Bundibugyo ebolavirus Survival Is Associated with Early Activation of Adaptive Immunity and Reduced Myeloid-Derived Suppressor Cell Signaling

Abstract

Ebolaviruses Bundibugyo virus (BDBV) and Ebola virus (EBOV) cause fatal hemorrhagic disease in humans and nonhuman primates. While the host response to EBOV is well characterized, less is known about BDBV infection. Moreover, immune signatures that mediate natural protection against all ebolaviruses remain poorly defined. To explore these knowledge gaps, we transcriptionally profiled BDBV-infected rhesus macaques, a disease model that results in incomplete lethality. This approach enabled us to identify prognostic indicators. As expected, survival (∼60%) correlated with reduced clinical pathology and circulating infectious virus, although peak viral RNA loads were not significantly different between surviving and nonsurviving macaques. Survivors had higher anti-BDBV antibody titers and transcriptionally derived cytotoxic T cell-, memory B cell-, and plasma cell-type quantities, demonstrating activation of adaptive immunity. Conversely, a poor prognosis was associated with lack of an appropriate adaptive response, sustained innate immune signaling, and higher expression of myeloid-derived suppressor cell (MDSC)-related transcripts (S100A8, S100A9, CEBPB, PTGS2, CXCR1, and LILRA3). MDSCs are potent immunosuppressors of cellular and humoral immunity, and therefore, they represent a potential therapeutic target. Circulating plasminogen activator inhibitor 1 (PAI-1) and tissue plasminogen activator (tPA) levels were also elevated in nonsurvivors and in survivors exhibiting severe illness, emphasizing the importance of maintaining coagulation homeostasis to control disease progression. IMPORTANCE Bundibugyo virus (BDBV) and Ebola virus (EBOV) are ebolaviruses endemic to Africa that cause severe, often fatal hemorrhagic disease. BDBV is considered a less pathogenic ebolavirus due to its reduced lethality during human outbreaks, as well as in experimentally infected nonhuman primates. The reduced mortality of BDBV in NHP models, resulting in a pool of survivors, afforded us the unique opportunity of identifying immune correlates that confer protection against ebolaviruses. In this study, we discovered that the survival of BDBV-infected nonhuman primates (NHPs) was dependent on early development of adaptive (memory) immune responses and reduced myeloid-derived suppressor cell (MDSC)-related signaling. MDSCs are a heterogenous group of immune cells implicated in a number of diseases that are powerful immunosuppressors of cellular and humoral immunity. Thus, MDSCs represent a novel therapeutic target to prevent ebolavirus disease. To our knowledge, this is the first study to link increased morbidity with recruitment of these potent immunosuppressive cells.

Keywords: Ebola virus; coagulation; filovirus; immunology; myeloid-derived suppressor cell; nonhuman primate; pathogenesis.

FIG 1

Survival of BDBV-infected macaques and comparison of peak viral loads and lesion severity scores in fatal and surviving subjects. (A) Survival curve of rhesus macaques (n = 10) infected intramuscularly with 1,000 PFU of BDBV-Uganda up to the ≥28-day endpoint. (B) Clinical scores of individual fatal versus surviving BDBV-infected macaques; criteria include behavior, posture and activity level, appetite, respiration, and the presence of hemorrhagic manifestations. (C) Peak viral loads were measured by RT-qPCR in whole blood and reported as log₁₀ copies/ml irrespective of the day on which the highest viremia was detected. The limit of detection for this assay was 1,000 copies/ml. (D) Peak viral loads were measured in plasma samples by standard plaque assay and reported as log₁₀ PFU/ml irrespective of the day on which the highest viremia was detected. The limit of detection for this assay was 25 PFU/ml. (C and D) Statistical significance was determined using the Mann-Whitney nonparametric two-tailed t test. ns, no statistically significant difference; ** P < 0.001; ***, P < 0.0001; ****, P < 0.00001. (E and F) Viral loads determined by PCR (E) or plaque assay (F) at each dpi sampled are displayed and are reported as log₁₀ copies/ml or log₁₀ PFU/ml, respectively. Red denotes fatal group, black denotes survivor group. Each replicate is shown with symbols denoting data for individual subjects (n = 10 biologically independent animals/samples per tissue type in a single experiment); each bar and error bar represents the group mean value ± the standard error of the mean (SEM).

FIG 2

Representative histologic lesions among BDBV-infected macaques. (A to E) Fatal 4 (animal identifier). (A) Expansion of medullary sinuses in lymph node (LN) and medullary histiocytosis. (B) Expansion of hepatic sinusoidal spaces with Kupffer cell hypertrophy and hyperplasia. (C) Numerous tingible body macrophages and loss of a defined marginal zone in splenic white pulp. (D) Expansion of alveolar septa with mixed inflammatory cells and increased numbers of alveolar macrophages. (E) Modest expansion of the choroid plexus mononuclear cells. (F to J) Fatal 2. (F) Expansion of medullary and subcapsular sinuses in lymph node and histiocytosis. (G) Modest expansion of hepatic sinusoidal spaces with mixed inflammatory cells and sinusoidal leukocytosis. (H) Loss of a defined marginal zone in splenic white pulp. (I) Extensive expansion of alveolar septa with mixed inflammatory cells. (J) Well-defined glial nodule within the brainstem. (K to O) Survivor 2. (K) No significant lesions (NSL) in lymph node. (L) NSL in liver. (M) NSL in spleen. (N) NSL in lung. (O) NSL in brainstem. All images were captured at 20× magnification; tissue samples were stained with hematoxylin and eosin.

