Cathelicidin Insufficiency in Patients with Fatal Leptospirosis

Abstract

Leptospirosis causes significant morbidity and mortality worldwide; however, the role of the host immune response in disease progression and high case fatality (>10-50%) is poorly understood. We conducted a multi-parameter investigation of patients with acute leptospirosis to identify mechanisms associated with case fatality. Whole blood transcriptional profiling of 16 hospitalized Brazilian patients with acute leptospirosis (13 survivors, 3 deceased) revealed fatal cases had lower expression of the antimicrobial peptide, cathelicidin, and chemokines, but more abundant pro-inflammatory cytokine receptors. In contrast, survivors generated strong adaptive immune signatures, including transcripts relevant to antigen presentation and immunoglobulin production. In an independent cohort (23 survivors, 22 deceased), fatal cases had higher bacterial loads (P = 0.0004) and lower anti-Leptospira antibody titers (P = 0.02) at the time of hospitalization, independent of the duration of illness. Low serum cathelicidin and RANTES levels during acute illness were independent risk factors for higher bacterial loads (P = 0.005) and death (P = 0.04), respectively. To investigate the mechanism of cathelicidin in patients surviving acute disease, we administered LL-37, the active peptide of cathelicidin, in a hamster model of lethal leptospirosis and found it significantly decreased bacterial loads and increased survival. Our findings indicate that the host immune response plays a central role in severe leptospirosis disease progression. While drawn from a limited study size, significant conclusions include that poor clinical outcomes are associated with high systemic bacterial loads, and a decreased antibody response. Furthermore, our data identified a key role for the antimicrobial peptide, cathelicidin, in mounting an effective bactericidal response against the pathogen, which represents a valuable new therapeutic approach for leptospirosis.

Fig 1. Transcriptional signatures associated with fatal cases.

(A) PCA of all probes for 3 patient groups and healthy volunteers. (B) Heatmap depicting hierarchical clustering of 471 probes with differential expression during acute illness in 3 deceased patients (D) and 13 acute survivors (S). For comparison, the same transcripts for 4 healthy volunteers are shown (H). Blue indicates down-regulation and Red indicates up-regulation in log₂. (C-E) Functional REACTOME pathways for 3 expression groups with negative log of p-values and number of genes in parentheses. In Groups 1 (black box) and 2 (blue box), transcripts were enriched in survivors vs fatal cases, while in Group 3 (red box), transcripts were enriched in fatal cases.

Fig 2. Specific transcripts associated with case fatality.

Values are the average normalized log₂ fold-change of signal intensities ± Standard Error of the Mean for select transcripts in Groups 1–3 described in Fig 1. The gene names are shown on the left and the functional annotation is shown on the right. Genes were selected based on their fold-change in Deceased vs Survivor (DvS) comparisons and had significant q values.

Fig 3. More robust T and B cell responses in patients lacking acute lung injury.

PBMCs from patients with ALI (N = 4) and hospitalized patients lacking pulmonary complications (No ALI; N = 9) during acute leptospirosis. Cells were labeled with fluorescent antibodies for immunophenotyping and analyzed by flow cytometry [36]. Live CD3⁺ cells or CD3^–/CD19⁺ cells were sampled and clustered by Citrus analysis, based on the expression of markers in each panel. Abundance of subsets was compared using SAM (FDR < 5%) between No ALI and ALI groups. Data shown represent fold change ratios of cell abundance for the indicated cell subsets.

Fig 4. Serum protein levels validate expression profiles of specific gene products identified by microarray.

Serum protein levels of cathelicidin (CAMP [LL-37]) (A) and RANTES (D) were higher in survivors, while HGF (B) and IL-18 (E) serum levels were higher in deceased patients. (C) Deceased patients had higher serum of CHI3L1 than survivors, in contrast to microarray results. (F) Elastase levels did not differ between outcomes. N = 23 for Survivors (S; blue triangles) and N = 22 for and deceased patients (D; red circles) for (A-E). N = 14 for survivors and N = 15 for deceased patients in (F). Filled symbols denote individuals included in microarray analyses. Values are medians +/- IQR in pg/mL (B-E) or ng/mL (A, F).

Fig 5. Cathelicidin (LL-37) protects hamsters from lethal Leptospira infection.

(A) Survival in hamsters pre-treated with 1 mg/kg of cathelicidin (LL-37) (n = 7) was significantly greater than LL-37 scrambled peptide (Scrambled) (n = 7) (P = 0.016) or ddH₂O-treated controls (ddH₂O) (n = 7) (P = 0.008) following lethal challenge with 100 Leptospira. (B) Bacterial loads (Leptospira genome equivalents per mL of whole blood) in 7 infected hamsters were significantly lower at 4 (P = 0.035; P = 0.146), 6 (P = 0.003; P = 0.001), and 8 days (P = 0.003; P = 0.001) post-infection in LL-37-treated hamsters relative to 7 scrambled peptide (black) or ddH₂O-treated controls (red), respectively. Shown are medians ± IQR. An * signifies a P-value ≤0.05; ** signifies P<0.01 as determined by Mantel Cox test for (A) or Mann-Whitney test for (B).