Priolo et al. Parasites Vectors (2019) 12:133 https://doi.org/10.1186/s13071-019-3386-y Parasites & Vectors RESEARCH Open Access Leishmania infantum‑specific IFN‑γ production in stimulated blood from cats living in areas where canine leishmaniosis is endemic Vito Priolo1, Pamela Martínez‑Orellana2, Maria Grazia Pennisi1*†, Marisa Masucci1, David Prandi2, Dorotea Ippolito1, Federica Bruno3, Germano Castelli3 and Laia Solano‑Gallego2† Abstract Background: Feline leishmaniosis caused by Leishmania infantum is considered a rare disease in endemic areas, whereas subclinical infections are common. Immune response plays a key role in driving the course of L. infantum infection in other host species; however, the feline cell-mediated immune response to L. infantum infection has not yet been investigated. The aim of this study was to determine the cell-mediated immune response specific to L. infan- tum by means of interferon (IFN)-γ release in whole blood assay from cats living in endemic areas (66 in Sicily and 113 in Catalonia) and to compare with antibody levels to L. infantum [enzyme-linked immunosorbent assay (ELISA) and immunofluorescence antibody test (IFAT)], blood parasite load and retroviral infections. Results: Most cats (n = 140) were L. infantum antibody negative and only 22% (n = 39) were positive. Only 9 and 2% of tested cats had a feline immunodeficency virus (FIV) infection or a feline leukemia virus (FeLV) infection, respec‑ tively. Thirty-two cats out of 179 (18%) produced IFN-γ after stimulation with L. infantum soluble antigen (LSA) while the majority of cats (93%) produced IFN-γ after stimulation with concanavalin A (ConA). Six LSA-IFN-γ-producer cats were seropositive (three to ELISA and five to IFAT) but they were polymerase chain reaction (PCR) negative, while only one cat was antibody- and PCR-positive. Significant positive correlations were found between IFN-γ concentrations after stimulation with LSA and ConA, and between serology and PCR testing. No association was found between FIV status and LSA or ConA-IFN-γ production. Combining PCR, serology and specific IFN-γ concentration results, we found that 36% of cats studied were exposed to L. infantum. Conclusions: As expected, cats from endemic areas produce IFN-γ after ex vivo blood stimulation with LSA and therefore are able to activate a cell-mediated adaptive immune response against the parasite that is variably associ‑ ated with antibody or blood PCR positivity. The association of this assay to serological and molecular tests provides a better estimate of cat exposure to L. infantum. Keywords: Adaptive immunity, Feline, Humoral immunity, IFN-γ release whole blood assay, Leishmania infantum, PCR, Retroviral infections Background [1–3]. Dogs are considered the main reservoir of L. infan- Leishmaniosis is a vector-borne disease of humans and tum but there is clear evidence that some wild and synan- animals caused in Europe by Leishmania infantum and thropic mammals and domestic cats are able to infect transmitted by female sand flies of the genus Phlebotomus sand flies and they play a variable role in a reservoir sys- tem according to local and ecological peculiarities [3–5]. Feline leishmaniosis (FeL) due to L. infantum infec- *Correspondence: pennipet@unime.it † tion was described, for the first time, in 1912 by Sergent Maria Grazia Pennisi and Laia Solano-Gallego contributed equally to this work et al. [6] and since then it has been globally reported in 1 Dipartimento di Scienze Veterinarie, Università di Messina, Messina, Italy endemic areas [5, 7]. While subclinical feline infections 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://creat​iveco​mmons​.org/ publi​cdoma​in/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Priolo et al. Parasites Vectors (2019) 12:133 Page 2 of 9 are common in areas where canine leishmaniosis (CanL) at the Fundació Hospital Clínic Veterinari (Bellaterra, is endemic, clinical illness due to FeL is rare [7, 8]. Clini- Barcelona), Hospital Clinic Xinesca (Vilassar de Mar, cal disease is frequently associated with possible impaired Barcelona) and Vetamic Hospital Veterinari Cambrils immunocompetence, as in case of retroviral co-infec- (Cambrils, Tarragona). Sicilian samples were collected at tions, immunosuppressive therapy, or malignant neopla- Ospedale Veterinario Didattico (Università degli Studi di sia [7, 8]. The immune response plays a crucial role in the Messina, Dipartimento di Scienze Veterinarie, Messina) control of Leishmania infection. Albeit some differences and at Ambulatorio Veterinario Santa Lucia (Lipari, according to the host species, T cells modulate and ori- Messina). Inclusion criteria for enrollment included the ent, through cytokine production, macrophage reaction exposure to at least one sand-fly season and no treatment to the parasite [9–11]. Antigen processing cells (dendritic with repellent ectoparasiticides (i.e. pyrethroid prod- cells and macrophages) present L. infantum antigens to ucts). Sex, age class, breed, clinical status, lifestyle of pet ­CD4+ T cells that modulate the type of immune response cats (indoors, outdoors), and number of stray cats at each [12]. A T helper 1 (Th1) oriented immune response, asso- sampling site