Primary and secondary hemophagocytic lymphohistiocytosis have different patterns of T-cell activation, differentiation and repertoire

Abstract

Hemophagocytic lymphohistiocytosis (HLH) is a life-threatening inflammatory syndrome characterized by hyperactivation of lymphocytes and histiocytes. T cells play a key role in HLH pathogenesis, but their differentiation pattern is not well characterized in patients with active HLH. We compared T-cell activation patterns between patients with familial HLH (1°HLH), 2°HLH without apparent infectious trigger (2°HLH) and 2°HLH induced by a viral infection (2°V-HLH). Polyclonal CD8⁺ T cells are highly activated in 1°HLH and 2°V-HLH, but less in 2°HLH as assessed by HLA-DR expression and marker combination with CD45RA, CCR7, CD127, PD-1 and CD57. Absence of increased HLA-DR expression on T cells excluded active 1° HLH with high sensitivity and specificity. A high proportion of polyclonal CD127^- CD4⁺ T cells expressing HLA-DR, CD57, and perforin is a signature of infants with 1°HLH, much less prominent in virus-associated 2°HLH. The similar pattern and extent of CD8⁺ T-cell activation compared to 2° V-HLH is compatible with a viral trigger of 1°HLH. However, in most 1°HLH patients no triggering infection was documented and the unique activation of cytotoxic CD4⁺ T cells indicates that the overall T-cell response in 1°HLH is different. This may reflect different pathways of pathogenesis of these two HLH variants.

Keywords: HLA-DR; Hemophagocytic lymphohistiocytosis; Perforin; Primary HLH; Secondary HLH; T-cell activation.

Figure 1

Activation state of CD4+ and CD8+ T cells by HLA‐DR expression and soluble CD25 levels. (A‐C) PBMCs from healthy controls (Ctr), patients with 2°HLH with viral infection (2°V‐HLH) or without viral infection (2°HLH) and patients with primary HLH (1°HLH) were analyzed by flow cytometry. PBMCs of all groups were analyzed once for each individual, no replicates were performed. Data were pooled and grouped from Ctr: n = 43, 2°HLH: n = 30, 2°V‐HLH: n = 32, 1°HLH: n = 31 donor samples. (A) Representative dot‐plots showing HLA‐DR expression by flow cytometry on CD8⁺ and CD8⁻ T cells from Ctr, 2°V‐HLH and patients with 1°HLH. Plots were gated on CD3⁺ lymphocytes. (B) Percentage of HLA‐DR⁺ CD8⁺ T cells among total CD3⁺CD8⁺ T cells (upper panel) and HLA‐DR⁺ CD4⁺ T cells among total CD3⁺CD4⁺ T cells (lower panel) in the indicated groups of patients. Differences between the three groups were analyzed by Mann–Whitney U test. Only significant differences are shown (defined as ***for p ≤ 0.0005 and ****for p < 0.0001). (C) ROC curves showing the cut‐off for optimal sensitivity and specificity for % HLA‐DR⁺ CD8⁺ (upper panel) and CD4⁺ T cells (lower panel) for patients with any form of HLH and controls. (D) Serum level of soluble CD25 (sCD25) were measured by Chemiluminescence Immunoassay (CLIA) patients with 2°HLH (n = 23), 2°V‐HLH (n = 25) and 1°HLH (n = 27) given in U/mL. Differences between the three groups were analyzed by Mann–Whitney U test mean +/− SD is shown. Only significant differences are shown (defined as ***for p ≤ 0.0005 and ****for p < 0.0001). (E) Correlation of sCD25 levels and percentage of HLA‐DR expressing CD8⁺ T cells (upper panel) or CD4⁺ T cells (lower panel) in patients with 2°V‐HLH. (F) Correlation of sCD25 and percentage of HLA‐DR expressing T cells in patients with 1°HLH. Linear regression analysis was performed for all data pairs and the level of significance is indicated.

Figure 2

CD8⁺ T‐cell differentiation is similar in 1° and 2° V‐HLH. PBMCs from healthy controls (Ctr), asymptomatic patients with a genetic defect (1°asym), patients with 2°HLH without viral infection (2°HLH) or with viral infection (V‐2°HLH) and patients with primary HLH (1°HLH) were analyzed by flow cytometry. PBMCs of all groups were analyzed once for each individual, no replicates are performed. Data were pooled and grouped from Ctr (n = 9), 1° asym (n = 4), 2°HLH (n = 6), 2°V‐HLH (n = 6) and 1°HLH (n = 18) donor samples. (A) Representative dot‐plots of naïve (CD45RA⁺,CCR7⁺), central memory TCM (CD45RA⁻, CCR7⁺), effector‐memory TEM (CD45RA⁻,CCR7⁻) and effector TE (CD45RA⁺,CCR7⁻) CD8⁺ T cells of a healthy control (Ctr), a patient with 2°V‐HLH and a patient with 1°HLH. Plots are gated on CD3⁺CD8⁺ T cells. (B) Summary of differentiation state of CD3⁺ CD8⁺ (upper panel) and CD3⁺ CD4⁺ (lower panel). The mean fraction (+/−SD) of each of the 4 cell populations in a given patient group is shown.

