Beta-Glucan Modulates Monocyte Plasticity and Differentiation Capacity to Mitigate DSS-Induced Colitis

Reviewer #1 (Public review):

Summary:

This study presents an interesting investigation into the role of trained immunity in inflammatory bowel disease, demonstrating that β-glucan-induced reprogramming of innate immune cells can ameliorate experimental colitis. The findings are novel and clinically relevant, with potential implications for therapeutic strategies in IBD. The combination of functional assays, adoptive transfer experiments, and single-cell RNA sequencing provides comprehensive mechanistic insights. However, some aspects of the study could benefit from further clarification to strengthen the conclusions.

Strengths:

(1) This study elegantly connects trained immunity with IBD, demonstrating how β-glucan-induced innate immune reprogramming can mitigate chronic inflammation.

(2) Adoptive transfer experiments robustly confirm the protective role of monocytes/macrophages in colitis resolution.

(3) Single-cell RNA sequencing provides mechanistic depth, revealing the expansion of reparative Cx3cr1⁺ macrophages and their contribution to epithelial repair.

(4) The work highlights the therapeutic potential of trained immunity in restoring gut homeostasis, offering new directions for IBD treatment.

Weaknesses:

While β-glucan may exert its training effect on hematopoietic stem cells, performing ATAC-seq on HSCs or monocytes to profile chromatin accessibility at antibacterial defense and mucosal repair-related genes would further validate the trained immunity mechanism. Alternatively, the authors could acknowledge this as a study limitation and future research direction.

Reviewer #2 (Public review):

Summary:

The study investigates whether β-glucan (BG) can reprogram the innate immune system to protect against intestinal inflammation. The authors show that mice pretreated with BG prior to DSS-induced colitis experience reduced colitis severity, including less weight loss, colon damage, improved gut repair, and lowered inflammation. These effects were independent of adaptive immunity and were linked to changes in monocyte function.

The authors show that the BG-trained monocytes not only help control inflammation but confer non-specific protection against experimental infections (Salmonella), suggesting the involvement of trained immunity (TI) mechanisms. Using single-cell RNA sequencing, they map the transcriptional changes in these cells and show enhanced differentiation of monocytes into reparative CX3CR1⁺ macrophages. Importantly, these protective effects were transferable to other mice via adoptive cell transfer and bone marrow transplantation, suggesting that the innate immune system had been reprogrammed at the level of stem/progenitor cells.

Overall, this study provides evidence that TI, often associated with heightened inflammatory programs, can also promote tissue repair and resolution of inflammation. Moreover, this BG-induced functional reprogramming can be further harnessed to treat chronic inflammatory disorders like IBD.

Strengths:

(1) The authors use advanced experimental approaches to explore the potential therapeutic use of myeloid reprogramming by β-glucan in IBD.

(2) The authors follow a data-to-function approach, integrating bulk and single-cell RNA sequencing with in vivo functional validation to support their conclusions.

(3) The study adds to the growing evidence that TI is not a singular pro-inflammatory program, but can adopt distinct functional states, including anti-inflammatory and reparative phenotypes, depending on the context.

Weaknesses:

(1) The epigenetic and metabolic basis of TI is not explored, which weakens the mechanistic claim of TI. This is especially relevant given that a novel reparative, anti-inflammatory TI program is proposed.

(2) The absence of a BG-only group limits interpretation of the results. Since the authors report tissue-level effects such as enhanced mucosal repair and transcriptional shifts in intestinal macrophages (colonic RNA-Seq), it is important to rule out whether BG alone could influence the gut independently of DSS-induced inflammation.
Without a BG-only control, it is hard to distinguish a true trained response from a potential modulation caused directly by BG.

(3) Although monocyte transfer experiments show protection in colitis, the fate of the transferred cells is not described (e.g., homing or differentiation into Cx3cr1⁺ macrophage subsets). This weakens the link between specific monocyte subsets and the observed phenotype.

(3) While scRNA-seq reveals distinct monocyte/macrophage subclusters (Mono1-3..), their specific functional roles remain speculative. The authors assign reparative or antimicrobial functions based on transcriptional signatures, but do not perform causal experiments (depletion or in vitro assays). The biological roles of these cells remain correlative.

