Elevated Cathepsin S Serum Levels in New-Onset Type 1 Diabetes and Autoantibody-Positive Siblings Free

Type 1 diabetes develops because of an immune-mediated loss of mass and function of insulin-producing β-cells in the pancreatic islets. During islet inflammation (insulitis), immune cell– and β-cell–secreted proinflammatory cytokines cause β-cell dysfunction and apoptosis (1). Recently, dysfunctional autophagy due to impaired lysosome function and leakage of lysosomal cathepsin proteases was identified as a downstream signaling event (2).

Studies on mice have linked the cysteine protease cathepsin S (CTSS) to the development of autoimmune diabetes, insulitis, and regulation of glucose homeostasis (3–6). Studies indicate that CTSS expression is elevated in islets/β-cells during insulitis and diabetes development (3,7,8), and CTSS plasma levels were reported to be higher in children with type 1 diabetes compared with healthy children (9). CTSS is known to have both intra- and extracellular functions (10–12). It can be induced and secreted from different cell types during inflammatory conditions and has been suggested as a biomarker in many autoimmune, inflammatory, and metabolic diseases (11–13).

In the current study, we examined the biomarker potential of CTSS in new-onset type 1 diabetes and investigated the expression and secretion of CTSS in human islets and β-cells.

Serum samples from 150 cases and 150 siblings were randomly selected from the population-based Danish Childhood Diabetes Register of children with new-onset type 1 diabetes (14). Cases and siblings were not necessarily from the same families; however, all the siblings included had a sibling diagnosed with type 1 diabetes before the age of 18 years. Two cases and two siblings were excluded because of insufficient material, one case was excluded because of reclassification as maturity-onset diabetes of the young, and one case was excluded because the blood sample was taken almost 2 years after diagnosis. All other samples were taken less than 3 months after onset (defined as the date of first insulin injection). Serum samples were stored at −80°C until analyses.

The study was performed in accordance with the Declaration of Helsinki and was approved by the Danish Ethical Committee, Copenhagen, Denmark (H-KA-20070009). Informed consent was given by all of the children, their parents, or guardians.

Autoantibodies against glutamic acid decarboxylase (GADA), islet antigen-2 (IA-2A), insulin (IAA), and zinc transporter 8 (ZnT8A) were measured by radioligand binding assays (14,– 16). Autoantibody positivity was defined by the following cutoffs: GADA >34 units/mL, IA-2A >5 units/mL, IAA >3.69 units/mL, and ZnT8A >59 units/mL.

Isolated pancreatic islets from human donors were obtained from Prodo Laboratories, Inc. (Aliso Viejo, CA), via Tebu-Bio (Paris, France). Donor information is provided in the Supplementary Material. Human islets were cultivated as previously described (7).

The human EndoC-βH5 cells (Human Cell Design, Toulouse, France) were maintained as previously described (17) and seeded with 200,000 and 50,000 cells per well in 24- and 96-well culture plates, respectively.

All islets/cells were precultured for 2–3 days in a humidified incubator at 37°C with 5% CO₂ before incubation for 24, 48, or 96 h in the presence or absence of 50 units/mL recombinant human interleukin-1β (IL-1β) (R&D Systems, Minneapolis, MN), 1,000 units/mL recombinant human interferon γ (IFNγ) (PeproTech, Rocky Hill, NJ), and/or 1,000 units/mL recombinant human tumor necrosis factor α (TNFα) (R&D Systems), or 1,000 units/mL recombinant human IFN-α (PeproTech). For the 96-h condition, media was changed after 48 h.

CTSS was measured in serum and supernatants using the Human CTSS ELISA (Thermo Scientific, Waltham, MA). The CTSS concentration was determined using a four-parameter logistic curve fit and an Infinite M200 PRO plate reader (Tecan, Männedorf, Switzerland). The intra- and interassay coefficient of variation was <10% and <12%, respectively. One sample (sibling) was excluded because of a coefficient of variation >40%.

Protein lysate preparation and immunoblotting were performed as previously described (7). All antibodies are listed in Supplementary Table 1. GAPDH was used as loading control.

RNA was extracted using the RNeasy Mini kit (Qiagen, Hilden, Germany). cDNA synthesis and real-time qPCR were performed as previously described (7). Relative expression levels were normalized to stably expressed housekeeping genes GAPDH and HPRT and presented as 2^-ΔCt values.

The CTSS expression in islet cells was extracted from publicly available single-cell RNA sequencing (RNAseq) data on human islets from nondiabetic, autoantibody-positive, type 1 diabetes, and type 2 diabetes donors using the Islet HPAP single-cell gene expression map (18). Supplementary Tables 2 and 3 detail donor information.