are summarized in Table 1. Age was classi- ciated with production of interferon-gamma (IFN-γ), fied as follows: young (6–18 months); adult (between 19 interleukin (IL)-2 and tumor necrosis factor alpha (TNF- and 96 months); and old (> 96 months). The clinical sta- α), stimulates phagocytosis by macrophages, their pro- tus of cats was defined as “sick” or “apparently healthy” duction of nitric oxide and reactive oxygen intermediate based on data available from history and physical exami- and consequent phagocyte-based parasite intracellular nation. Sick cats were considered when clinical signs elimination [11, 13]. Conversely, in dogs susceptibility compatible with FeL (i.e. lymph node enlargement, skin, to infection and disease progression is mediated pre- mucosal or eye lesions, stomatitis, weight loss, chronic dominantly by a non-protective T helper 2 (Th2) immune kidney disease, anaemia) were found as previously response and the production of cytokines such as IL-4, described [7]. Apparently healthy cats were considered IL-10, IL-13 and transforming growth factor beta (TGF- when no clinical signs were present with the exception β) which are associated with downregulation of the cel- of four cats admitted because of trauma that were also lular immune response, a high level of antibodies, and L. included in this group. infantum dissemination [9–11]. One milliliter of blood, aseptically put into a heparin Different feline innate and adaptive immune responses tube, was used for whole blood assay within 24 h after might account for the observed lower prevalence of L. blood sampling. EDTA blood and blood serum were also infantum infection as well as clinical leishmaniosis in obtained and immediately aliquoted and stored at -20 °C cats as compared to dogs [7, 14]. It is well recognized that until processed for DNA extraction and serological inves- cats appear to be less frequently affected by arthropod- tigations, respectively. borne diseases when compared to dogs, although no important differences are known between the canine and Feline IFN‑γ release whole blood assay feline immune systems [15]. Interestingly, to the best of Whole blood assays were performed as previously our knowledge, no studies have so far evaluated L. infan- described in dogs [10] with some minor modifications to tum-specific cell-mediated immunity in cats. However, adapt the technique to the smaller volume of blood avail- detection of IFN-γ in antigen stimulated whole blood or able. Heparinized blood was diluted at a ratio of 1:10 with peripheral blood mononuclear cells (PBMCs) was used in medium and after five days the supernatant was collected cats to evaluate cellular immune response mechanisms to from unstimulated cultured cells. Two aliquots of diluted other pathogens (e.g. feline coronavirus or Toxoplasma blood were respectively cultured with Concanavalin- gondii) [16–18]. A (ConA) or L. infantum soluble antigen (LSA) to sub- The main aim of this study was to determine the L. sequently recover supernatants from stimulated blood infantum-specific cellular immune response in cats by cells. specific DuoSet® ELISA (Development Sistems R&D™, means of the evaluation of IFN-γ production in stimu- Feline IFN-γ concentrations were determined using a lated blood from cats living in endemic areas of CanL (Catalonia and Sicily) and correlate it with L. infantum Abingdon, UK) according to the manufacturer’s instruc- antibody levels, blood parasitemia, and retroviral status. tions with some modifications. The standard curve for IFN-γ started with 4000 pg/ml and serial 2-fold dilu- Methods tions were made until a concentration of 31.25 pg/ml was Study areas, cats and sampling reached. Duplicates of supernatants obtained from whole Cats were sampled from March 2016 to April 2017 in blood culture were tested in the ELISA plates. The opti- two Mediterranean endemic areas of CanL: Sicily (Italy) cal density was measured with an ELISA reader (Anthos and Catalonia (Spain). Catalan samples were collected 2020, Cambridge, UK) at wavelength of 450 nm. The Priolo et al. Parasites Vectors (2019) 12:133 Table 1 Age class, sex, breed, lifestyle and clinical status of cats studied according to their geographical distribution Provenience Age class Sex Breed Categorya Clinical status (N; %) Young Adult Old Male Female DSH Pure breed Indoor Outdoor Stray Apparently healthy Sick n (%) n (%) n (%) n (%) n (%) n (%) n (%) n (%) n (%) n (%) n (%) n (%) Sicily Messina City (N = 24; 36) 6 (25) 18 (75) 0 (0) 9 (38) 15 (62) 24 (100) 3 (12) 3 (12) 5 (21) 16 (67) 11 (46) 13 (54) Aeolian Islands (N = 42; 64) 9 (21) 26 (62) 7 (17) 19 (45) 23 (55) 41 (98) 1 (2) 1 (2) 40 (96) 1 (2) 19 (45) 23 (55) Total (N = 66; 37) 15 (23) 44 (67) 7 (10) 28 (42) 38 (58) 65 (99) 1 (1) 4 (6) 45 (68) 17 (26) 30 (46) 36 (54) Catalonia Vilassar de Mar (N = 22;19) 11 (50) 7 (32) 4 (18) 13 (59) 9 (41) 17 (77) 5 (23) 7 (32) 11 (50) 4 (18) 17 (77) 5 (23) Cambrils (N = 45; 40) 3 (7) 40 (89) 2 (4) 18 (40) 27 (60) 42 (93) 3 (7) 0 (0) 0 (0) 45 (100) 40 (89) 5 (11) Bellaterra (N = 46; 41) 6 (13) 39 (85) 1 (2) 25 (54) 21 (46) 44 (96) 2 (4) 0 (0) 15 (33) 31 (67) 32 (70) 14 (30) Total (N = 113; 63) 20 (18) 86 (76) 7 (6) 56 (49) 57 (51) 103 (91) 10 (9) 7 (6) 26 (23) 80 (71) 