Metaproteomics characterizes host–microbiota responses to the time in range derived from continuous glucose monitoring in type 1 diabetes

To the Editor: Type 1 diabetes (T1D) is an autoimmune disease resulting from the interplay between genomic predisposition and environmental influences, including gut microbiome dysbiosis.[1] Chronic complications of T1D have been linked to abnormal glycemic fluctuations. However, the conventional glycemic control metric, glycosylated hemoglobin, reflects only average glucose levels over several months and does not capture short-term variability. In contrast, time in range (TIR), a metric derived from continuous glucose monitoring (CGM), measures the proportion of time that glucose concentrations remain within the target range (typically 3.9–10.0 mmol/L).[2] TIR thus provides a dynamic assessment of glycemic control, incorporating both the frequency and duration of hypoglycemia and hyperglycemia. Recent studies have suggested that TIR is a predictor of future diabetic complications,[3] yet the underlying mechanisms remain largely unclear. The gut microbiota plays a key role in metabolic diseases,[4] but its response to TIR status and its contribution to the link between TIR and diabetic complications in T1D remain unexplored. Fecal metaproteomics, a technique that enables the comprehensive profiling of microbial and host proteins, offers a powerful approach to investigate these interactions. This cross-sectional study enrolled 37 adults with T1D and 36 healthy controls for fecal metaproteomic analysis. Subjects with T1D were stratified into high TIR (TIR-H >0.70, n = 11) and low TIR (TIR-L ≤0.70, n = 26) groups based on American Diabetes Association recommendations. The clinical features are detailed in Supplementary Table 1, https://links.lww.com/CM9/C854. All participants with T1D underwent CGM for 14 days to calculate glycemic indices, including mean blood glucose, TIR, time above range and time below range, and glycemic variability metrics. Fecal samples were collected and analyzed using metaproteomics to characterize both host- and microbiota-derived proteins. Taxonomic and functional annotation was performed, and network analyses explored associations between microbial taxa, microbial proteins, and host proteins. To validate the metabolite consequences of the observed changes, a second T1D cohort (n = 91) underwent targeted serum short-chain fatty acids (SCFAs) profiling, and their discriminatory power for TIR was tested by receiver-operating characteristic (ROC) analysis. The clinical characteristics of the new cohort are shown in Supplementary Table 2, https://links.lww.com/CM9/C854. To ensure the comparability between the two cohorts, their clinical characteristics were compared and summarized in Supplementary Table 3, https://links.lww.com/CM9/C854. The Human Care and Use Committee of Peking Union Medical College Hospital approved the study protocol (No. 13275). Written informed consent was obtained from all participants. Detailed methods are provided in the Supplementary Methods, https://links.lww.com/CM9/C854. Our analyses revealed distinct microbial and host protein signatures separating individuals with T1D from healthy controls [Supplementary Figure 1A, B, and C, https://links.lww.com/CM9/C854]. At the host level, the differentially abundant proteins were involved in pancreatic secretion, fatty acid degradation, peroxisome proliferator-activated receptor (PPAR) signaling, complement and coagulation cascades, metabolic pathways, and protein digestion and absorption [Supplementary Figure 1D, https://links.lww.com/CM9/C854]. At the microbial level, functional annotation indicated increased glycolysis, amino acid degradation, and sucrose metabolism, while pathways such as galacturonate degradation were reduced [Supplementary Figure 1E, https://links.lww.com/CM9/C854]. Altered taxa included enrichment of Leptolyngbya, which positively correlated with immunomodulatory host proteins; and depletion of Micromonospora and Clostridioides, positively correlated with exocrine pancreas proteins chymotrypsin like elastase family member 3A (CELA3A) and chymotrypsin like elastase family member 3A (CELA3B) [Supplementary Figure 1F, https://links.lww.com/CM9/C854]. These findings suggest that T1D is associated not only with shifts in microbial composition but also with functional alterations in host–microbiota protein interactions that may contribute to its pathogenesis. We further identified specific host protein responses to TIR and found notable differences in their profiles between the TIR-H and TIR-L groups [Supplementary Figure 2A and B, https://links.lww.com/CM9/C854]. Functional analysis of these differential proteins showed that they were involved in cardiomyocyte contraction regulation, chronic neurodegenerative diseases, and lipid metabolism [Supplementary Figure 2C, https://links.lww.com/CM9/C854]. This observation is consistent with epidemiological data. Besides, metaproteomic data clearly distinguished TIR-L from TIR-H subjects at the microbial level [Supplementary Figure 2D, https://links.lww.com/CM9/C854]. Both SCFAs-producing genera (Acidiphilium, Anaerotignum, and Peptoniphilus) and bacterial proteins involved in SCFAs production decreased with TIR decline [Supplementary Figure 2E, https://links.lww.com/CM9/C854]. SCFAs are produced in large quantities by the gut microbiota through anaerobic fermentation of dietary fiber, with acetate, propionate, and butyrate ranking in the top three. It has been widely reported that they exert positive effects on cardiovascular diseases, a leading cause of mortality in people with diabetes, by regulating inflammation, the immune system, and associated G protein-coupled receptors, with the participation of multiple neurohumoral regulatory mechanisms.