Hyaluronan deposition in islets may precede and direct the location of islet immune-cell infiltrates

Abstract

Aims/hypothesis: Substantial deposition of the extracellular matrix component hyaluronan (HA) is characteristic of insulitis in overt type 1 diabetes. We investigated whether HA accumulation is detectable in islets early in disease pathogenesis and how this affects the development of insulitis and beta cell mass.

Methods: Pancreas tissue from 15 non-diabetic organ donors who were positive for islet autoantibodies (aAbs) and from 14 similarly aged aAb^- control donors were examined for the amount of islet HA staining and the presence of insulitis. The kinetics of HA deposition in islets, along with the onset and progression of insulitis and changes in beta cell mass, were investigated in BioBreeding DRLyp/Lyp rats (a model of spontaneous autoimmune diabetes) from 40 days of age until diabetes onset.

Results: Abundant islet HA deposits were observed in pancreas tissues from n = 3 single- and n = 4 double-aAb⁺ donors (aAb⁺HA^high). In these seven tissues, the HA-stained areas in islets measured 1000 ± 240 μm² (mean ± SEM) and were fourfold larger than those from aAb^- control tissues. The aAb⁺HA^high tissues also had a greater prevalence of islets that were highly rich in HA (21% of the islets in these tissues contained the largest HA-stained areas [>2000 μm²] vs less than 1% in tissues from aAb^- control donors). The amount of HA staining in islets was associated with the number of aAbs (i.e. single- or double-aAb positivity) but not with HLA genotype or changes in beta cell mass. Among the seven aAb⁺HA^high tissues, three from single- and one from double-aAb⁺ donors did not show any islet immune-cell infiltrates, indicating that HA accumulates in aAb⁺ donors independently of insulitis. The three aAb⁺HA^high tissues that exhibited insulitis had the largest HA-stained areas and, in these tissues, islet-infiltrating immune cells co-localised with the most prominent HA deposits (i.e. with HA-stained areas >2000 μm²). Accumulation of HA in islets was evident prior to insulitis in 7-8-week-old presymptomatic DRLyp/Lyp rats, in which the islet HA-stained area measured 2370 ± 170 μm² (mean ± SEM), which was threefold larger than in 6-week-old rats. This initial islet HA deposition was not concurrent with beta cell loss. Insulitis was first detected in 9-10-week-old rats, in which the HA-stained areas were 4980 ± 500 μm². At this age, the rats also exhibited a 44% reduction in beta cell mass. Further enlargement of the HA-positive areas (mean ± SEM: 7220 ± 880 μm²) was associated with invasive insulitis. HA deposits remained abundant in the islets of rats with destructive insulitis, which had lost 85% of their beta cells.

Conclusions/interpretation: This study indicates that HA deposition in islets occurs early in type 1 diabetes and prior to insulitis, and points to a potential role of HA in triggering islet immune-cell infiltration and the promotion of insulitis.

Keywords: Autoantibodies; Extracellular matrix; Hyaluronan; Insulitis; Islet; Type 1 diabetes.

Fig. 1

Islet HA deposits in aAb⁺ donors. HA staining (brown) in islets from (a) aAb⁻ and (b) aAb⁺ donors. Yellow arrows indicate HA at the islet periphery, while blue arrows indicate HA within the islet. Arrowheads point to the islet border. Scale bars, 50 μm. (c, d) Violin plots of HA⁺ areas measured in islets from each donor for aAb⁻ donors (c) and aAb⁺ donors (d; light red, single-aAb⁺; dark red, double-aAb⁺). Islet HA⁺ areas are shown on a log₁₀ scale. The dotted horizontal lines indicate the median of HA⁺ area in the control aAb⁻ group

Fig. 2

HA accumulates in islets from a subset of aAb⁺ donors. HA staining (brown) in islets from (a) aAb⁻, (b, c) single-aAb⁺ and (d) double-aAb⁺ donors. Arrowheads point to the islet border. Scale bars, 50 μm. (e) Morphometric quantification of islet HA⁺ areas, shown on a log₁₀ scale. Each circle denotes an individual donor; blue, aAb⁻; light red, single-aAb⁺; dark red, double-aAb⁺. Data are mean values of islet HA⁺ areas for each individual donor. The numbers (1–7) indicate the aAb⁺HA^high tissues ranked according to the size of islet HA⁺ areas. The dotted horizontal line indicates the upper cut-off value (mean of HA⁺ area in aAb⁻control + 3 SD). In total, 4598, 3210 and 2272 islets from aAb⁻, single-aAb⁺ and double-aAb⁺ donor tissues were analysed, respectively. (f) Islet HA⁺ area size distribution. The pie charts represent the percentage of islets with HA⁺ areas falling within each of the HA⁺ area size categories. Islets were analysed in aAb⁻ tissues (n=4598 islets) and in aAb⁺HA^high tissues from n=3 single-aAb⁺ donors (n=982 islets) and n=4 double-aAb⁺ donors (n=1828 islets). *p<0.001,single-aAb⁺HA^high vs controls, and double-aAb⁺HA^high vs single-aAb⁺HA^high or controls; Mann–Whitney U test

