Ocular toxoplasmosis: susceptibility in respect to the genes encoding the KIR receptors and their HLA class I ligands

The objective of this study was to investigate the influence of the genes encoding the KIR receptors and their HLA ligands in the susceptibility of ocular toxoplasmosis. A total of 297 patients serologically-diagnosed with toxoplasmosis were selected and stratified according to the presence (n = 148) or absence (n = 149) of ocular scars/lesions due to toxoplasmosis. The group of patients with scars/lesions was further subdivided into two groups according to the type of ocular manifestation observed: primary (n = 120) or recurrent (n = 28). Genotyping was performed by PCR-SSOP. Statistical analyses were conducted using the Chi-square test, and odds ratio with a 95% confidence interval was also calculated to evaluate the risk association. The activating KIR3DS1 gene was associated with increased susceptibility for ocular toxoplasmosis. The activating KIR together with their HLA ligands (KIR3DS1-Bw4-80Ile and KIR2DS1+/C2++ KIR3DS1+/Bw4-80Ile+) were associated with increased susceptibility for ocular toxoplasmosis and its clinical manifestations. KIR-HLA inhibitory pairs -KIR2DL3/2DL3-C1/C1 and KIR2DL3/2DL3-C1- were associated with decreased susceptibility for ocular toxoplasmosis and its clinical forms, while the KIR3DS1−/KIR3DL1+/Bw4-80Ile+ combination was associated as a protective factor against the development of ocular toxoplasmosis and, in particular, against recurrent manifestations. Our data demonstrate that activating and inhibitory KIR genes may influence the development of ocular toxoplasmosis.

Ocular Toxoplasmosis, the most common form of posterior uveitis, results from Toxoplasma gondii infection 1 . The prevalence varies widely between different countries however, both the frequency and the severity of the resulting ocular manifestations are higher in Brazil than in many other parts of the world 2,3 . Eye injuries affect the retina and the choroid with local inflammatory reactions being observed in ocular tissues infected by T. gondii 1 .
The damage to ocular tissues led to the proposal of phenomena that may be related to pathogenic mechanisms of ocular toxoplasmosis, including autoimmune mechanisms 1,4 . Now it is known that an exaggerated T-helper 1 (Th-1) response, in particular by Th-17 cells, can cause tissue damage and contribute to the severity of ocular toxoplasmosis due to the production of interleukin-17 (IL-17), a potent inducer of inflammation 5,6 . In addition to Th-17, other sources of IL-17, including natural killer cells (NK), can contribute to the development of inflammatory conditions 7 .
Increases in the numbers of circulating proinflammatory monocytes and NK CD56 dim cytotoxic cells and a decrease in immunoregulatory NK CD56 bright cells have been identified in children congenitally infected with T. gondii who have active eye lesions. Furthermore, subsets of NK cells and CD8 + T cells play a crucial role as biomarkers of cicatricial lesion of the eye 8 . There is also evidence that NK cells have a predominantly proinflammatory profile in vitro during T. gondii infections, due to an increased production of interferon-gamma (IFN-γ ) in patients with congenital ocular toxoplasmosis 6 .
The effector function of NK cells is regulated by a set of receptors named killer immunoglobulin-like receptors (KIR) expressed on the cell surface that recognize human leukocyte antigen (HLA) class I molecules of target cells 9,10 . KIR genes are responsible for coding the KIR receptors of NK cells. These genes comprise a family of 15 genes located on chromosome 19q13.4 characterized as inhibitors (KIR2DL1, -2DL2, -2DL3, -2DL5A,-2DL5B -3DL1, -3DL2, and -3DL3) or activators (KIR2DS1, -2DS2, -2DS3, -2DS4, -2DS5, -3DS1 and -2DL4,), and two pseudogenes (KIR2DP1 and -3DP1). Moreover, based on the content of the genes, KIR genotypes are divided into two groups designated AA and BX (BB and AB) that differ in the number and type of KIR genes 11 .