(4) While Rag1⁻/⁻ mice were used to rule out adaptive immunity, the potential role of innate lymphoid cells (ILCs), particularly ILC2s and ILC3s, which are known to promote mucosal repair (PMID: 27484190), was not explored. Given the reparative phenotype observed, the contribution of ILCs remains a confounding factor.

Reviewer #3 (Public review):

Summary:

In the present work, Yinyin Lv et al offer evidence for the therapeutic potential of trained immunity in the context of inflammatory bowel disease (IBD). Prior research has demonstrated that innate cells pre-treated (trained) with β-glucan show an enhanced pro-inflammatory response upon a second challenge.

While an increased immune response can be beneficial and protect against bacterial infections, there is also the risk that it will worsen symptoms in various inflammatory disorders. In the present study, the authors show that mice preconditioned with β-glucan have enhanced resistance to Staphylococcus aureus infection, indicating heightened immune responses.

The authors demonstrate that β-glucan training of bone marrow hematopoietic progenitors and peripheral monocytes mitigates the pro-inflammatory effects of colitis, with protection extending to naïve recipients of the trained cells.

Using a dextran sulfate sodium (DSS)-induced model of colitis, β-glucan pre-treatment significantly dampens disease severity. Importantly, the use of Rag1^-/- mice, which lack adaptive immune cells, confirms that the protective effects of β-glucan are mediated by innate immune mechanisms. Further, experiments using Ccr2^-/- mice underline the necessity of monocyte recruitment in mediating this protection, highlighting CCR2 as a key factor in the mobilization of β-glucan-trained monocytes to inflamed tissues. Transcriptomic profiling reveals that β-glucan training upregulates genes associated with pattern recognition, antimicrobial defense, immunomodulation, and interferon signaling pathways, suggesting broad functional reprogramming of the innate immune compartment. In addition, β-glucan training induces a distinct monocyte subpopulation with enhanced activation and phagocytic capacity. These monocytes exhibit an increased ability to infiltrate inflamed colonic tissue and differentiate into macrophages, marked by increased expression of Cx3cr1. Moreover, among these trained monocyte and macrophage subsets, other gene expression signatures are associated with tissue and mucosal repair, suggesting a role in promoting resolution and regeneration following inflammatory insult.

Strengths:

(1) Overall, the authors present a mechanistically insightful investigation that advances our understanding of trained immunity in IBD.

(2) By employing a range of well-characterized murine models, the authors investigate specific mechanisms involved in the effects of β-glucan training.

(3) Furthermore, the study provides functional evidence that the protection conferred by the trained cells persists within the hematopoietic progenitors and can be transferred to naïve recipients. The integration of transcriptomic profiling allows the identification of changes in key genes and molecular pathways underlying the trained immune phenotype.

(4) This is an important study that demonstrates that β-glucan-trained innate cells confer protection against colitis and promote mucosal repair, and these findings underscore the potential of harnessing innate immune memory as a therapeutic approach for chronic inflammatory diseases.

Weaknesses:

However, FPKM is not ideal for between-sample comparisons due to its within-sample normalization approach. Best practices recommend using raw counts (with DESeq2) for more robust statistical inference.

Author response:

Public Reviews:

Reviewer #1 (Public review):

Summary:

This study presents an interesting investigation into the role of trained immunity in inflammatory bowel disease, demonstrating that β-glucan-induced reprogramming of innate immune cells can ameliorate experimental colitis. The findings are novel and clinically relevant, with potential implications for therapeutic strategies in IBD. The combination of functional assays, adoptive transfer experiments, and single-cell RNA sequencing provides comprehensive mechanistic insights. However, some aspects of the study could benefit from further clarification to strengthen the conclusions.

We are grateful for the reviewer’s positive assessment of our study and constructive suggestions to improve the manuscript.

Strengths:

(1) This study elegantly connects trained immunity with IBD, demonstrating how β-glucan-induced innate immune reprogramming can mitigate chronic inflammation.

(2) Adoptive transfer experiments robustly confirm the protective role of monocytes/macrophages in colitis resolution.

(3) Single-cell RNA sequencing provides mechanistic depth, revealing the expansion of reparative Cx3cr1⁺ macrophages and their contribution to epithelial repair.

(4) The work highlights the therapeutic potential of trained immunity in restoring gut homeostasis, offering new directions for IBD treatment.