89 (79) 24 (21) All cats (N = 179; 100) 35 (19) 130 (73) 14 (8) 84 (47) 95 (53) 168 (94) 11 (6) 11 (6) 71 (40) 97 (54) 119 (66) 60 (34) a Indoor and outdoor columns concern the lifestyle of pet cats Abbreviation: DSH, domestic short-hair Page 3 of 9 Priolo et al. Parasites Vectors (2019) 12:133 Page 4 of 9 standard curve was calculated using a computer generat- antigen by INgezim-FeLV DAS (Ingenasa, Madrid, Spain) ing four parameters logistic curve-fit with the program and only 16 feline sera for anti-FIV antibodies by INge- MyAssays (http://www.myass​ays.com/). Cats were classi- zim-FIV (Ingenasa). fied as IFN-γ producers (IFNγ-p) when the L. infantum- specific IFN-γ concentration, after subtracting the value Blood DNA extraction and Leishmania real‑time PCR obtained with supernatant of unstimulated cell cultures, Total DNA was extracted from EDTA blood using the was higher than the last detectable dilution of the stand- DNA Gene extraction kit (Sigma Aldrich, Saint Louis, ard curve (31.25 pg/ml). Similarly, cats were classified as MO, USA) following the manufacturer’s instructions IFN-γ non-producers (IFNγ-np) when the L. infantum- with some modifications. Forty microliters of protein- specific IFN-γ concentration, after subtracting the value ase K solution were added to all samples. Four hundred obtained with supernatant of unstimulated cell cultures, microliters of whole blood were used for all the samples. was less than 31.25 pg/ml or at undetectable levels. The Blood from a clinically healthy non-infected cat was same cat classification was made for IFN-γ production used as a control for DNA contamination in every DNA obtained by ConA stimulation of whole blood cultures. extraction performed. The real-time polymerase chain reaction (RT-PCR) was carried out in a CFX96 Real-time System (Bio-Rad Laboratories s.r.l., Hercules, CA, USA) Leishmania infantum antibody detection using TaqMan Master Mix (Applied Biosystems by Ther- IFAT moFisher, Waltham, MA, USA) and performed as previ- Anti-L. infantum IgG antibodies were detected using the ously described [22]. L. infantum (strain MHOM/IT/80/IPT1) antigen pro- duced by C.Re.Na.L. (Centro di Referenza Nazionale per Statistical analysis la Leishmaniosi, Palermo, Italy). Fluoresceinated anti-cat A minimum sample size of 173 cats was calculated for cat immunoglobulin G (IgG) antibody (working anti-feline enrollment, based on the prevalence of feline L. infantum IgG (H+L)-FITC, Fuller Laboratories, Fullertone, CA, infection in areas under study (Catalonia and Sicily) [14, USA) was used. The manufacturer’s protocol was fol- 19, 20, 23], and on the assumptions of 99% confidence lowed, and the end point titer of positive samples was level and 5% precision [24]. determined preparing serial 2-fold dilutions of serum Fisher’s exact test was used to compare groups defined starting from 1:20. The cut-off value for positivity was set by categorical variables. Results of tests for L. infantum at 1:80 [19, 20]. Fluorescence microscope readings were exposure of cats did not pass the D’Agostino-Pearson made by a unique operator (MM). normality test. Accordingly, the Mann-Whitney U-test was used to compare unmatched continuous data and In‑house ELISA the Wilcoxon signed-rank test to compare paired con- ELISA was performed as previously described [20, 21]. tinuous variables. Spearman’s correlation coefficient was All plates included the serum from a sick cat from Cyprus calculated to evaluate relationships between ELISA, IFAT with a confirmed infection with L. infantum as positive and PCR results and between levels of IFN-γ, anti-Leish- control, and the serum of a cat from an area where leish- mania antibodies, and L. infantum DNA in blood of cats maniosis was not endemic as a negative control. All sam- studied. P-values<0.05 were considered significant. The ples were run in duplicate. The cut-off was established at D’Agostino-Pearson normality test and Fisher’s exact test 12.3 ELISA units (EU) (mean ± 3 standard deviations of were performed using Prism 7 for Mac IOS and all other sera from 81 cats from the UK, a non-endemic area). tests were performed using SPSS software v.17.0 for Win- dows. Finally, Cohen’s kappa coefficient was measured Detection of anti‑FIV antibodies and FeLV p27 antigen analyzing results of ELISA and IFAT with blood PCR Due to insufficient amounts of serum, only 149 cats out (https​://idost​atist​ics.com/cohen​-kappa​-free-calcu​lator​). of 179 were tested for feline immunodeficiency virus (FIV) and 171 for feline leukemia virus (FeLV) infections. Results A total of 133 feline sera were tested for the detection Cats of FeLV p27 antigen and anti-FIV antibodies by a rapid Sex, breed, age class, lifestyle, origin (region and munici- ELISA (SNAP Combo Plus FeLV antigen and FIV anti- pality) and clinical status of 179 cats included in the study body test, Idexx Laboratories, Westbrook, ME, USA), are summarized in Table 1. The median of age of cats was according to the manufacturer’s protocol. Because of 24 months [25–75 percentile (25–75) = 18–48 months; serum paucity, other cats were tested by a commer- interquartile range (IQR) = 30 months]. There were no cial ELISA: 38 serum samples were tested for FeLV p27 differences