[5] Correlation networks highlighted a striking functional link between microbial and host proteins. SCFAs-producing bacteria were positively associated with host proteins essential for myocardial contraction, including troponin C1 (TNNC1), tropomyosin 4 (TPM4), and myosin light chain 4 (MYL4) [Supplementary Figure 2F, https://links.lww.com/CM9/C854]. They are essential components of the troponin complexes and myosin, and the cyclical interactions between these structures drive cardiac contraction. This may provide a novel molecular mechanism, suggesting that gut microbiota and their metabolic products, SCFAs, might be involved in the underlying mechanisms by which TIR affects the onset and progression of diabetic myocardial dysfunction. To further explore the clinical relevance of SCFAs in assessing glycemic control, we conducted targeted metabolomics followed by a ROC curve analysis in a second, independent cohort of T1D, to evaluate the performance of various SCFAs as biomarkers for distinguishing between TIR-H and TIR-L. Consistently, the concentrations of serum SCFAs (acetic acid, propionic acid, isobutyric acid, butyric acid, valeric acid, and hexanoic acid) were significantly decreased in individuals with T1D whose TIR was ≤0.7 compared with those who had TIR >0.7 [Supplementary Figure 3A, https://links.lww.com/CM9/C854], with valeric acid exhibiting the highest discriminatory power [Supplementary Figure 3B, https://links.lww.com/CM9/C854]. As a relatively little-known SCFA, the role of valeric acid in glucose metabolism homeostasis and diabetic complications has been limitedly reported. Prior studies have suggested potential roles in cardiovascular protection and glucose tolerance. Therefore, our research extended the understanding of valeric acid in glycemic management in T1D. The use of SCFAs as biomarkers offers a novel, objective method for evaluating TIR in T1D subjects, complementing traditional glucose monitoring techniques. This study had several limitations. First, the participants in this study were from a single center, and the sample size was limited, potentially restricting the generalizability of our results. A large-scale study from multiple centers is needed to validate these results and ensure broader applicability. Second, the functions of some differential bacterial taxa and proteins remain unknown, limiting insights into their roles in the gut microbiota and metabolic regulation. Future molecular biology studies and related animal models are needed to clarify these possible mechanisms. Third, the lack of dietary intake data may have introduced confounding effects, and incorporating detailed dietary assessments in future research would help account for this variable. Last, the potential effects of α-glucosidase inhibitors and metformin on the gut microbiota in individuals with T1D may also introduce confounding effects, although these medications were evenly distributed between the TIR-H and TIR-L subgroups in this study. Future studies with more detailed stratification based on medication use are needed to comprehensively evaluate these effects. This study identified novel associations among the gut microbiota composition, microbiota-derived proteins, host proteins, and the TIR of T1D. Collectively, our findings suggest that the composition of the gut microbiota was modulated in T1D patients with different levels of TIR. These changes result in modified microbe-derived proteins that interact with host proteins, potentially contributing to the development of diabetic complications. However, these observations are speculative, and further investigation is required to validate their therapeutic potential. We detected positive correlations among the abundance of SCFAs-producing genera, microbial proteins involved in SCFAs production, and TIR levels. Furthermore, we established that serum SCFAs serve as promising biomarkers that exhibit significant discriminatory performance for glycemic control, with valeric acid being particularly noteworthy [Figure 1]. This is an intriguing area for future research, and we hypothesize that valeric acid supplementation could serve as a promising therapeutic strategy to modulate TIR levels in T1D patients. These findings may shed new light on the glycemic management of T1D.Figure 1: Metaproteomics revealed gut microbiota remodeling and host-microbial protein networks linking SCFAs-producing genera to diabetic complications in T1D subjects with different TIR. Targeted metabolomics indicated SCFAs as promising biomarkers for TIR control in T1D subjects. ROC: Receiver-operating characteristic; SCFAs: Short-chain fatty acids; T1D: Type 1 diabetes; TIR: Time in range; TIR-H: High TIR; TIR-L: Low TIR.