Weaknesses:

While β-glucan may exert its training effect on hematopoietic stem cells, performing ATAC-seq on HSCs or monocytes to profile chromatin accessibility at antibacterial defense and mucosal repair-related genes would further validate the trained immunity mechanism. Alternatively, the authors could acknowledge this as a study limitation and future research direction.

We agree that further epigenetic profiling—such as ATAC-seq analysis on HSCs or monocytes—would provide additional mechanistic depth to our current findings. We will acknowledge this as a limitation of the present study and highlight it as an important direction for future research.

Comment (1): It’s better to include a schematic summarizing the proposed mechanism for reader clarity.

We agree that a visual summary will enhance the clarity and accessibility of our findings. We will add a new schematic diagram (Figure 6) illustrating the proposed mechanism of β-glucan–induced myeloid reprogramming and its protective effects in the experimental colitis model.

Comment (2): Discuss potential off-target effects of β-glucan-induced trained immunity (e.g., risk of exacerbated inflammation in other contexts).

We appreciate this important comment regarding the potential off-target effects of β-glucan pretreatment. As trained immunity is known to amplify inflammatory responses upon heterologous stimulation and has been implicated in chronic inflammation–prone conditions such as atherosclerosis, this is an important consideration. Previous in vivo studies have shown that β-glucan pretreatment can enhance antibacterial or antitumor responses without inducing basal inflammation after one week of administration (PMID: 22901542, PMID: 30380404, PMID: 36604547, PMID: 33125892). Nevertheless, it remains possible that β-glucan–induced trained immunity could have unintended effects in certain contexts, which warrants further investigation and caution. We will expand the Discussion section to include a dedicated paragraph addressing these potential off-target effects.

Reviewer #2 (Public review):

Summary:

The study investigates whether β-glucan (BG) can reprogram the innate immune system to protect against intestinal inflammation. The authors show that mice pretreated with BG prior to DSS-induced colitis experience reduced colitis severity, including less weight loss, colon damage, improved gut repair, and lowered inflammation. These effects were independent of adaptive immunity and were linked to changes in monocyte function.

The authors show that the BG-trained monocytes not only help control inflammation but confer non-specific protection against experimental infections (Salmonella), suggesting the involvement of trained immunity (TI) mechanisms. Using single-cell RNA sequencing, they map the transcriptional changes in these cells and show enhanced differentiation of monocytes into reparative CX3CR1⁺ macrophages. Importantly, these protective effects were transferable to other mice via adoptive cell transfer and bone marrow transplantation, suggesting that the innate immune system had been reprogrammed at the level of stem/progenitor cells.

Overall, this study provides evidence that TI, often associated with heightened inflammatory programs, can also promote tissue repair and resolution of inflammation. Moreover, this BG-induced functional reprogramming can be further harnessed to treat chronic inflammatory disorders like IBD.

Strengths:

(1) The authors use advanced experimental approaches to explore the potential therapeutic use of myeloid reprogramming by β-glucan in IBD.

(2) The authors follow a data-to-function approach, integrating bulk and single-cell RNA sequencing with in vivo functional validation to support their conclusions.

(3) The study adds to the growing evidence that TI is not a singular pro-inflammatory program, but can adopt distinct functional states, including anti-inflammatory and reparative phenotypes, depending on the context.

We are grateful for the reviewer’s positive assessment of our study and recognition of its translational implications. We particularly appreciate the acknowledgment that our work expands the therapeutic potential of β-glucan–mediated trained immunity in ameliorating colitis.

Weaknesses:

(1) The epigenetic and metabolic basis of TI is not explored, which weakens the mechanistic claim of TI. This is especially relevant given that a novel reparative, anti-inflammatory TI program is proposed.

We appreciate the reviewer’s valuable comment highlighting the importance of the epigenetic and metabolic basis of TI in providing mechanistic insight. While previous studies, including work from our group (S.-C. Cheng), have extensively characterized the epigenetic and metabolic signatures of monocytes from BG-trained mice—primarily in the context of inflammatory genes—we acknowledge that these aspects are not directly addressed in our current manuscript.

To strengthen the mechanistic component, we plan to: 1. Reanalyze relevant public datasets, focusing on pathways related to reparative and antibacterial function. 2. Perform monocyte ATAC-seq in our current model to validate the epigenetic changes in these pathways.