between the two regions related to the vari- ables sex, breed, age class and lifestyle of cats. Conversely, Priolo et al. Parasites Vectors (2019) 12:133 Page 5 of 9 cats from Sicily were more frequently found sick com- than the result observed in Catalan cats [median: 3.7 EU; pared to Catalan cats (Fisher’s exact test, P < 0.0001). (25–75) = 1.9–6.4 EU] (Mann-Whitney U-test, Z = -249, P < 0.0001). Leishmania infantum serological tests Results of serological tests are shown in Table 2 and Leishmania real‑time PCR Fig. 1. A slight positive correlation (Spearman’s cor- PCR results based on cat provenience and clinical status relation coefficient, rs = 0.342, P = 0.0001) was found are summarized in Table 2. Only 9 cats (5%) were Leish- between ELISA and IFAT. However, only 11 of the 39 cats mania-PCR positive (Fig. 1) and the median parasite load positive to IFAT and/or ELISA were positive to both tests was 60 L. infantum amastigotes/ml [(25–75) = 10–128.5]. (28%) and an agreement of 78.1% was found when these Leishmania-PCR positive cats were more frequently pos- tests were compared (Cohen’s kappa coefficient = 0.18, itive by serology (89% by IFAT and 78% by ELISA) than i.e. slight agreement beyond chance). No differences were Leishmania-PCR negative cats (13% by IFAT and 7% by found regarding seropositivity and variables studied. ELISA) (Fisher’s exact test, P<0.0001). In addition, anti- body level measured by IFAT (Mann-Whitney U-test, IFAT Z = 86.5, P < 0.0001) and ELISA (Mann-Whitney-U test, IFAT results based on cat provenience and clinical status Z = 219, P < 0.0001) of Leishmania-PCR positive cats are summarized in Table 2. The median IFAT titer was 80 was significantly higher when compared to Leishmania- [(25–75) = 80–320]. No differences of IFAT titers were PCR negative cats. Moreover, a positive correlation was found between Sicilian and Catalonian cats. However, found between PCR and ELISA (Spearman’s correlation Catalonian cats from Bellaterra were more frequently coefficient, rs = 0.272, P = 0.0001) with an agreement of found positive to IFAT than those from Vilassar de Mar 86.1% (Cohen’s kappa coefficient = 0.25, i.e. fair agree- (Fisher’s exact test, P = 0.0453). ment) as well as a positive correlation between PCR and IFAT (Spearman’s correlation coefficient, rs = 0.347, ELISA P = 0.0001) with an agreement of 80.1% (Cohen’s kappa ELISA results based on cat provenience and clinical coefficient = 0.18, i.e. slight agreement). No other differ- status are summarized in Table 2. The median value of ences were found in positive PCR prevalence according positive samples was 19 EU [(25–75) = 13.5–107.3 EU]. to variables under study. Old cats (4/14, 29%) were more frequently positive than young cats (1/35, 3%) (Fisherʼs exact test, P = 0.0194). Anti‑FIV antibodies and FeLV p27 antigen Sicilian cats were statistically more frequently found Thirteen cats (9%) were antibody-positive to FIV and positive to ELISA than Catalan cats (Fisher’s exact test, 4 (2%) were antigen-positive to FeLV (Table 2). FIV P = 0.0282) and the ELISA median value for Sicilian cats and FeLV co-infections were never detected. All FeLV- was significantly higher [6.1 EU; (25–75) = 1.2–9.5 EU] positive cats were from Catalonia while there was no Table 2 Number and percentage of cats positive to L. infantum and retroviruses according to their geographical distribution and clinical status Provenience or clinical status Serology Blood PCR IFN-γ assay Retroviruses (N; %) Total ELISA IFAT LSA ConA FIV FeLV Sicily Messina City (N = 24; 36) 5 (21) 4 (17) 3 (12) 1 (4) 6 (25) 24 (100) 2 (8) 0 (0) Aeolian Islands (N = 42; 64) 9 (21) 7 (17) 6 (14) 2 (5) 5 (12) 39 (93) 4 (9) 0 (0) Total (N = 66; 37) 14 (21) 11 (17) 9 (14) 3 (4) 11 (17) 63 (95) 6 (9) 0 (0) Catalonia Vilassar de Mar (N = 22;19) 1 (4) 0 (0) 1 (4) 1 (4) 2 (9) 22 (100) 1 (5) 0 (0) Cambrils (N = 45; 40) 8 (18) 2 (4) 6 (13) 1 (2) 5 (11) 40 (89) 4 (12) 1 (2) Bellaterra (N = 46; 41) 16 (35) 6 (13) 15 (33) 4 (9) 14 (30) 41 (89) 2 (6) 3 (7) Total (N = 113; 63) 25 (22) 8 (7) 22 (19) 6 (5) 21 (19) 103 (91) 7 (8) 4 (4) Apparently healthy (N = 119; 66) 21 (18) 6 (5) 19 (16) 3 (2) 21 (18) 108 (91) 8 (7) 2 (2) Sick (N = 60, 34) 18 (30) 13 (22) 12 (20) 6 (10) 11 (18) 58 (97) 5 (8) 2 (3) All cats (N = 179; 100) 39 (22) 19 (11) 31 (17) 9 (5) 32 (18) 166 (93) 13 (9) 4 (2) Abbreviations: FIV, feline immunodeficiency virus; FeLV, feline leukemia virus; LSA, Leishmania soluble antigen; ConA, concanavalin A Priolo et al. Parasites Vectors (2019) 12:133 Page 6 of 9 Fig. 1 Percentage of cats positive to L. infantum according respectively to: all specific tests performed, serology (IFAT and/or ELISA) and blood PCR, serology (IFAT and/or ELISA), IFAT, ELISA, blood PCR, and LSA IFN-γ assay difference in FIV prevalence between cats enrolled in significantly lower than concentration obtained with Sicily (9%) and in Catalonia (8%) (Table 2). Moreover, ConA stimulation [1115 pg/ml (25–75) = 199.9–2931 pg/ there were no statistical differences in prevalence of anti- ml] (Wilcoxon signed-rank test, Z = -11.108, P < 0.0001). L. infantum antibodies or Leishmania DNA positivity in The unique IFNγ-p cat positive to Leishmania-PCR (125 cats positive to FIV or FeLV compared to