A TCRβ Repertoire Signature Can Predict Experimental Cerebral Malaria

Abstract

Cerebral Malaria (CM) is associated with a pathogenic T cell response. Mice infected by P. berghei ANKA clone 1.49 (PbA) developing CM (CM+) present an altered PBL TCR repertoire, partly due to recurrently expanded T cell clones, as compared to non-infected and CM- infected mice. To analyse the relationship between repertoire alteration and CM, we performed a kinetic analysis of the TRBV repertoire during the course of the infection until CM-related death in PbA-infected mice. The repertoires of PBL, splenocytes and brain lymphocytes were compared between infected and non-infected mice using a high-throughput CDR3 spectratyping method. We observed a modification of the whole TCR repertoire in the spleen and blood of infected mice, from the fifth and the sixth day post-infection, respectively, while only three TRBV were significantly perturbed in the brain of infected mice. Using multivariate analysis and statistical modelling, we identified a unique TCRβ signature discriminating CM+ from CTR mice, enriched during the course of the infection in the spleen and the blood and predicting CM onset. These results highlight a dynamic modification and compartmentalization of the TCR diversity during the course of PbA infection, and provide a novel method to identify disease-associated TCRβ signature as diagnostic and prognostic biomarkers.

Fig 1. Kinetic analysis of the TCR TRBV-TRBC repertoire during the course of PbA infection.

(A) Experimental procedure showing the preparation of samples from both infected and non-infected mice, as well as the parallel analysis of mice during the course of the infection (“kinetic” samples) and at the time of ECM onset (“CM+” samples). Days post-infection (days p-i) indicate the time at which animals of the corresponding groups are sacrificed. Organs harvested for each group are indicated. (B) Modification of the TRBV-TRBC repertoire in spleen, blood and brain of B10.D2 mice during the course of Plasmodium berghei ANKA infection. Average DBV-BC perturbations across all TRBVs and individuals in each group (μμDBV-BC) in the spleen (black), the blood (red) and the brain (green) are shown for the control uninfected (CTR), day 3 p-i (d3), day 4 p-i (d4), day 5 p-i (d5), day 6 p-i (d6) and CM+ groups. DBV-BC perturbations were computed with ISEApeaks using CTR Spleen as the reference group.

Fig 2. Differential kinetics of TCR TRBV-TRBC repertoire perturbation in the spleen and blood of infected mice.

(A-B) Kinetic representation of global perturbation scores across all TRBVs in the spleen (A) and the blood (B) using principal component analysis (PCA). Progressive modification of the repertoire is diagrammed by the shift of day p-i-related groups from the right to the left on the first PCA component (PC1). Colors correspond to analyzed groups. (C-D) Mean DBV-BC perturbations across all TRBVs (μDBV-BC) in the spleen (C) and the blood (D). DBV-BC were computed as in Fig 1B. Black dots represent individual mouse global perturbation scores. Red dots represent the average global perturbation score for each group. Statistical comparisons were performed using the two-way ANOVA test for the difference between organs and between infected groups. Tests were significant for organs (p<0.0001) and day post infection (p<0.0001) at α = 0.05.

Fig 3. TRBV-TRBC repertoire differences in spleen, blood and brain of naïve B10.D2 mice.

(A) Mean DBV-BC perturbations across all TRBVs (μDBV-BC) in the spleen (left), blood (center) and brain (right) of CTR uninfected mice are shown for each individual. DBV-BC perturbations were computed with ISEApeaks using CTR Spleen as the reference group. Statistical comparisons were performed using pairwise t-test with correction for FDR between groups depicting significant p-value (p<0.0001) for each comparison. (B) PCA on DBV-BC log perturbation scores separating repertoires of the spleen, blood and brain on the x axis (PC1).

Fig 4. Modification of the TRBV-TRBC repertoire in the brain of B10.D2 mice during the course of PbA infection.

(A) Mean DBV-BC perturbations across all TRBVs (μDBV-BC) in the brain of CTR uninfected (CTR), day 6 p-i (Day 6) and CM⁺ mice are shown for each individual. DBV-BC were computed as in Fig 1B. Black dots represent individual mouse scores. Lines represent the average score for each group. Average DBV-BC of CTR spleen is indicated as the reference (Ctr_sp). (B) Survival curve of B10.D2 mice infected with 10⁶ PbA-PRBC. All mice developed ECM symptoms. 75% of the mice died between day 6 and day 8. (C) Regression slope and 95% confidence band of parasitemia (%) over day post infection. The infection has no effect on the % of parasitemia (no significant slope). (D) DBV-BC perturbations scores of five TRBV. Statistical tests were performed as described in Fig 2.

Fig 5. A unique TRBV signature discriminates CTR from CM⁺ spleen, blood and brain repertoires.

(A) Panels of GSEA report for the peak set BL_PVCLUST_37 showing significant enrichment for CM⁺ condition compared to uninfected group in spleen (left) and blood (middle) and in the brain (right). The enrichment figures show ranked peaks according to the pooled t-statistic across imputed dataset (bottom). Peak positions are indicated on the ranked list (middle). The enrichment score (ES) is the maximum of the running sum (top). (B) Normalized GSEA enrichment scores post-infection growth curves in three compartments for the peak set BL_PVCLUST_37. In comparison to the CTR groups, the set is significantly enriched in day 5 post-infection in spleen (solid line), in day 6 in blood (dashed line) and in ECM state in brain (dotted line).