Data are presented with individual values and means ± SD or in box plots with lower quartile, median, upper quartile, and whiskers representing the range of the remaining data points. Cellular data were analyzed using one- or two-tailed paired or unpaired Student t test, as appropriate. Clinical data analyses included two-tailed unpaired Student t test, χ² test, and linear regression models. Single-cell RNAseq data were analyzed using one-way ANOVA with follow-up tests comparing the means of the groups. P values <0.05 were considered statistically significant. Statistical analyses were performed using GraphPad Prism 10.1.2, Microsoft Excel, and RStudio 2022.07.1 + 554. Graphs were constructed using GraphPad Prism 10.1.2.

Data sets are available upon request. No applicable resources were generated during this study.

To examine the biomarker potential of CTSS in type 1 diabetes, CTSS was analyzed in serum from 146 children with new-onset type 1 diabetes and from 147 healthy siblings. The two groups were similar according to sex, age at sampling, and ethnicity (Table 1). The mean CTSS level was 7.5% higher in the children with new-onset type 1 diabetes compared with the healthy siblings (P < 0.01) (Fig. 1A and Table 1). Five of the healthy siblings developed type 1 diabetes 1–8 years after the blood sampling, and, interestingly, a post hoc test showed that CTSS was higher in these preonset siblings compared with the other two groups (P < 0.05) (Fig. 1B and Table 1). Linear regression analysis showed that CTSS was inversely associated with age at sampling in both healthy siblings (P = 0.0010) (Fig. 1C) and new-onset children (P = 0.0002) (Fig. 1D). No associations were observed between CTSS and glycemic control 1 year after diagnosis, as evaluated by continuous or categorized HbA_1c (19) (Fig. 1E and F) (Supplementary Table 4).

Because CTSS is elevated in other autoimmune diseases (11,12) and has been associated with autoantibody levels in rheumatoid arthritis (20), we examined whether CTSS was associated with autoantibody status. In the healthy siblings, the CTSS level was significantly higher in autoantibody-positive compared with autoantibody-negative children, as evaluated by the presence of GADA, IA-2A, IAA, and/or ZnT8A autoantibodies (P = 0.0065) (Fig. 1G). This association was, however, not present in the children with new-onset type 1 diabetes, evaluated by the presence of GADA and/or IA-2A autoantibodies (Fig. 1H). The CTSS level appeared to be higher in the two ZnT8A-positive siblings compared with the other autoantibody-positive siblings (Fig. 1I).

We recently identified CTSS to be highly upregulated in human islets and β-cell lines exposed to proinflammatory cytokines (7). Therefore, the cytokine-mediated induction of CTSS was examined at the protein level in two human in vitro models. In EndoC-βH5 cells, CTSS had a low basal mRNA expression and was highly upregulated by IFNγ without further induction by IL-1β(+TNFα) (P < 0.001) (Fig. 2A); concordantly, CTSS protein was undetectable at baseline and highly induced by IL-1β+IFNγ+TNFα (P < 0.01) (Fig. 2B). Likewise, CTSS was present in the culture media from EndoC-βH5 cells after exposure to IL-1β+IFNγ+TNFα (P < 0.05) (Fig. 2C). In human islets, CTSS mRNA was highly induced by IL-1β+IFNγ but only modestly induced by IFNα (P < 0.01) (Fig. 2D). Induction of CTSS protein was confirmed after exposure to IL-1β+IFNγ(+TNFα) (P < 0.01) (Fig. 2E). CTSS was present in the culture media of both treated and untreated human islets and tended to be increased upon exposure to IL-1β+IFNγ(+TNFα) (P = 0.087) (Fig. 2F).

To gain knowledge of the cell type(s) responsible for the CTSS expression in human islets, we investigated the single-cell CTSS expression in islets from nondiabetic, autoantibody-positive, type 1 diabetes, and type 2 diabetes donors using publicly available single-cell RNAseq data. The highest CTSS expression level was observed in macrophages (Supplementary Table 5); however, the type 1 diabetes–associated induction in CTSS was exclusive for the β-cells (P < 0.05) (Fig. 2G–I and Supplementary Table 6).

In the current study, we explored the biomarker potential of CTSS in type 1 diabetes. We found a 7.5% increase in the CTSS serum level in children with new-onset type 1 diabetes compared with healthy siblings. This supports a previous study reporting ∼30% and ∼50% increased CTSS plasma levels in children with new-onset and long-standing type 1 diabetes, respectively, as compared with healthy children (9). The more modest increase observed in our study may be due to the comparison with first-degree relatives rather than unrelated healthy controls. The different sample and assay types applied may also add to the observed differences. Our data showing higher CTSS levels in autoantibody-positive and preonset siblings indicate that circulating CTSS increases during early disease development. The lack of association between CTSS and autoantibody status in the children with new-onset type 1 diabetes is likely due to most of the children being autoantibody-positive.