the negatives. L. infantum amastigotes/ml) tested high positive by both Similarly, no statistical difference was found for the lev- IFAT (titer 40960) and ELISA (> 346.3 EU). This cat was els of anti-L. infantum antibodies or L. infantum parasite adult, FeLV-positive, suffered from squamous cell carci- load between FIV- or FeLV-positive cats and the negative noma and Leishmania amastigotes were detected at cyto- ones. Interestingly, four FIV- and two FeLV-positive cats logical evaluation of the neoplastic cutaneous lesion. were also positive to one or more L. infantum diagnostic Sick and apparently healthy cats had the same preva- tests. Only one FeLV- and two FIV-positive cats produced lences of IFNγ-p individuals (Table 2); however, sick IFN-γ after LSA stimulation. Moreover, two of them were IFNγ-p cats had a significantly higher level of LSA IFN-γ positive to both serological tests and blood PCR. Con- (Mann-Whitney U-test, Z = 47, P = 0.0056) (Table 3). versely, 107 cats were L. infantum and retrovirus nega- The prevalence of cats producing IFN-γ after stimula- tive and 21 produced IFN-γ after LSA stimulation. There tion with ConA was above 90% in both groups (Table 2); were no statistical differences in IFN-γ productions (after however, among 64 cats exposed to L. infantum (i.e. posi- LSA or ConA stimulations) between FIV- or FeLV-posi- tive to at least one specific test), sick individuals (n = 25) tive and negative cats. had a significantly higher median concentration [2159 pg/ml; (25–75) = 965.3–4508 pg/ml] than apparently healthy cats (n = 39) [1712.5 pg/ml; (25–75) = 87.1–2946 IFN‑γ production pg/ml] (Mann-Whitney U-test, Z = 339, P = 0.0117). The frequency of cats producing IFN-γ after stimulation The concentration of IFN-γ produced after stimulation with LSA or ConA is summarized in Table 2 according with ConA was significantly higher in the IFNγ-p group to their provenience and clinical status. A higher num- (Mann-Whitney U-test, Z = 1473, P = 0.0008) (Table 3) ber of cats produced IFN-γ after stimulation with ConA and IFN-γ concentrations after ConA or LSA stimula- (166/179, 93%) than with LSA (32/179, 18%) (Fisher’s tions were positively correlated (Spearman’s correlation exact test, P < 0.0001) (Table 2). Levels of IFN-γ accord- coefficient, rs = 0.264, P = 0.0001). Additionally, ELISA ing to serological and PCR results, and clinical status are antibody levels were positively correlated with IFN-γ summarized in Table 3. The median of the IFN-γ con- centration of LSA-IFN-γ producer cats (IFNγ-p) was Priolo et al. Parasites Vectors (2019) 12:133 Page 7 of 9 Table 3 Concentrations of IFN-γ and level of positivity to L. infantum tests according to results of tests and the clinical status in IFN-γ producer and non-producer cats L. infantum tests or clinical status No. of cats (%) LSA IFN-γ ConA IFN-γ ELISA IFAT PCR (pg/ml)a (pg/ml)a (EU)a (titer)a (amastigotes/ml)a IFN-γ producer cats Serology- and PCR-negative 25 (14) 174 (64.2; 659.9) 2280 (500.7; 4068) 3.4 (2.1; 6.8) 20 (0; 40) 0 (0; 0) Serology-positive and PCR-negative 6 (3) 193.9 (103.4; 685.3) 4456 (1756; 5734) 8.9 (1.4; 16.1) 80 (70;140) 0 (0; 0) Serology- and PCR-positive 1 (1) 276.3 2083.9 > 346.3 40960 125 Apparently healthy 21 (18) 116 (63.7; 238.9) 2097 (205.7; 3976) 3.2 (1.4; 6.1) 20 (0; 40) 0 (0; 0) Sick 11 (18) 567.1 (276.3; 1747) 3509 (2326; 4967) 6.9 (4.1; 12.7) 20 (0; 80) 0 (0; 0) Total 32 (18) 203.5 (65.9; 586.7) 2360 (979.1; 4297) 3.7 (2.1; 8.1) 20 (0; 40) 0 (0; 0) IFN-γ non-producer cats Serology- and PCR-negative 115 (64) 0 (0; 0) 646.1 (137.9; 2279) 4.2 (2.4; 6.3) 20 (0; 40) 0 (0; 0) Serology-positive and PCR-negative 24 (13) 0 (0; 0) 803.2 (176; 2553) 7.65 (4.7; 14.2) 80 (50; 160) 0 (0; 0) Serology- and PCR-positive 8 (4) 0 (0; 0) 1334 (303.8; 2370) 25.5 (9.7; 263.5) 320 (80; 1120) 60 (7.5; 170) Apparently healthy 98 (82) 0 (0; 0) 587.5 (127.4; 2449) 4.3 (2.6; 6.7) 40 (0; 40) 0 (0; 0) Sick 49 (82) 0 (0; 0) 1019 (204.7; 2219) 5.8 (3.0; 8.9) 20 (10; 40) 0 (0; 0) Total 147 (82) 0 (0; 0) 742.5 (150.7; 2407) 4.5 (2.8; 7.2) 40 (0; 40) 0 (0; 0) a Median values (25th and 75th percentiles) concentrations from ConA stimulated cultures (Spear- and IFNγ-np cats positive to serological or blood PCR man’s correlation coefficient, rs = 0.209, P = 0.0001). tests. However, sick IFNγ-p cats produced a significantly higher level of IFN-γ. Moreover, the unique cat with clin- Overall L. infantum frequency of infection ical FeL confirmed by cytology was IFNγ-p and, despite An overall L. infantum prevalence of 36% calculated by possible immunosuppression due to FeLV infection and PCR, serology and LSA IFN-γ assay was obtained in the to neoplasia, this cat reached an IFN-γ level around the population studied (Fig. 1) and no differences were found median value. In the present study, retroviral infection between Catalonia (35%) and Sicily (36%), the sites of per se did not affect adaptive immune response of cats to each region studied, or between apparently healthy (33%) L. infantum, but a limitation was the very low number of and sick cats (42%). FIV (n = 13) or FeLV (n = 4) infected enrolled cats. Dogs with mild or moderate CanL produce IFN-γ and Discussion the lack of production is restricted to severe