezes-Souza D, Correa-Oliveira R, Guerra-Sa R, Giunchetti R, Teixeira- Carvalho A, Martins-Filho O, et al. Cytokine and transcription factor Ethics approval and consent to participate profiles in the skin of dogs naturally infected by Leishmania (Leishmania) No procedure different from those necessary for the diagnosis and staging of chagasi presenting distinct cutaneous parasite density and clinical status. the disease was performed in the dogs (Groups A and B). Dog owners were Vet Parasitol. 2011;177:39–49. informed about the study objectives and the procedure. An informed consent 16. Hosein S, Rodríguez-Cortés A, Blake D, Allenspach K, Alberola J, Solano- was obtained from all owners for participation in the study. The procedure Gallego L. Transcription of Toll-Like Receptors 2, 3, 4 and 9, FoxP3 and regarding healthy Beagle dogs (Group C) followed ethical guidelines in Th17 cytokines in a susceptible experimental model of canine Leishmania accordance with European Union regulations and was approved by the local infantum infection. PLoS ONE. 2015;10:e0140325. Ethical Review Committee from the Royal Veterinary College (University of 17. Pereira-Fonseca DCM, Oliveira-Rovai FM, Rodas LAC, Beloti CAC, Torrecilha London, UK). RBP, Ito PKRK, Avanço SV, Cipriano RS, Utsunomiya YT, Hiramoto RM, Calvo-Bado L, Courtenay O, Machado GF, Lima VMF, Nunes CM. Dog Consent for publication skin parasite load, TLR-2, IL-10 and TNF-α expression and infectiousness. Not applicable. Parasite Immunol. 2017;39(11):1–7. 18. Amorim I, Silva S, Figueiredo M, Moura E, Castro R, Lima T, et al. Toll recep‑ Competing interests tors Type-2 and CR3 expression of canine monocytes and its correlation The authors declare that they have no competing interests. with immunohistochemistry and xenodiagnosis in visceral leishmaniasis. PLoS ONE. 2011;6:e27679. Author details 19. Figueiredo M, Amorim I, Pinto A, Barbosa V, Pinheiro LJ, Deoti B, et al. 1 Department de Medicina i Cirurgia Animals, Facultat de Veterinària, Uni‑ Expression of Toll-like receptors 2 and 9 in cells of dog jejunum and colon versitat Autònoma de Barcelona, Bellaterra, Spain. 2 Hospital Clínic Veterinari, naturally infected with Leishmania infantum. BMC Immunol. 2013;14:22. Universitat Autònoma de Barcelona, Bellaterra, Spain. 20. Melo GD, Silva JES, Grano FG, Homem CG, Machado GF. Compartmental‑ ized gene expression of toll-like receptors 2, 4 and 9 in the brain and Received: 21 April 2019 Accepted: 25 November 2019 peripheral lymphoid organs during canine visceral leishmaniasis. Parasite Immunol. 2014;36:726–31. 21. Vidya M, Kumar V, Sejian V, Bagath M, Krishnan G, Bhatta R. Toll-like recep‑ tors: significance, ligands, signaling pathways, and functions in mammals. Int Rev Immunol. 2018;37:20–36. References 22. Turchetti A, da Costa L, Romão Ede L, Fujiwara R, da Paixão T, Santos R. 1. Solano-Gallego L, Koutinas A, Miro G, Cardoso L, Pennisi M, Ferrer L, et al. Transcription of innate immunity genes and cytokine secretion by canine Directions for the diagnosis, clinical staging, treatment and prevention of macrophages resistant or susceptible to intracellular survival of Leishma- canine leishmaniosis. Vet Parasitol. 2009;165:1–18. nia infantum. Vet Immunol Immunopathol. 2015;163:67–76. 