Figure 3

A high percentage of CD127⁻PD1⁺CD57⁺ CD4⁺ T cells are characteristic for 1°HLH. PBMCs from healthy controls (Ctr), asymptomatic patients with a genetic defect (1°asym), patients with 2°HLH without viral infection (2°HLH) or with viral infection (V‐2°HLH) and patients with primary HLH (1°HLH) were analyzed by flow cytometry. PBMCs of all groups were analyzed once for each individual, no replicates were performed. Data were pooled and grouped from Ctr (n = 9), 1° asym (n = 4), 2°HLH (n = 6), 2°V‐HLH (n = 6) and 1°HLH (n = 18) donor samples. (A) Representative dot‐plots of CD127 and CD45RA expression gated on CD3⁺CD8⁺ T cells in a healthy control (Ctr), a patient with 2°V‐HLH and a patient with 1°HLH. (B) Percentage of CD127⁻ cells among CD8⁺ (upper panel) and CD4⁺ (lower panel) T cells in the indicated patient groups. (C) Representative dot‐plots of PD1 and CD45RA expression gated on CD3⁺CD8⁺ T cells in a healthy control (Ctr), a patient with 2°V‐HLH and a patient with 1°HLH. (D) Percentage of PD1⁺ cells among CD3⁺CD8⁺ (upper panel) and CD3⁺CD4⁺ (lower panel) T cells in the indicated control and patient groups. (E) Representative dot‐plots of CD57 expression gated on CD3⁺CD8⁺ T cells in a healthy control (Ctr), a patient with 2°V‐HLH and a patient with 1°HLH. Plots are gated on CD3⁺ lymphocytes (F) Percentage of CD57⁺ cells among CD8⁺ (upper panel) and CD4⁺ (lower panel) T cells in the indicated control and patient groups. Significance was defined by Mann–Whitney U test as *for p ≤0.05, **for p ≤0.005, ***for p ≤ 0.0005 and ****for p < 0.0001. Data are shown as mean +SD and are from a single experiment.

Figure 4

A high percentage of perforin‐expressing CD4 ⁺ T cells is characteristic for 1° HLH. PBMCs from healthy controls (Ctr), patients with 2°HLH without viral infection (2°HLH) or with viral infection (V‐2°HLH) and patients with primary HLH (1°HLH) were analyzed by flow cytometry. PBMCs of all groups were analyzed once for each individual, no replicates were performed. Data were pooled and grouped from Ctr (CD8⁺ n = 3 and CD4⁺ n = 7 (4 beyond 1year), 2°HLH (n = 5), 2°V‐HLH (n = 8 (2 beyond 1year) and 1°HLH (n = 10) donor samples. (A) Representative dot‐plots of intracellular perforin versus CCR7 expression in CD3⁺CD8⁺ (left panels) and CD3⁺CD4⁺ (right panels) T cells in a healthy control (Ctr), a patient with 2°V‐HLH and a patient with 1°HLH. (B) Percentage of perforin⁺ cells among CD3⁺CD8⁺ (upper panel) and CD3⁺CD4⁺ (lower panel) T cells in the indicated control and patient groups (for 1° HLH, the analysis was restricted to MUNC13‐4 and MUNC18‐2 deficient patients and perforin deficient patients with residual protein expression). Data are shown as mean +SD and are from a single experiments. Differences between the three HLH groups were analyzed by Mann–Whitney U test U test as *for p ≤0.05, **for p ≤0.005, ***for p ≤ 0.0005 and ****for p < 0.0001. Mean +/− SD is shown. (C) Representative dot‐plot of CD27 and CD28 expression gated on CCR7⁻ CD8⁺ (upper panel) and CCR7⁻ CD4⁺ (lower panel) T cells in two patients representing different patterns of expression: one patient with MUNC13‐4 deficiency (P1; same patient as in A) and one patient with perforin deficiency (with residual protein expression (P2).

Figure 5

Moderate perturbation of the V beta repertoire analyses among CD8 ⁺ and CD4 ⁺ T cells in 1° and 2°V‐HLH (A, B) Relative expression of the 24 most common TCR V beta chains on CD8⁺ (A) and CD4⁺ (B) T cells. T cells were analyzed by flow cytometry. In the individual charts, each vertical bar represents a fraction of T cells expressing one of the 24 chains, “others” indicates the remaining T‐cell population not covered by the 24 antibodies. The bars are ordered in increasing size and not according to the V beta chain they represent. The chain numbers of the 2 most abundant V beta expressing populations are indicated in the individual charts. Below each chart, the number of V beta populations extending to more than 2SD above the mean value of controls is indicated. The chart labeled Ctr represents the mean value of controls, the other charts in the upper lane represent three infants with 2°HLH, the charts in the middle lane represent 4 infants with 2°V‐HLH and the charts in the lower lane six infants with 1°HLH.