disease [10, To our knowledge, this study demonstrated for the first 28]. Prospective clinical studies with cats with confirmed time that cats naturally exposed to L. infantum infection clinical leishmaniosis are needed to further evaluate if produced IFN-γ after ex vivo whole blood stimulation IFN-γ is as possible marker for staging the severity of dis- with L. infantum antigens, as occurs in dogs, humans and ease and evaluating efficacy of therapy as documented in laboratory animals [10, 25, 26]. dogs [10, 28]. Parasite-specific IFN-γ production was found in 18% IFN-γ production after stimulation with ConA was of enrolled cats and it was associated with antibody pro- obtained in almost all cats and it was not associated with duction in only seven of the 32 IFNγ-p cats of this study. any variable, including clinical status, retroviral infec- Both types of adaptive immune responses were therefore tions or exposure to L. infantum. These findings are sim- variably combined in single cats and a wide immunologi- ilar to data from studies of CanL [10, 28]. As expected, cal spectrum may exist also in cats as already reported in concentrations of ConA-IFN-γ were significantly higher dogs and humans [10, 27]. than LSA-IFN-γ and they were positively correlated. Pro- The present study did not aim to evaluate the para- duction of ConA-IFN-γ suggests non-specific T cell acti- site T cell mediated immunity of clinical cases of FeL vation. Surprisingly, when considering cats exposed to but based on data available from physical examination, L. infantum (i.e. positive to at least one the specific tests the frequency of IFNγ-p individuals was not different in performed), sick individuals produced a significantly apparently healthy and sick cats. Similarly, there was no higher level of ConA compared to apparently healthy difference in the prevalence of sick cats between IFNγ-p individuals. This finding is unusual, and it is difficult to Priolo et al. Parasites Vectors (2019) 12:133 Page 8 of 9 interpret due to the limitations of a clinical evaluation Conclusions only based on physical examination and clinical history. As expected, cats from endemic areas produce IFN-γ In dogs with CanL, ConA-IFN-γ concentrations do not after ex vivo blood stimulation with LSA and therefore vary significantly in different LeishVet clinical stages, are able to activate a cell-mediated adaptive immune unless in the severe stage of disease (LeishVet stage IV) response against the parasite that is variably associated when a significantly lower concentration is observed [10]. with antibody or blood PCR positivity. The association The prevalence of anti-L. infantum antibodies obtained of this assay to serological and molecular tests provides a in this study by combination of both IFAT and ELISA better estimate of cat exposure to L. infantum. results was similar in the two regions with an approxi- mate value of 20%. It is not easy, therefore, to explain the Abbreviations higher frequency and level of positivity of ELISA in Sicil- CanL: canine leishmaniosis; ConA: concanavalin A; ConA-IFNγ: IFN-γ pro‑ ian cats but these cats were more frequently found sick duced in whole blood assay after stimulation with ConA; DTH: delayed type compared to Catalan cats (Table 2). Previous studies con- hypersensitivity reaction; ELISA: enzyme-linked immunosorbent assay; FeL: feline leishmaniosis; FeLV: feline leukemia virus; FIV: feline immunodeficiency ducted in Sicily were based on IFAT and, when the same virus; IFAT: indirect fluorescence antibody test; IFN-γ: interferon-γ; IFNγ-np: cut-off was used, anti-L. infantum antibody prevalence cats IFN-γ non-producers in whole blood assay after stimulation with LSA; ranged between 6.6–29% [19, 29–32]. Conversely, only IFNγ-p: cats IFN-γ producers in whole blood assay after stimulation with LSA; IgG: immunoglobulin G; LPA: lymphocyte proliferation assay; LSA: Leishmania two studies evaluated anti-L. infantum antibody preva- soluble antigen; PBMCs: peripheral blood mononuclear cells; RT-PCR: real-time lence in Catalonia; they both used ELISA and antibody polymerase chain reaction. prevalence was 1.7 and 5.3%, respectively [14, 33]. How- Acknowledgements ever, as already reported, we found a positive correlation The authors are grateful to Sara Montserrat-Sangrà, Anna Morist (Departament between ELISA and IFAT [21]. de Medicina i Cirurgia Animals, Facultat de Veterinària, Universitat Autònoma The parasite load of cats was measured by blood PCR, de Barcelona, Bellaterra, Spain), Angela Burrascano, Vittorio Fisichella and Maria Foti (Dipartimento di Scienze Veterinarie, Università degli Studi di thus some positive cats were possibly missed because Messina) for technical collaboration. The authors also thank Laura Gulotta blood is not the most sensitive tissue for the detection of (Ambulatorio Veterinario Santa Lucia, Lipari), Ares Burballa (Center veterinari L. infantum DNA in dogs and likewise with cats [7, 34]. Xinesca), Simona Foti, Anna Marti (Vetamic Hospital Veterinari Cambrils) and Laura Ordeix (Hospital Clinic Veterinari), for the collection of blood samples. However, all PCR-positive cats were also antibody-posi- Publication of this paper has been sponsored by Bayer Animal Health