2. Papadogiannakis E, Koutinas A. Cutaneous immune mechanisms in 23. Montserrat-Sangrà S, Alborch L, Ordeix L, Solano-Gallego L. TLR-2 and canine leishmaniosis due to Leishmania infantum. Vet Immunol Immuno‑ TLR-4 transcriptions in unstimulated blood from dogs with leishmaniosis pathol. 2015;163:94–102. due to Leishmania infantum at the time of diagnosis and during follow- 3. Pinelli E, Gebhard D, Mommaas A, van Hoeij M, Langermans J, Ruiten‑ up treatment. Vet Parasitol. 2016;228:172–9. berg E, et al. Infection of a canine macrophage cell line with Leishmania 24. Grano F, Dos S Silva J, Melo G, de Souza M, Lima V, Machado G. Toll-like infantum: determination of nitric oxide production and anti-leishmanial receptors and cytokines in the brain and in spleen of dogs with visceral activity. Vet Parasitol. 2000;92:181–9. leishmaniosis. Vet Parasitol. 2018;253:30–8. 4. Solano-Gallego L, Cardoso L, Pennisi MG, Petersen C, Bourdeau P, Oliva 25. Esteve L, Saz S, Hosein S, Solano-Gallego L. Histopathological findings G, et al. Diagnostic challenges in the era of canine Leishmania infantum and detection of Toll-like receptor 2 in cutaneous lesions of canine leish‑ vaccines. Trends Parasitol. 2017;33:706–17. maniosis. Vet Parasitol. 2015;209:157–63. 5. Ordeix L, Fondati A. Manifestaciones clínicas cutáneas. In: Solano-Gallego 26. Esch K, Juelsgaard R, Martinez P, Jones D, Petersen C. Programmed death L, editor. Leishmaniosis canina. Una revisión actualizada. Zaragoza: Servet; 1- mediated T cell exhaustion during visceral leishmaniasis impairs 2013. p. 69–95. phagocyte function. J Immunol. 2013;191:5542–50. 6. Saridomichelakis M, Koutinas A. Cutaneous involvement in canine 27. Schaut R, Lamb I, Toepp A, Scott B, Mendes-Aguiar C, Coutinho J, et al. leishmaniosis due to Leishmania infantum (syn. L. chagasi). Vet Dermatol. Regulatory IgDhi B cells suppress T Cell function via IL-10 and PD-L1 dur‑ 2014;25:61–71. ing progressive visceral leishmaniasis. J Immunol. 2016;196:4100–9. 7. Ordeix L, Dalmau A, Osso M, Llull J, Montserrat-Sangrà S, Solano-Gallego 28. Schaut R, Grinnage-Pulley T, Esch K, Toepp A, Duthie M, Howard R, et al. L. Histological and parasitological distinctive findings in clinically- Recovery of antigen-specific T cell responses from dogs infected with lesioned and normal-looking skin of dogs with different clinical stages of Leishmania (L.) infantum by use of vaccine associated TLR-agonist adju‑ leishmaniosis. Parasites Vectors. 2017;10:121. vant. Vaccine. 2016;34:5225–34. Ordeix et al. Parasites Vectors (2019) 12:575 Page 13 of 13 29. Chiku V, Silva K, de Almeida B, Venturin G, Leal A, de Martini C, et al. PD-1 39. Silva-Barrios S, Smans M, Duerr C, Qureshi S, Fritz J, Descoteaux A, et al. function in apoptosis of T lymphocytes in canine visceral leishmaniasis. Innate immune B Cell activation by Leishmania donovani exacerbates dis‑ Immunobiology. 2016;221:879–88. ease and mediates hypergammaglobulinemia. Cell Rep. 2016;15:2427–37. 30. Solano-Gallego L, Di Filippo L, Ordeix L, Planellas M, Roura X, Altet L, et al. 40. Fabié A, Mai L, Dagenais-Lussier X, Hammami A, van Grevenynghe J, Early reduction of Leishmania infantum-specific antibodies and blood Stäger S. IRF-5 promotes cell death in CD4 T cells during chronic infec‑ parasitemia during treatment in dogs with moderate or severe disease. tion. Cell Rep. 2018;24:1163–75. Parasites Vectors. 2016;9:235. 