Elevated circulating CTSS levels have also been reported in type 2 diabetes (21,22), and CTSS has been proposed as a link between obesity and diabetes, for example, via adipose tissue–derived inflammatory activity. Noteworthy is that no change in adipocyte-secreted CTSS was demonstrated upon exposure to plasma from children with type 1 diabetes (9). We speculate whether the increased circulating CTSS levels observed in type 1 diabetes derive, at least in part, from inflammation-induced secretion from β-cells (Fig. 3). Our observation that CTSS is induced and secreted from β-cells under inflammatory conditions supports this, and is in line with previous studies demonstrating higher CTSS expression in nonobese diabetic (NOD) mouse islets with insulitis (3) and in FACS-sorted β-cells from donors with type 1 diabetes (8). Secretion of cathepsins via lysosomal exocytosis has previously been coupled to their overexpression (10); however, it is currently unknown whether the observed cytokine-induced secretion of CTSS from β-cells is exocytosis mediated.

In the current study, we report that CTSS is markedly upregulated by the diabetogenic cytokines IL-1β+IFNγ(+TNFα) in human islets, and only moderately induced by the early-response cytokine IFNα. It is unclear which cytokine is responsible for the observed induction. In EndoC-βH5 cells, CTSS is mainly induced by IFNγ, whereas several cytokines can induce CTSS expression in other cells, for example, IL-1β, IFNγ, and TNFα (13). IFNγ has been found to induce CTSS transcription through an IFN-stimulated response element in the CTSS promoter region via the transcriptional regulator IFN regulatory factor 1 (12,23). Furthermore, both IFNγ and TNFα can induce CTSS expression via adenosine-to-inosine RNA editing by adenosine deaminase acting on RNA 1 (ADAR1) (24). Noteworthy, a recent study showed that ADAR1 is induced by proinflammatory cytokines in human islets and β-cells and that it regulates the β-cell transcriptome during inflammation (25).

The induction of CTSS by cytokines may be anchored exclusively to one or more types of islet cells. Publicly available single-cell RNAseq data on human islets showed that CTSS was highly expressed in macrophages, but elevated expression was exclusive for the β-cells in donors with type 1 diabetes. Whether the induction and secretion of CTSS from β-cells reflects an active role in type 1 diabetes pathogenesis remains to be elucidated. CTSS has been proposed to be involved in the degradation of the extracellular matrix or basement membrane surrounding the islets (3). This could explain why a previous study reported reduced insulitis and partial protection from diabetes in Ctss^−/− NOD mice (4). Future studies should address the possible therapeutic potential of CTSS in type 1 diabetes.

The current study warrants further investigations to clarify whether CTSS could be an early biomarker in type 1 diabetes, potentially reflecting insulitis, and thus be used for diagnostic screening before the onset of symptoms.

Acknowledgments. The authors thank Rebekka Gerwig, Fie Hillesø, and Tine Wille from Steno Diabetes Center Copenhagen for their excellent technical assistance. The authors acknowledge the help of Sara Juul Mansachs, Herlev and Gentofte Hospital, as well as Gitte Lindved Petersen and Simranjeet Kaur, Steno Diabetes Center Copenhagen, with the Danish Childhood Diabetes Register sample preparation and data analysis, respectively. This article used data acquired from the Human Pancreas Analysis Program (HPAP-RRID:SCR_016202) Database (https://hpap.pmacs.upenn.edu), a Human Islet Research Network (RRID:SCR_014393) consortium (UC4-DK-112217, U01-DK-123594, UC4-DK-112232, and U01-DK-123716).

Funding. This work was funded by grants to T.F. from JDRF International, New York, NY (3-PDF-2020-938-A-N), and Læge Sofus Carl Emil Friis og Hustru Olga Doris Friis’ Legat, Denmark, and to F.P. from Danmarks Frie Forskningsfond (Independent Research Fund Denmark), Denmark (DFF-0134-00267 and DFF-4183-00031).

Duality of Interest. K.B., J.S., and T.F. hold shares in Novo Nordisk. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. T.F. developed the original idea of the study. F.P. and T.F. designed the experiments. C.F., M.H.J., and T.F. researched data. C.F., M.H.J., M.H.-J., K.B., J.S., F.P., and T.F. contributed to discussion. F.P. and T.F. acquired funding for the study. F.P. and T.F. performed supervision. C.F., M.H.-J., and T.F. wrote the manuscript. All authors reviewed and edited the manuscript. T.F. is the guarantor of this work and, as such, had full access to all data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented at the 58th annual meeting of the European Association for the Study of Diabetes, Stockholm, Sweden, 19–23 September 2022.