in the tive and the level of L. infantum DNA in blood was posi- framework of the 14th CVBD World Forum Symposium. tively correlated with the antibody level of cats. This is Funding not a common finding in cats from L. infantum endemic Not applicable. areas but, compared to other studies, we used two sero- logical techniques and therefore the sensitivity of serol- Availability of data and materials Data supporting the conclusions of this article are included within the article. ogy was increased [19, 20, 35]. Raw data are available from the corresponding author upon reasonable In this study, most IFNγ-p cats were negative to L. request. infantum antibody and DNA detection in blood (25/32, Authors’ contributions 78%). This finding means that the combination of sero- LSG and MGP conceived the research study. VP collected samples and carried logical and molecular tests with L. infantum-specific out laboratory work with the supervision and/or collaboration of MM, MGP, IFN-γ evaluation offered a more accurate overall esti- LSG, PMO, DI, DP, FB and GC. VP and PMO performed statistical analysis. VP and MGP wrote the first draft of the manuscript. PMO, LSG and MGP revised the mation of the exposure to L. infantum of cats under manuscript. All authors read and approved the final manuscript. study. In fact, the prevalence of L. infantum was 5% for blood PCR and 22% for antibody detection but the Ethics approval and consent to participate Cats were sampled during the annual health check, elective surgery, or combination of results obtained by serological and trap-neuter-release programs and always after the signature of an informed molecular testing did not increase the percentage of consent by the owner. Residual samples of blood were mostly used in this positivity because all PCR positive cats were also posi- study; however, for cats sampled in Sicily an ethical committee evaluated and approved the protocol (code: 015/2017). tive for L. infantum antibodies (Fig. 1). However, when also considering positivity for L. infantum-specific Consent for publication IFN-γ production, the overall L. infantum prevalence Not applicable. raised to 36%. These data confirm that a consider- Competing interests able percentage of the studied cats had contact with L. The authors declare that they have no competing interests. infantum and that detection of cell-mediated immune response by measuring specific IFN-γ production pro- Publisher’s Note vides a better estimation of exposure of cats to L. infan- Springer Nature remains neutral with regard to jurisdictional claims in pub‑ lished maps and institutional affiliations. tum in endemic areas as seen in dogs with different techniques [34, 36]. Priolo et al. Parasites Vectors (2019) 12:133 Page 9 of 9 Author details 18. Yin Q, El-Ashram S, Liu X-Y, Suo X. Early detection of Toxoplasma 1 Dipartimento di Scienze Veterinarie, Università di Messina, Messina, Italy. gondii-infected cats by interferon-gamma release assay. Exp Parasitol. 2 Departament de Medicina i Cirurgia Animals, Facultat de Veterinària, 2015;157:145–9. Universitat Autònoma de Barcelona, Bellaterra, Barcelona, Spain. 3 Centro di 19. Pennisi MG, Lupo T, Malara D, Masucci M. Serological and molecular Referenza Nazionale per la Leishmaniosi (CReNaL), Istituto Zooprofilattico prevalence of Leishmania infantum infection in cats from southern Italy. J Sperimentale della Sicilia “A. Mirri”, Palermo, Italy. Feline Med Surg. 2012;14:656–7. 20. Sherry K, Trotta M, Miranda C, Montoya A, Miró G, Trotta M, et al. A sero‑ Received: 25 November 2018 Accepted: 7 March 2019 logical and molecular study of Leishmania infantum infection in cats from the island of Ibiza (Spain). Vector-Borne Zoonotic Dis. 2011;11:239–45. 21. Persichetti MF, Solano-Gallego L, Vullo A, Masucci M, Marty P, Delaunay P, et al. Diagnostic performance of ELISA, IFAT and Western blot for the detection of anti- Leishmania infantum antibodies in cats using a Bayes‑ References ian analysis without a gold standard. Parasit Vectors. 2017;10:119. 1. World Organization for Animal Health. Leishmaniosis. Manual of diagnos‑ 22. Vitale F, Reale S, Vitale M, Petrotta E, Torina A, Caracappa S. TaqMan-based tic tests and vaccines for terrestrial animals. 2018. http://www.oie.int/en/ detection of Leishmania infantum DNA using canine samples. Ann N Y inter​natio​nal-stand​ard-setti​ng/terre​stria​l-manua​l/acces​s-onlin​e. Accessed Acad Sci. 2004;1026:139–43. 05 Mar 2019. 23. Persichetti MF, Pennisi MG, Vullo A, Masucci M, Migliazzo A, Solano- 2. Maroli M, Feliciangeli MD, Bichaud L, Charrel RN, Gradoni L. Phlebotomine Gallego L. Clinical evaluation of outdoor cats exposed to ectoparasites sandflies and the spreading of leishmaniases and other diseases of public and associated risk for vector-borne infections in southern Italy. Parasit health concern. Med Vet Entomol. 2013;27:123–47. Vectors. 2018;11:136. 3. Maroli M, Pennisi MG, Di Muccio T, Khoury C, Gradoni L, Gramiccia M. 24. Kasiulevičius V, Šapoka V, Filipavičiūtė R. Sample size calculation in epide‑ Infection of sandflies by a cat naturally infected with Leishmania infan- miological studies. Gerontologija. 2006;7:225–31. tum. Vet Parasitol. 2007;45:357–60. 