Fig. 3

Islet HA accumulation takes place in the absence of insulitis. HA staining (brown) in islets from (a) aAb⁻ and (b–d) aAb⁺ donors. Arrowheads point to the islet border. The area of insulitis in (d) is shown magnified in the inset. (e–h) Adjacent sections of the islets shown in (a–d), respectively, stained for LCA (brown) and SYN (red). Scale bars, 50 μm. (i) Prevalence of islets with LCA⁺ cell infiltrates plotted as a function of islet HA⁺ areas, shown on a log₁₀ scale. Each circle denotes mean values for an individual donor. The dotted lines indicate the upper cut-off values (mean + 3 SD) of the measurements obtained from the aAb⁻ controls. The numbers (1–7) indicate the aAb⁺HA^high tissues ranked according to the size of islet HA⁺ areas. Blue, aAb⁻; light red, single-aAb⁺; dark red, double-aAb⁺

Fig.4

Insulitis occurs exclusively in the islet regions containing the largest HA deposits. Violin plots of HA⁺ areas in islets without (LCA⁻) or with (LCA⁺) immune cells for the (a) aAb⁻, (b) aAb⁺HA^low, (c) aAb⁺HA^high and (d) aAb⁺HA^highLCA^high tissues. Islet HA⁺ areas are shown on a log₁₀ scale. The total number of LCA⁻ or LCA⁺ islets in each group is indicated. LCA⁺ cells/100 islets: 3 ± 1 in (a) and (b); 5 ± 2 in (c); 47 ± 36 in (d)

Fig. 5

Relationship between islet HA and beta cell mass. Immunohistochemical staining for insulin (brown) in islets from (a) control aAb⁻ and (b) aAb⁺HA^high tissues. Scale bars, 100 μm. (c) Individual measurements of beta cell mass plotted as a function of islet HA⁺ area, shown on a log₁₀ scale. (d) Insulin⁺ area relative to pancreas section area. Data are presented as the mean values of measurements for each individual donor. In (d), the horizontal lines represent the mean value in each group. The numbers (1–7) indicate the aAb⁺HA^high tissues ranked according to the size of islet HA⁺ areas. Blue, aAb⁻; light red, single-aAb⁺; dark red, double-aAb⁺

Fig. 6

(a) Islet HA⁺ areas, insulitis and beta cell mass in individual aAb⁺ donors. Islet HA⁺ areas are shown as mean values obtained from the islets examined in each pancreas. Between 250 and 560 islets were analysed per pancreas. For ‘islet HA⁺ area’ and ‘beta cell mass’, the size of each red circle is proportional to the mean value, which is indicated within the circle. The pie charts represent the percentage of islets with HA⁺ areas falling within each of the HA⁺ area size categories or the percentage of islets with LCA⁺ cells. HLA genotype and aAb status are also shown for each donor. See ESM Fig. 6 for further analyses. (b) Whole slide images of pancreas tissue sections from donor 1 (aAb⁺HA^high) and donor 13 (aAb⁺HA^low). The blue, green, and red circles show the islet border and the indicate islets with HA-stained areas ≤500 μm², 501–2000 μm² and >2000 μm², respectively. Scale bars, 2000 μm. N/D, not determined

Fig. 6

(a) Islet HA⁺ areas, insulitis and beta cell mass in individual aAb⁺ donors. Islet HA⁺ areas are shown as mean values obtained from the islets examined in each pancreas. Between 250 and 560 islets were analysed per pancreas. For ‘islet HA⁺ area’ and ‘beta cell mass’, the size of each red circle is proportional to the mean value, which is indicated within the circle. The pie charts represent the percentage of islets with HA⁺ areas falling within each of the HA⁺ area size categories or the percentage of islets with LCA⁺ cells. HLA genotype and aAb status are also shown for each donor. See ESM Fig. 6 for further analyses. (b) Whole slide images of pancreas tissue sections from donor 1 (aAb⁺HA^high) and donor 13 (aAb⁺HA^low). The blue, green, and red circles show the islet border and the indicate islets with HA-stained areas ≤500 μm², 501–2000 μm² and >2000 μm², respectively. Scale bars, 2000 μm. N/D, not determined

Fig. 7

Islet HA accumulation precedes insulitis in presymptomatic DRLyp/Lyp rats. (a–e) HA staining (brown) in islets from (a) diabetes-resistant DRLyp/+ and (b–e) diabetes-prone DRLyp/Lyp rats. Arrowheads point to the islet border. (f–j) SYN (red) staining of islets from (f) DRLyp/+ and (g–j) DRLyp/Lyp rats. Scale bars, 50 μm. (k) Islet HA⁺ areas (mean in 300–400 islets/group), islet HA⁺ area size distribution, insulitis and beta cell mass in DRLyp/+ and DRLyp/Lyp rats. For ‘islet HA⁺ area’ and ‘beta cell mass’, the size of each circle is proportional to the mean value, which is indicated within the circle. The pie charts in represent the percentage of islets with HA⁺ areas falling within each of the HA area size categories or the proportion of islets with insulitis grade 0, 1, 2 or 3 in the rats within each age group (see ESM Table 3 for further details)