KIR genes have been described as risk or protective factors in different types of non-toxoplasmic uveitis and inflammatory ocular diseases. These diseases include Behcet's uveitis 12 , uveitis in patients with spondyloarthropathies 13,14 and Vogt-Koyanagi-Harada syndrome 15,16 , all of which are triggered by autoimmune processes. KIR genes are also associated with many other infectious diseases [17][18][19] . Additionally, both murine and human studies have shown that major histocompatibility complex (MHC) class I (called HLA class I in human) are associated with Toxoplasma susceptibility [20][21][22][23][24][25][26] .
NK cells have great importance in the control of T. gondii infection 27 however, the role of KIR genes that encode the immune receptors of NK cells and can trigger local inflammation in the eye has not been elucidated in ocular toxoplasmosis yet. The objective of this study was to investigate the influence of the genes encoding the KIR receptors and their HLA ligands in the resistance or susceptibility to the development of ocular toxoplasmosis.

Results
General characteristics of patients with and without ocular manifestations of toxoplasmosis. The characteristics of the study population with respect to age, gender, clinical diagnosis and serological profile are shown in Table 1. The median ages were significantly different between the groups: A higher mean age was observed for the group of patients without ocular toxoplasmosis compared to the group of patients with ocular toxoplasmosis (P < 0.0001, t = 7.00), with the subgroup of patients with primary manifestations of ocular toxoplasmosis (P < 0.0001; t = 5.48) and with the subgroup of patients with the recurrent form of the disease (P < 0.0001; t = 7.51). A higher mean age was also observed for the subgroup of patients with primary manifestations than those who had recurrent manifestations (P = 0.002; t = 3.12).
Distribution of KIR genes and KIR genotypes in patients with and without ocular toxoplasmosis. The distribution of genotype frequencies of KIR2DL2/3 and KIR3DL1/S1 were in Hardy-Weinberg equilibrium (P > 0.05) in this study population. However, KIR3DL1/S1 for the patient group that developed ocular toxoplasmosis was not in Hardy-Weinberg equilibrium (P = 0.03). KIR framework genes, KIR2DL4, KIR3DL2, KIR3DL3 and KIR3DP1, as expected were present in all samples, which are important internal controls to check the quality of genotyping.
The distribution of KIR gene frequencies and AA and BX genotype frequencies are shown in Table 2. An increased susceptibility for developing ocular toxoplasmosis (OR = 2.15; CI = 1.31-3.50; P = 0.003, Pc = 0.04) was observed for the KIR3DS1 activating gene. There was also a positive association between KIR3DS1 and recurrent manifestations of disease (OR = 3.25; CI = 1.42-7.44, P = 0.007; Pc = 0.1) when this patient group was compared  to patients without ocular toxoplasmosis, although it was lost after applying the Bonferroni correction. On the other hand, the KIR2DS2 activating gene was associated with decreased susceptibility for ocular toxoplasmosis (OR = 0.55; CI = 0.31-0.97; P = 0.05; Pc = 0.80), but the significance was also lost after applying the Bonferroni correction. No significant difference was observed in the AA and BX genotype frequencies between all groups investigated in this study.
Distribution of the HLA class I KIR-ligands in in patients with and without ocular toxoplasmosis. Human leucocyte antigen frequencies was in Hardy-Weinberg equilibrium (P > 0.05) in all groups studied.
The frequencies of the HLA class I ligands of KIR (A3 or A11, Bw4-80Ile and − 80Thr, C1 and C2, in homozygosity and heterozygosity) were analyzed and were similar between groups ( Table 3).
Distribution of KIR and their respective HLA ligands in patients with and without ocular toxoplasmosis. Data Table 4). Figure 1 shows the influence of the number of KIR-HLA class I activating and inhibitory ligands on the development of ocular toxoplasmosis and its clinical manifestations. There were significant differences between patients with and without ocular toxoplasmosis (OR = 2.29; CI = 1.19-4.39; P = 0.01; Pc = 0.04) and patients with primary manifestations and without ocular toxoplasmosis (OR = 2.29; CI = 1.16-4.52; P = 0.01; Pc = 0.04) when only two pairs of activating ligands were present (Fig. 1A). Significant associations were not found in the analysis of pairs of inhibitory KIR-HLA class I ligands (Fig. 1B).