25. Singh OP, Gidwani K, Kumar R, Nylén S, Jones SL, Boelaert M, et al. 4. Del Río L, Chitimia L, Cubas A, Victoriano I, De la Rúa P, Gerrikagoitia X, Reassessment of immune correlates in human visceral leishmaniasis et al. Evidence for widespread Leishmania infantum infection among as defined by cytokine release in whole blood. Clin Vaccine Immunol. wild carnivores in L. infantum periendemic northern Spain. Prev Vet Med. 2012;19:961–6. 2014;113:430–5. 26. Nieto A, Domínguez-Bernal G, Orden JA, De La Fuente R, Madrid-Elena 5. Pennisi MG. Leishmaniosis of companion animals in Europe: an update. N, Carriòn J. Mechanisms of resistance and susceptibility to experimental Vet Parasitol. 2015;208:35–47. visceral leishmaniosis: BALB/c mouse versus Syrian hamster model. Vet 6. Sergent E, Sergent E, Lombard J, Quilichini M. La leishmaniose à Alger. Res. 2011;42:39. Infection simultanée d’un enfant, d’un chien et d’un chat dans la même 27. Kumar T, Nylén S. Immunobiology of visceral leishmaniasis. Front Immu‑ habitation. Bull Soc Pathol Exot. 1912;5:93–8. nol. 2012;3:251. 7. Pennisi MG, Cardoso L, Baneth G, Bourdeau P, Koutinas A, Miró G, et al. 28. Martínez-Orellana P, Marí-Martorell D, Montserrat-Sangrà S, Ordeix L, Leishvet update and recommendations on feline leishmaniosis. Parasit Baneth G, Solano-Gallego L. Leishmania infantum-specific IFN-γ produc‑ Vectors. 2015;8:302. tion in stimulated blood from dogs with clinical leishmaniosis at diagno‑ 8. Pennisi MG, Hartmann K, Lloret A, Addie D, Belák S, Boucraut-Baralon C, sis and during treatment. Vet Parasitol. 2017;248:39–47. et al. Leishmaniosis in cats: ABCD guidelines on prevention and manage‑ 29. Pennisi MG, Masucci M, Catarsini O. Presenza di anticorpi anti-Leish- ment. J Feline Med Surg. 2013;15:638–42. mania in gatti FIV+ che vivono in zona endemica. Atti Soc Ital Sci Vet. 9. Baneth G, Koutinas AF, Solano-Gallego L, Bourdeau P, Ferrer L. Canine 1998;52:265–6. leishmaniosis - new concepts and insights on an expanding zoonosis: 30. Pennisi MG, Maxia L, Vitale F, Masucci M, Borruto G, Caracappa S. Studio part one. Trends Parasitol. 2008;24:324–30. dell’infezione da Leishmania mediante PCR in gatti che vivono in zona 10. Solano-Gallego L, Montserrat-Sangrà S, Ordeix L, Martínez-Orellana endemica. Atti Soc Ital Sci Vet. 2000;54:215–6. P, Montserrrat-Sangrà S, Ordeix L, et al. Leishmania infantum-specific 31. Brianti E, Falsone L, Napoli E, Gaglio G, Giannetto S, Pennisi MG, et al. production of IFN-γ and IL-10 in stimulated blood from dogs with clinical Prevention of feline leishmaniosis with an imidacloprid 10%/flumethrin leishmaniosis. Parasit Vectors. 2016;9:317. 4.5% polymer matrix collar. Parasit Vectors. 2017;10:334. 11. Hosein S, Blake DP, Solano-Gallego L. Insights on adaptive and innate 32. Otranto D, Napoli E, Stefania M, Annoscia G, Domenica V, Greco G, et al. immunity in canine leishmaniosis. Parasitology. 2017;144:95–115. Feline and canine leishmaniosis and other vector-borne diseases in the 12. Papadogiannakis EI, Koutinas AF. Cutaneous immune mechanisms in Aeolian Islands: pathogen and vector circulation in a confined environ‑ canine leishmaniosis due to Leishmania infantum. Vet Immunol Immuno‑ ment. Vet Parasitol. 2017;236:144–51. pathol. 2015;163:94–102. 33. Portús M, Gállego M, Riera C, Aisa MJ, Fisa R, Castillejo S. Wild and 13. Zafra R, Jaber JR, Pérez-Ecija RA, Barragán A, Martínez-Moreno A, Pérez J. domestic mammals in the life cycle of Leishmania infantum in Southwest High iNOS expression in macrophages in canine leishmaniasis is associ‑ Europe. A literature review and studies performed in Catalonia (Spain). ated with low intracellular parasite burden. Vet Immunol Immunopathol. Rev Iber Parasitol. 2002;62:72–6. 2008;123:353–9. 34. Lombardo G, Pennisi MG, Lupo T, Migliazzo A, Caprì A, Solano-Gallego L. 14. Solano-Gallego L, Rodríguez-Cortés A, Iniesta L, Quintana J, Pastor J, Detection of Leishmania infantum DNA by real-time PCR in canine oral Espada Y, et al. Cross-sectional serosurvey of feline leishmaniasis in ecore‑ and conjunctival swabs and comparison with other diagnostic tech‑ gions around the Northwestern Mediterranean. Am J Trop Med Hyg. niques. Vet Parasitol. 2012;184:10–7. 2007;76:676–80. 35. Attipa C, Papasouliotis K, Solano-Gallego L, Baneth G, Nachum-Biala Y, 15. Day MJ. Cats are not small dogs: is there an immunological explanation Sarvani E, et al. Prevalence study and risk factor analysis of selected bacte‑ for why cats are less affected by arthropod-borne disease than dogs? rial, protozoal and viral, including vector-borne, pathogens in cats from Parasit Vectors. 2016;9:507. Cyprus. Parasit Vectors. 2017;10:130. 16. Rossi G, Cornaro C, Battilani M, Pocacqua V, Paltrinieri S. Production 36. Cardoso L, Neto F, Sousa JC, Rodrigues M, Cabral M. Use of a leishmanin of IFN-γ in feline whole blood after incubation with potential T-cell skin test in the detection of canine Leishmania-specific cellular immunity. epitopes of the nucleocapsid protein of feline coronavirus. Vet Microbiol. Vet Parasitol. 1998;79:213–20. 2011;150:248–56. 17. Gelain ME, Meli M, Paltrinieri S. Whole blood cytokine profiles in cats infected by feline coronavirus and healthy non-FCoV infected specific pathogen-free cats. J Feline Med Surg. 2006;8:389–99.