Blood immune cell profiling in adults with longstanding type 1 diabetes is associated with macrovascular complications

Abstract

Aims/hypothesis: There is increasing evidence for heterogeneity in type 1 diabetes mellitus (T1D): not only the age of onset and disease progression rate differ, but also the risk of complications varies markedly. Consequently, the presence of different disease endotypes has been suggested. Impaired T and B cell responses have been established in newly diagnosed diabetes patients. We hypothesized that deciphering the immune cell profile in peripheral blood of adults with longstanding T1D may help to understand disease heterogeneity.

Methods: Adult patients with longstanding T1D and healthy controls (HC) were recruited, and their blood immune cell profile was determined using multicolour flow cytometry followed by a machine-learning based elastic-net (EN) classification model. Hierarchical clustering was performed to identify patient-specific immune cell profiles. Results were compared to those obtained in matched healthy control subjects.

Results: Hierarchical clustering analysis of flow cytometry data revealed three immune cell composition-based distinct subgroups of individuals: HCs, T1D-group-A and T1D-group-B. In general, T1D patients, as compared to healthy controls, showed a more active immune profile as demonstrated by a higher percentage and absolute number of neutrophils, monocytes, total B cells and activated CD4+CD25+ T cells, while the abundance of regulatory T cells (Treg) was reduced. Patients belonging to T1D-group-A, as compared to T1D-group-B, revealed a more proinflammatory phenotype characterized by a lower percentage of FOXP3+ Treg, higher proportions of CCR4 expressing CD4 and CD8 T cell subsets, monocyte subsets, a lower Treg/conventional Tcell (Tconv) ratio, an increased proinflammatory cytokine (TNFα, IFNγ) and a decreased anti-inflammatory (IL-10) producing potential. Clinically, patients in T1D-group-A had more frequent diabetes-related macrovascular complications.

Conclusions: Machine-learning based classification of multiparameter flow cytometry data revealed two distinct immunological profiles in adults with longstanding type 1 diabetes; T1D-group-A and T1D-group-B. T1D-group-A is characterized by a stronger pro-inflammatory profile and is associated with a higher rate of diabetes-related (macro)vascular complications.

Keywords: heterogeneity; hierarchical classification; immune cell profile; machine learning; multiparameter flow cytometry; type 1 diabetes mellitus.

Figure 1

Different circulating immune cell subsets in T1D and HC. (A) Biplot of the first two principal components (PCs) of percentages of immune cell subsets determined by multiparameter flow cytometry. (B) Violin plots show the univariate comparison of cell percentages between T1D (n = 242) and HC (n = 37), adjusted for the covariates age and sex. Each dot represents one individual. Grey dot indicates the mean value, and grey vertical line represents the standard deviation. *, p<0.05; **, p<0.01; ***, p<0.001; ns, not significant. mCD4, memory CD4+ cells.

Figure 2

Elastic net based ranking of circulating immune cell subsets and hierarchical clustering of T1D patients and healthy controls. The response variable of the EN model was the individual’s cohort status, i.e. HC vs T1D. The input variables were the percentages of immune cell subsets plus individual’s age and sex (see details in the Method section). The scaled variable importance value was used to rank the weight of each explanatory variables. (A, top) Variable importance plot showed all explanatory variables selected by the EN-model. The most explanatory immune cell subsets contributing to 90% of the classification (top-35) are those above the dashed line. (B, bottom) Hierarchical clustering using the top-35 most explanatory immune cell subsets revealed three clusters; one healthy control cluster and two distinct T1D-patient subgroups. The example heatmap shows the top-10 of immune cell subsets contributing to the classification. Cohort status and sex of every individual are listed on the top of heatmap. T1D (n = 242), HC (n = 37). F: female; M: male; mCD4, memory CD4+ cells.

Figure 3

Discriminating peripheral immune cell profiles classify two T1D- patient groups. Elastic net (EN) modelling of T1D patients only. The response variable of the EN model was patient’s subgroups, i.e. T1D-group-A vs T1D-group-B. The input variables were the percentages of immune cell subsets plus patient’s age and sex. (A) Variable importance plot shows all variables selected by the EN-model (33 out of 114 input explanatory variables, Y-axis). (B-E) Violin plots showing univariate comparison, adjusted for the covariates age and sex, between T1D-group-Avs T1D-group-B patients for (B) age (top) and sex (bottom), (C) top-5 explanatory immune cell subsets as selected by the EN model, (D) percentages of FOXP3+ Treg and ratio of CD4+ regulatory T cells: CD4+ conventional T cells (Treg/Tconv) based on absolute cell counts of CD4+CD25++Foxp3+ and CD4+CD25- cells, and (E) percentages of monocyte subsets. T1D-group-A (n = 50), T1D-group-B (n = 191). Grey dot indicates the mean value, and grey vertical line represents the standard deviation. **, p<0.01; ***, p<0.001.; ns, not significant. mCD4, memory CD4+ cells. F, Female; M, Male.

Figure 4

T1D-group-A shows increased pro-inflammatory cytokine production and exclusive cytokine / immune cell subset correlations. Cytokines production as measured by ELISA of PBMCs stimulated in vitro with S. aureus (top panels) or C. albicans (bottom panels). Innate cytokines TNFα and IL-10 were measured 24 hours after stimulation and adaptive cytokines IFNγ and IL-17 were measured 7 days after stimulation. (A) Violin plots show the cytokines production in T1D-group-A vs T1D-group-B. Age and sex covariates were adjusted. (B) Pearson’s correlation of the percentages of cell subsets and cytokines produced upon stimulation with S.aureus in T1D-group-A (left panel) or T1D-group-B (right panel). Dot size represents the absolute correlation coefficients and the colour scale indicates the correlation direction. Arrows indicate the mutual exclusive correlation patterns in T1D-group-A (n = 50) vs T1D-group-B (n = 191). Asterisk(s) within the dot indicate the significance of correlation p-values. *, p<0.05; **, p<0.01; ***, p<0.001; ns, not significant. memCD4, memory CD4+ cells; Mono., monocytes; EM: effector memory.

Figure 5

Higher prevalence of macrovascular disease in T1D-group-A. Bar plots showing the percentage of patients associating with or without diabetes-related complications regardless of macro-/micro-complications (left panel), only macrovascular complications (middle panel) or only microvascular complications (right panel)in T1D-group-A (n = 50) versus T1D-group-B (n = 191). The p-value of the Chi-squared test is listed on the top of the bars. Yes indicating patient associating with at least one of the corresponding complications listed on the top of the bars.