Discussion
Possible causes of ocular manifestations of toxoplasmosis have not been fully elucidated, but it is believed that factors related both to the parasite and the host contribute to the development of this disease 4,28 . Therefore, the study of the impact of genes on infections, such as toxoplasmosis, is extremely important because it provides information about the contribution of the host's genetic factors on the development of this type of disease. We have previously shown that MICA polymorphisms are not associated with the development of ocular toxoplasmosis 29 .
Using the same samples, the current study found that KIR receptors genes and their HLA ligands are associated  with the development of ocular lesions resulting from T. gondii infection. To the best of our knowledge, this is the first study of KIR genes and HLA ligands in the immunopathology of ocular toxoplasmosis. The difference in the mean ages of the patient groups of this study was carefully discussed previously 29 . Briefly, T. gondii infection can occur at any time of life, and although most cases of ocular toxoplasmosis occur due to infections acquired after birth, a significant number of patients can acquire the disease congenitally and the resulting scars trend to be persistent 1 . It has also been demonstrated that the risk of recurrence is higher in the year following the first infection than in future years 30,31 . As no distinction was made between congenital and acquired disease in the analysis of the characteristics of eye injuries, this may be one of the possible explanations for the lower mean age observed for patients who developed ocular toxoplasmosis, including those who present with recurrent signs of the disease. Furthermore, other eye diseases, those without scars/lesions due to toxoplasmosis, are prevalent in older patients 32 .
Regarding the distribution of KIR genes, after Bonferroni correction only the KIR3DS1 activating gene was associated with increased risk of developing ocular toxoplasmosis with the other associations being lost. The Bonferroni correction decreases the chance of a significant difference by chance alone, making the data more robust 33 . The association observed for the KIR3DS1 with ocular toxoplasmosis can be explained by the absence of KIR3DL1, because 3DS1 and 3DL1 segregate as alleles of a single locus. Thus, the presence of KIR3DS1, and consequently the absence of KIR3DL1, can create an increased potential for the activation of NK cells owing to decreases in the ratio of inhibitory/activating receptors. Previous studies have shown involvement of the KIR3DS1/L1 alleles in various types of non-toxoplasmic uveitis and inflammatory eye diseases triggered by autoimmune factors. Levinson et al. 15 observed that individuals with the KIR3DS1 gene in their haplotype have an increased risk of developing Vogt-Koyanagi-Harada syndrome, while the presence of KIR3DL1 was associated to protection against the development of the disease. A similar result was observed by Moon et al. 14 for the development of uveitis related to ankylosing spondylitis: KIR3DS1 was associated as a risk factor and KIR3DL1 was associated as a protective factor.
In this study, KIR3DL1/S1 for the patient group that developed the ocular toxoplasmosis were not in Hardy-Weinberg equilibrium. However, some authors claim that this equilibrium should only be investigated in the control group, because it represents the general population 34,35 . Thus, the high frequency of KIR3DS1 might be changing the distribution of these alleles in individuals with ocular toxoplasmosis, resulting in a deviation from the Hardy-Weinberg equilibrium. Considering that numerous precautions were adopted to prevent bias in this study, we can safely say that the 3DS1 gene exerts a real influence on the development of ocular toxoplasmosis, even though KIR3DL1/S1 were not in Hardy-Weinberg equilibrium.
The function of KIR genes in immune response is highly dependent on the HLA molecules expressed on the target cell surface. Therefore, KIR receptors influence susceptibility for or protection against certain illnesses by means of a balance in activation and inhibition signals that regulate the NK cell effector function 9,10 . The recognition of specific HLA ligands by inhibiting KIR is well established 36 . There is clear evidence that HLA-Bw4-80Ile is powerfully recognized by KIR3DL1, but there is still controversy as to whether its homologue, KIR3DS1, interacts with the same ligand, as only indirect evidence was found 37,38 . There is also a hierarchy of inhibition related to KIR2DL receptors in which KIR2DL3-C1 has lower inhibitory potential than KIR2DL2-C1 and KIR2DL1-C2 39 .
In this study, where the KIR genes were analyzed in the presence of their respective ligands (KIR-HLA), the KIR3DS1-Bw4-80Ile pair was associated with the development of ocular toxoplasmosis irrespective of the type of clinical manifestation (primary or recurrent); while KIR2DL3/2DL3-C1/C1 and KIR2DL3/2DL3-C1 were associated with protection against the development of ocular toxoplasmosis and its clinical manifestations. In accord with our results, other studies found an association involving the KIR3DS1-Bw4-80Ile pair in other human diseases [40][41][42][43] , suggested an interaction between these molecules. KIR2DL3 and KIR2DL2 are considered alleles, as are KIR3DL1 and KIR3DS1. Although the KIR2DL2-C1 interaction is stronger than the KIR2DL3-C1 interaction, the inhibitory signal generated by the absence of KIR2DL2 (KIR2DL3/2DL3-C1/C1 and KIR2DL3/2DL3-C1) appears to be sufficient to inhibit the effector function of NK cells and protect against elevated inflammation and tissue damage. It has been shown that abnormal expressions of inhibitory receptors of NK cells, including the KIR2DL3 receptor, may be associated with the development of Behcet's disease 44 .
It is important to highlight the results observed for the number of pairs of KIR-HLA ligands and the correlation between the distribution of activating KIR and their respective HLA ligands. A higher frequency of only two pairs of ligands was observed in patients with ocular toxoplasmosis and in patients with primary manifestations compared to patients without ocular toxoplasmosis. The combination responsible for this association was KIR2DS1 + /C2 ++ KIR3DS1 + /Bw4-80Ile + , although the KIR2DS1 and KIR3DS1 genes are not in linkage disequilibrium as observed among patients with ocular toxoplasmosis (Δ ′ = 0.06; P = 0.68) and patients without ocular toxoplasmosis (Δ ′ = 0.15; P = 0.45). The combination of one pair (inhibitory) in the absence of the other pair (activating) was analyzed. It was possible to observe that the KIR3DS1 − /KIR3DL1 + /Bw4-80Ile + combination decreases the risk of developing ocular toxoplasmosis and its recurrent clinical forms.
These results may suggest that the activating function mediated by KIR2DS1 plus KIR3DS1 and their respective HLA ligands is, in fact, an important factor interfering in ocular toxoplasmosis, since such a combination may affect the balance of inhibiting/activating signals. NK cells are activated when there is an increase of activation signals, even if there is a combination of strong or weak inhibitory signals 45 . Yet, the absence of KIR3DS1 in the presence of its inhibitor homologue, KIR3DL1, was indicative of the inhibitory or protective role played by NK cells, which perhaps avoids subsequent immune responses that may trigger inflammation and autoimmunity. In support of our findings, Levinson et al. 46 demonstrated both negative and positive associations mediated by KIR-HLA pairs in another ocular inflammatory disease, birdshot chorioretinopathy.
NK, CD4 and CD8 T cells, and type 1 cytokines, such as IFN-γ and IL-2, play a protective role in T. gondii infection. IL-12 stimulates NK cells to produce IFN-γ and to promote the development of Th-1 cells that produce IFN-γ , a cytokine involved in the activation of macrophages, the main phagocytes in chronic inflammation 47,48 . In the eye, although the immune response is usually suppressed to prevent tissue damage, there is experimental evidence that T. gondii infection promotes the production of factors such as IFN-γ that suppress the immune privilege of this organ 49 . This possibly leads to increased severity of lesions with marked necrosis or inflammation of the retina and the choroid 50,51 .
