High-throughput sequencing of islet-infiltrating memory CD4+ T cells reveals a similar pattern of TCR Vβ usage in prediabetic and diabetic NOD mice

Abstract

Autoreactive memory CD4(+) T cells play a critical role in the development of type 1 diabetes, but it is not yet known how the clonotypic composition and TCRβ repertoire of the memory CD4(+) T cell compartment changes during the transition from prediabetes to diabetes. In this study, we used high-throughput sequencing to analyze the TCRβ repertoire of sorted islet-infiltrating memory CD4(+)CD44(high) T cells in 10-week-old prediabetic and recently diabetic NOD mice. We show that most clonotypes of islet-infiltrating CD4(+)CD44(high) T cells were rare, but high-frequency clonotypes were significantly more common in diabetic than in prediabetic mice. Moreover, although the CD4(+)CD44(high) TCRβ repertoires were highly diverse at both stages of disease development, dominant use of TRBV1 (Vβ2), TRBV13-3 (Vβ8.1), and TRBV19 (Vβ6) was evident in both prediabetic and diabetic mice. Our findings strongly suggest that therapeutic targeting of cells specifically expressing the dominant TCRβ might reduce pancreatic infiltration in prediabetic mice and attenuate the progression to diabetes.

Figure 1. Distribution of unique CD4⁺CD44^high clonotypes in prediabetic and diabetic NOD mice.

The frequency of individual TCRβ clonotypes was determined in CD4⁺CD44^high cells sorted from the islets of 10-week-old prediabetic (n = 7) and newly diabetic (n = 9) NOD female mice. Distribution of raw (non-normalized) frequencies among the unique TCRβ clonotypes in the memory repertoire is shown for each prediabetic (A) and diabetic (B) mouse. Each unique TCRβ clonotype is shown as a dot. The number of unique TCRβ clonotypes for each mouse is shown in Table 1.

Figure 2. High-frequency CD4⁺CD44^high clonotypes are more common in the islets of diabetic than prediabetic NOD mice.

CD4⁺CD44^high cells from the islets of 10-week-old prediabetic (n = 7) and newly diabetic (n = 9) NOD female mice were sorted and subjected to high-throughput sequencing of the CDR3β regions. (A) Scatter plot showing the normalized frequencies of all CDR3β sequences represented by unique TCRβ clonotypes (15,923 in prediabetic mice and 1,471 in diabetic mice, shown as dots). Clonotypes accounting for <0.02% and ≥0.02% of all CDR3βs were defined as rare and high-frequency clones, respectively. (B) The proportion of rare (white bar) and high-frequency (gray bar) clonotypes in prediabetic and diabetic mice. Rare clonotypes are significantly more common than high-frequency clonotypes in prediabetic (p = 0.0006) and diabetic mice (p<0.0001). Rare clonotypes are significantly more frequent in prediabetic than in diabetic mice (p = 0.0118), and high-frequency clonotypes are significantly more frequent in diabetic than in prediabetic mice (p = 0.0118). (C) The proportion of CDR3β sequences derived from rare (white bar) and high-frequency (gray bar) clonotypes in prediabetic or diabetic mice. The proportion of CDR3β sequences derived from rare clonotypes is significantly higher in prediabetic than in diabetic mice (p = 0.0168). The proportion of CDR3β sequences derived from high-frequency clonotypes is significantly higher in diabetic than in prediabetic mice (p = 0.0168). The proportion of CDR3β sequences derived from high-frequency clonotypes is significantly higher than sequences derived from rare clonotypes in diabetic mice (p = 0.0399).

Figure 3. Islet-infiltrating CD4⁺CD44^high cells from prediabetic and diabetic mice exhibit the same dominant TRBV gene restriction.

Islet-infiltrating CD4⁺CD44^high cells were sorted from prediabetic (A, C, E) and diabetic (B, D, F) mice and the CDR3β regions were sequenced. The average frequency of TRBV gene segment usage is shown for all clonotypes (A and B), rare clonotypes (C and D), and high-frequency clonotypes (E and F). Each TRBV gene segment is represented by a slice proportional to its frequency. TRBV4, 20, 21, 22, 23, 24, and 30 genes were used at very low frequencies and are not shown in the graphs.

Figure 4. High variation in TRBJ usage in islet-infiltrating CD4⁺CD44^high T cell clonotypes.

The average frequency of TRBJ gene usage by all CD4⁺CD44^high clonotypes (A) and high-frequency clonotypes (B) from prediabetic mice (**p = 0.0054 for TRBJ1-5*1) and diabetic mice. The average frequency of TRBJ gene usage by all CD4⁺CD44^high clonotypes and high-frequency clonotypes from (C) prediabetic mice (**p = 0.0081 for TRBJ2-7) and (D) diabetic mice (*p = 0.019 for TRBJ1-1, *p = 0.04 for TRBJ2-3, and *p = 0.022 for TRBJ2-5). All data are the mean ± SEM.

Figure 5. CD4⁺CD44^high clonotypes from prediabetic and diabetic mice have similar CDR3β length distribution.

(A) The distribution of CDR3 lengths for all unique CD4⁺CD44^high clonotypes are presented as means ± SEM of prediabetic (white bar) and diabetic (black bar) mice. The distribution of CDR3β lengths are shown for individual prediabetic mice (B) or diabetic mice (C).

Figure 6. High-throughput sequencing analysis reveals monoclonal expansions in islet-infiltrating CD4⁺CD44^high cells from diabetic mice.

The frequencies of V and J gene segment usage within total CDR3β sequences amplified from islet-infiltrating CD4⁺CD44^high cells isolated from individual diabetic. Each panel represents one mouse.

Figure 7. Absence of monoclonal expansions in prediabetic mice.

The frequencies of V and J gene segment usage within total CDR3β sequences amplified from islet-infiltrating CD4⁺CD44^high cells isolated from individual prediabetic (B) mice. Each panel represents one mouse.