FIG 3

Representative immunohistochemistry (IHC) for anti-BDVD antigen in lesions among BDBV-infected macaques. (A to E) Fatal 4 (animal identifier). (A) IHC-positive histiocytes within the expanded sinuses of a lymph node. (B) Rare IHC-positive hepatocytes and scattered IHC-positive Kupffer cells. (C) IHC-positive macrophages throughout the splenic white pulp. (D) IHC-positive alveolar macrophages and mononuclear cells within the alveolar septa. (E) Scattered IHC-positive ependymal cells of the choroid plexus. (F to J) Fatal 2. (F) IHC-positive histiocytes within the expanded sinuses of a lymph node. (G) Rare IHC-positive Kupffer cells. (H) IHC-positive macrophages throughout the splenic white pulp. (I) IHC-positive mononuclear cells within the alveolar septa. (J) Focal IHC-positive glial nodule within the brainstem. (K to O) Survivor 2. (K) No significant immunolabeling (NSI) in lymph node. (L) NSI in liver. (M) NSI in spleen. (N) NSI in lung. (O) NSI in brainstem. All images captured at 20× magnification; IHC-positive cells are brown.

FIG 4

Comparison of transcriptional changes in surviving versus nonsurviving rhesus macaques infected with BDBV. (A) Shown are principal component (PC) analyses of all normalized transcripts delineated by disposition (left; fatal subjects [n = 4], survivors with mild-to-moderate disease [M] [n = 2], and survivors with severe disease [S] [n = 3]) and disease stage (right; early [6 to 8 dpi], middle [10 to 11 dpi], and late [13 to 19 dpi]). (B) Heatmap depicting the overall most differentially expressed transcripts in survivor versus fatal subjects. Only differentially expressed transcripts with a Benjamini-Hochberg false discovery rate (FDR)-corrected P value of less than 0.05 are shown. (C) Pearson correlation plots for myeloid-derived suppressor cell-related (S100A8 and S100A9) and B cell receptor-affiliated (CD79A and CD79B) transcripts. (D) Heatmap of the most significantly upregulated and downregulated upstream regulators in survivor versus fatal subjects. (E) Network plots depicting gene clusters associated with BDBV infection in each fatal (left) or survivor (right) data set. Networks of enriched terms are colored by cluster identification; nodes that share the same cluster ID are typically close to each other. Terms with a P value of <0.01, a minimum count of 3, and an enrichment factor of >1.5 are collected and grouped into clusters based on their membership similarities. The most statistically significant term within a cluster is chosen to represent the cluster. (F) Heatmap of the most significantly upregulated and downregulated canonical pathways in survivor versus fatal subjects. Functional enrichment of all normalized transcripts at early, middle, and late stages of disease was accomplished using Ingenuity Pathway Analysis. For heatmaps, red indicates high expression, blue indicates low expression, and white indicates no difference in expression. All canonical pathways and upstream regulators had a Benjamini-Hochberg false discovery rate (FDR)-corrected P value of <0.05. M, survivor with mild-to-moderate disease; S, survivor with severe disease; PC1, principal component 1; PC2, principal component 2.

FIG 5

Immune cell type profiling of survivor and fatal samples. (A) Overall respective cell type quantities for each disease stage and data set (M, mild-to-moderate disease survivor [n = 3]; S, severe disease survivor [n = 2]) compared to the fatal group (n = 4), determined using the NanoString nSolver Advanced Analysis plugin. (B) Comparative heatmap of predicted immune cell type frequencies in each group at early, middle, and late stages of disease using CIBERSORT deconvolution software. The algorithm infers a relative increase (red) or decrease (blue) for each cell subset.

FIG 6

Antibody titers and levels of thrombosis-associated markers in BDBV-infected macaques. (A and B) BDBV glycoprotein-specific immunoglobulin G (IgG) (A) and immunoglobulin M (IgM) (B) titers in serum samples of fatal (n = 4) and survivor (n = 6 [n = 4 with mild-to-moderate disease and n = 2 with severe disease]) subjects were measured at early, middle, and late stages of disease. (C) Neutralizing antibody titers in BDBV-infected macaques determined by plaque reduction neutralization tests. (D to F) Fold change increases or decreases in thrombosis-associated markers in BDBV-infected macaques for each group. Each replicate is shown with symbols denoting data for individual subjects (n = 10 biologically independent animals/samples in a single experiment); each bar and error bar represents the group mean value ± SEM. Statistical significance was determined using two-way ANOVA with Greenhouse-Geisser correction. *, P < 0.05; **, P < 0.001; ***, P < 0.0001; ****, P < 0.00001.

FIG 7

Comparison of plasma cytokine/chemokine levels in BDBV-infected rhesus macaques. (A to K) Fold change increases or decreases in selected cytokines, chemokines, or other soluble mediators grouped by disposition (n = 4 fatal and n = 6 survivor [n = 4 with mild-to-moderate disease and n = 2 with severe disease] subjects) and disease stage. Each replicate is shown with symbols denoting data for individual subjects (n = 10 biologically independent animals/samples in a single experiment); each bar and error bar represents the group mean value ± SEM. Statistical significance was determined using two-way ANOVA with Greenhouse-Geisser correction. *, P < 0.05; **, P < 0.001; ***, P < 0.0001; ****, P < 0.00001.