The immune response can determine the development of eye injuries resulting from T. gondii infection, and the mechanisms involved may be associated with both the pathogenesis and protective effects that control tissue damage. It has been shown that increases in the frequency of circulating NK cells and proinflammatory monocytes in children infected by T. gondii, particularly in those with active ocular lesions, are indicative of a strong and persistent proinflammatory response. Moreover, subsets of NK cells and CD8 + T cells act as biomarkers for cicatricial lesions of the eye 8 . It has also been demonstrated that NK cells show increased production of IFN-γ in patients with congenital ocular toxoplasmosis 6 . Furthermore, the high imunopathogenicity responsible for tissue damage and deterioration of ocular toxoplasmosis may be due to the presence of a potent inducer of inflammation, IL-17 5,6 , which is also produced by NK cells 7 . On the other hand, ocular toxoplasmosis may be linked to autoimmunity 1,4 .
The current study investigated the KIR-HLA ligand as a risk factor in ocular toxoplasmosis and these results may improve to understanding of the immunopathogenic mechanism involving NK cells in ocular manifestations related to toxoplasmosis. However, others studies should be performed such as histological analyses of the ocular tissue affected by T. gondii and NK cytotoxicity assays to better understanding the role of NK cells and the expression of KIR in the immunopathogenesis of ocular toxoplasmosis.
In conclusion, the results of this study show that activating and inhibitory KIR in the presence of their respective HLA ligands may have influence on the development of ocular toxoplasmosis and its clinical form in this population. In particular this is seen with the strong presence of activating signals as risk factors (KIR3DS1, KIR3DS1-Bw4-80Ile and KIR3DS1 + /Bw4-80Ile ++ KIR2DS1 + /C2 + ) and inhibitory signals as protective factors (KIR2DL3/2DL3-C1, KIR2DL3/2DL3-C1/C1 and KIR3DS1 -/KIR3DL1 + /Bw4-80Ile + ).

Methods
The design of this study aimed to follow, as close as possible, the criteria recommended by STrengthening the REporting of Genetic Association Studies (STREGA) 52 . The study subjects have been described previously 29 . Patients were grouped according to the presence of ocular scars/lesions due to toxoplasmosis (n = 148; 79 men and 69 women; mean age: 42.3 ± 20.6 years) or to the presence of ocular diseases not related to toxoplasmosis (n = 149; 73 men and 76 women; mean age: 57.7 ± 16.9 years). The group of patients with scars/lesions due to toxoplasmosis was further subdivided into two groups according to the type of ocular manifestation observed during a follow up period of at least two years: primary manifestations (n = 120; 65 men and 55 women; mean age: 44.9 ± 20.9 years) and recurrent manifestations characterized by the presence of satellite lesions 53 (n = 28; 14 men and 14 women; mean age: 31.8 ± 30.5 years) ( Table 1).

Ethics Information. This study was approved by the Research Ethics
All individuals who participated in this study were monitored and evaluated in respect to clinical symptoms, serology for T. gondii and epidemiological data. Besides, although patients self-reported themselves as European descent, mixed African and European descent, and African descent, due to high miscegenation of the Brazilian population they were defined as a population of mixed ethnicity 54 .
In this study, in order to avoid bias in the results, all patients were selected after clinical examination using the same criteria. Additionally, the probability of variations in the allele frequencies due to ethnic background was minimized by matching patients with ocular toxoplasmosis and patients without ocular toxoplasmosis from similar ethnic backgrounds. Furthermore, gender and residence in the same geographical areas were carefully matched during group selection.
The number of patients enrolled is sufficient to demonstrate whether there is an association between ocular toxoplasmosis and KIR genes with statistical power of more than 90% and it was chosen according to frequency of KIR genes recorded in Allele*Frequencies database (http://www.allelefrequencies.net) observed in a population located in the southeast region of Brazil and defined as a population of mixed ethnicity.