41. Day M. The immunopathology of canine vector-borne diseases. Parasites 31. Peters I, Peeters D, Helps C, Day M. Development and application of mul‑ Vectors. 2011;13:48. tiple internal reference (housekeeper) gene assays for accurate normali‑ 42. Belkaid Y, Piccirillo CA, Mendez S, Shevach E. CD4 1 CD25 1 regulatory zation of canine gene expression studies. Vet Immunol Immunopathol. T cells control Leishmania major persistence and immunity. Nature. 2007;117:55–66. 2002;420:502–7. 32. Wood S, Clements D, McEwan N, Nuttall T, Carter S. Reference genes for 43. Nylén S, Maurya R, Eidsmo L, Manandhar K, Sundar S, Sacks D. Splenic canine skin when using quantitative real-time PCR. Vet Immunol Immu‑ accumulation of IL-10 mRNA in T cells distinct from CD4+ CD25+ nopathol. 2008;126:392–5. (Foxp3) regulatory T cells in human visceral leishmaniasis. J Exp Med. 33. Solano-Gallego L, Montserrat-Sangrà S, Ordeix L, Martínez-Orellana P. 2007;204:805–17. Leishmania infantum-specific production of IFN-γ and IL-10 in stimulated 44. Boggiatto P, Ramer-Tait A, Metz K, Kramer E, Gibson-Corley K, Mullin K, blood from dogs with clinical leishmaniosis. Parasites Vectors. 2016;9:317. et al. Immunologic indicators of clinical progression during canine Leish- 34. Campos M, Lima L, de Lima A, Vasconselos Dos Santos T, Ramos P, Gomes mania infantum infection. Clin Vaccine Immunol. 2010;17:267–73. C, et al. Toll-like receptors 2, 4, and 9 expressions over the entire clinical 45. Barroso D, Falcão S, da Motta J, Sevilha Dos Santos L, Takano G, Gomes and immunopathological spectrum of American cutaneous leishmaniasis C, et al. PD-L1 may mediate T-cell exhaustion in a case of early diffuse due to Leishmania (V.) braziliensis and Leishmania (L.) amazonensis. PLoS leishmaniasis caused by Leishmania (L.) amazonensis. Front Immunol. ONE. 2018;13:e0194383. 2018;9:1021. 35. Montserrat-Sangrà S, Ordeix L, Martínez-Orellana P, Solano-Gallego L. 46. Silva K, Melo L, Perosso J, Oliveira B, Santos P, Eugênio FR, et al. CD95 (FAS) Parasite specific antibody levels, interferon-γ and TLR2 and TLR4 tran‑ and CD178 (FASL) induce the apoptosis of CD4+ and CD8+ cells isolated scripts in blood from dogs with different clinical stages of leishmaniosis. from the peripheral blood and spleen of dogs naturally infected with Vet Sci. 2018;5:31. Leishmania spp. Vet Parasitol. 2013;197:470–6. 36. Ajdary S, Ghamilouie M, Alimohammadian M. Toll-like receptor 4 47. Perosso J, Silva K, Ferreira S, Avanço S, dos Santos P, Eugênio FR, et al. polymorphisms predispose to cutaneous leishmaniasis. Microbes Infect. Alteration of sFAS and sFAS ligand expression during canine visceral 2011;13:226–31. leishmaniosis. Vet Parasitol. 2014;205:417–23. 37. Ziakas P, Prodromou M, El Khoury J, Zintzaras E, Mylonakis E. The role of TLR4 896 A>G and 1196 C>T in susceptibility to infections: a review and meta-analysis of genetic association studies. PLoS ONE. 2013;8:e81047. Publisher’s Note 38. Okui Y, Kano R, Maruyama H, Hasegawa A. Cloning of canine Toll-like Springer Nature remains neutral with regard to jurisdictional claims in pub‑ receptor 7 gene and its expression in dog tissues. Vet Immunol Immuno‑ lished maps and institutional affiliations. pathol. 2008;121:156–60. Ready to submit your research ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions