Circulating Biomarkers of Immune Activation, Oxidative Stress and Inflammation Characterize Severe Canine Visceral Leishmaniasis

Abstract

Clinical manifestations in canine visceral leishmaniasis (CVL) have not been clearly associated with immunological status or disease progression. We simultaneously assessed biomarkers of inflammation, immune activation, oxidative stress, and anti-sand fly saliva IgG concentrations in dog sera with different clinical manifestations to characterize a biosignature associated with CVL severity. In a cross-sectional exploratory study, a random population of 70 dogs from an endemic area in Brazil was classified according to CVL clinical severity and parasitological evaluation. A panel of biomarkers and anti-sand fly saliva IgG were measured in canine sera. Assessment of protein expression of profile biomarkers identified a distinct biosignature that could cluster separately animal groups with different clinical scores. Increasing severity scores were associated with a gradual decrease of LTB4 and PGE2, and a gradual increase in CXCL1 and CCL2. Discriminant analyses revealed that combined assessment of LTB4, PGE2 and CXCL1 was able to distinguish dogs with different clinical scores. Dogs with the highest clinical score values also exhibited high parasite loads and higher concentrations of anti-saliva antibodies. Our findings suggest CVL clinical severity is tightly associated with a distinct inflammatory profile hallmarked by a differential expression of circulating eicosanoids and chemokines.

Figure 1. Distinct expression of immune and inflammatory markers in serum from dogs presenting with different VL clinical severity scores.

(A) Hierarchical cluster analysis (Ward’s method) with bootstrap was performed to depict the overall expression profile of the indicated serum biomarkers in the different study groups. (B) Scatter plots of biomarkers displaying significant statistical differences (P < 0.005) between the study groups using Kruskal-Wallis test. Non-parametric linear trend ad hoc tests were employed to examine the variation of the biomarker levels following the clinical severity score. SOD levels did not exhibit linear trend, thus data were compared using the Dunn’s multiple comparisons test (**P < 0.01).

Figure 2. Network analysis of the circulating biomarkers in dogs with VL reveals a distinct biosignature of VL clinical severity.

(A) The network analysis (host interactome) shows statistically significant Spearman correlations (P < 0.05) between of the biomarkers measured. See Supplemental File 1 for additional details on the strength (r value) and level of significance (P-value) for each individual correlation. (B) Heatmap of the number of statistically significant correlations involving each biomarker measured in dogs with different VL disease severity is shown. Details on the numbers of correlations are shown is Supplemental File 1. (C) Comparisons of the network densities is shown (density calculations are described in Methods). Data was compared using permutation test. ns, non significant.

Figure 3. Using inflammatory markers to predict VL disease severity.

(A–D) ROC curve analyses were performed to estimate in a quantitative way the performance of the different combinations of biomarkers used in the cluster analysis in segregating dogs diverging in VL clinical severity. AUC: area under the curve. *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 4. Associations between antibodies production against salivary recombinant proteins LJM11 and LJM17 and parasite load of infected dogs.

(A) Correlations between parasite load and antibodies production against salivary recombinant proteins LJM11 and LJM17 in negative and infected dogs showing different clinical signs (n = 70). Dotted lines on the X-axis represent the median value of parasite load within the group of infected dogs, while dotted lines on Y-axis indicate the cut-off value for antibodies production against LJM11 and LJM17. In (B), the correlations were stratified according to clinical score presented by the dogs. White areas designate the quadrants that include the dogs displaying values of parasite load below the median and negative antibodies production; light green area designate the quadrants that include the dogs displaying values of parasite load below the median and positive antibodies production; dark green areas designate the quadrants that include the dogs displaying values of parasite load above the median and positive antibodies production; and grey areas designate the quadrants that include the dogs displaying values of parasite load above the median and negative antibodies production. In (C) the percentage of individuals within each area were compared between the groups with different clinical score using a chi-square analysis.