Inclusion/exclusion criteria. The inclusion criteria of patients with ocular toxoplasmosis were positive laboratory diagnosis of toxoplasmosis, the presence of ocular scars/lesions due to toxoplasmosis and live in municipalities in the northwest region of the State of Sao Paulo (located in the southeast region of Brazil, between 20°49'13″ S and 49°22′ 47″ W). The inclusion criteria of patients without ocular toxoplasmosis were positive laboratory diagnosis of toxoplasmosis but without ocular scars/lesions due to toxoplasmosis and living in the same geographical region as the patients with ocular toxoplasmosis. All patients were clinically evaluated by two experienced physicians.
The exclusion criteria were: patients with other infectious and parasitic diseases, patients with any type of mental disability, patients with blood dyscrasia and using oral anticoagulants and related patients. Clinical diagnosis. All patients were clinically assessed using an indirect binocular ophthalmoscope (Binocular Ophthalmoscope ID10, Topcon Corporation, USA). The evaluation of visual acuity followed the log-MAR Early Treatment Diabetic Retinopathy Study chart (ETDRS) criteria 55 . Intraocular pressure was measured by Goldmann applanation tonometry, and stereoscopic biomicroscopy was performed using a 78-diopter lens (Volk) and a slit lamp and all they were classified according to the ETDRS criteria 55 .
As no invasive test was performed, the ocular toxoplasmosis diagnostic criteria used were the same as in the clinical practice: injury identified by ophthalmoscopy associated with positive serology for T. gondii. This is therefore a presumptive diagnosis.
KIR and HLA genotyping. Blood samples were also collected into tubes containing EDTA anticoagulant for DNA extraction. DNA of all patients was extracted using the commercial kit for silica column extraction (QIAamp ® DNA Blood Mini Kit, QIAGEN, the Netherlands) following the manufacturer's instructions. All DNA samples were subjected to an evaluation of concentration and purity using the ratio of the absorbance at optical densities (OD) of 260 and 280 nm with Epoch ™ equipment (BioTek, Winooski, Vermont, USA). KIR and HLA-A, -B and -C were genotyped according to manufacturer's instructions by Polymerase chain reaction-sequence specific oligonucleotide probe (PCR-SSOP) protocols with Luminex ® technology (One Lambda Inc., Canoga Park, CA, USA). This technique uses PCR-amplified DNA with specific biotinylated primers. The amplified product is hybridized by complementary DNA probes conjugated to fluorescently coded microspheres, with detection using R-Phycoerythrin-conjugated Streptavidin (SAPE). The data were interpreted using a computer program (HLA Fusion, 2.0 Research, One Lambda).
Scientific RepoRts | 6:36632 | DOI: 10.1038/srep36632 Two types of KIR genotypes have been described based on the content of the genes: AA and BX (BB and AB) (defined according to http://www.allelefrequencies.net). Individual genotypes were determined to be AA when the genes KIR2DL1, KIR2DL3, KIR2DL4, KIR2DS4, KIR3DL1, KIR3DL2, KIR3DL3, KIR2DP1 and KIR3DP1 were present. The presence of one or more of the following genes: KIR2DL5, KIR2DS1, KIR2DS2, KIR2DS3, KIR2DS5 and KIR3DS1 characterized the BX genotype.
Statistical analysis. KIR, HLA and KIR-HLA frequencies were obtained by direct counting. The comparisons of the frequencies of HLA ligands, KIR genes, KIR AA and BX genotypes and KIR with or without ligands between groups of patients were performed with the Chi-square test with Yates' correction or, when necessary, Fisher's exact test using the program Graph Pad Instat (http://www.graphpad.com/quickcalcs/contingency1.cfm). Odds ratio (OR) with a 95% confidence interval (95% CI) was also calculated to evaluate the risk association. The mean ages were compared using the t-test. Differences with P-values < 0.05 corrected by the Bonferroni inequality method for multiple comparisons (Pc) were considered statistically significant. A Hardy-Weinberg equilibrium fit was performed by calculating expected genotype frequencies and comparing that with the observed values for KIR2DL2/3, KIR3DL1/S1, and the HLA alleles using ARLEQUIN software, version 3.1. (http://cmpg. unibe.ch/software/arlequin3).