(2) The absence of a BG-only group limits interpretation of the results. Since the authors report tissue-level effects such as enhanced mucosal repair and transcriptional shifts in intestinal macrophages (colonic RNA-Seq), it is important to rule out whether BG alone could influence the gut independently of DSS-induced inflammation.

Without a BG-only control, it is hard to distinguish a true trained response from a potential modulation caused directly by BG.

We thank the reviewer for this important suggestion. Although we did not perform qPCR for mucosal repair genes in Figure S1C and Figure S1D, our colon RNA-seq analysis in Figure 5G included a BG-only control group (Colitis_d0). The results from this group indicate that BG preconditioning alone does not alter baseline expression of colon mucosal repair genes, supporting the conclusion that the observed effects occur in the context of DSS-induced inflammation.

(3) Although monocyte transfer experiments show protection in colitis, the fate of the transferred cells is not described (e.g., homing or differentiation into Cx3cr1⁺ macrophage subsets). This weakens the link between specific monocyte subsets and the observed phenotype.

(4) While scRNA-seq reveals distinct monocyte/macrophage subclusters (Mono1-3.), their specific functional roles remain speculative. The authors assign reparative or antimicrobial functions based on transcriptional signatures, but do not perform causal experiments (depletion or in vitro assays). The biological roles of these cells remain correlative.

We agree that the functional role of CX3CR1⁺ macrophages is not comprehensively validated and is currently inferred from scRNA-seq clustering. While our flow cytometry data show increased CX3CR1⁺ macrophages in the BG-TI group, and our CCR2 KO and monocyte adoptive transfer experiments indicate these macrophages are monocyte-derived, we lack direct depletion experiments due to the unavailability of effective depletion antibodies for this subset.

We acknowledge this as a limitation and will clarify in the Discussion that our conclusions regarding CX3CR1⁺ macrophage function are based on transcriptional profiling and association with protective phenotypes, rather than direct causal evidence.

(5) While Rag1^-/- mice were used to rule out adaptive immunity, the potential role of innate lymphoid cells (ILCs), particularly ILC2s and ILC3s, which are known to promote mucosal repair (PMID: 27484190IF: 7.6 Q1 IF: 7.6 Q1 IF: 7.6 Q1 IF: 7.6 Q1 IF: 7.6 Q1 IF: 7.6 Q1 ), was not explored. Given the reparative phenotype observed, the contribution of ILCs remains a confounding factor.

We appreciate the reviewer’s valuable comment regarding the potential role of ILCs in the observed mucosal repair. Indeed, in examining the BG-trained immunity effect, the contribution of ILCs was not evaluated. We will explicitly acknowledge in the Discussion that Rag1⁻/⁻ mice retain ILCs (including ILC3s) and that BG-induced activation of these cells remains possible.

The literature (PMID: 21502992; PMID: 32187516) supports a role for ILC3-mediated IL-22 production in tissue repair, which could overlap with our observed effects. However, our monocyte adoptive transfer experiments show that monocytes alone can alleviate DSS-induced colitis, suggesting a dominant role for monocytes in this context. Nonetheless, we will make it clear that ILC contributions cannot be excluded.

Reviewer #3 (Public review):

Summary:

In the present work, Yinyin Lv et al offer evidence for the therapeutic potential of trained immunity in the context of inflammatory bowel disease (IBD). Prior research has demonstrated that innate cells pre-treated (trained) with β-glucan show an enhanced pro-inflammatory response upon a second challenge.

While an increased immune response can be beneficial and protect against bacterial infections, there is also the risk that it will worsen symptoms in various inflammatory disorders. In the present study, the authors show that mice preconditioned with β-glucan have enhanced resistance to Staphylococcus aureus infection, indicating heightened immune responses.

The authors demonstrate that β-glucan training of bone marrow hematopoietic progenitors and peripheral monocytes mitigates the pro-inflammatory effects of colitis, with protection extending to naïve recipients of the trained cells.

Using a dextran sulfate sodium (DSS)-induced model of colitis, β-glucan pre-treatment significantly dampens disease severity. Importantly, the use of Rag1^-/- mice, which lack adaptive immune cells, confirms that the protective effects of β-glucan are mediated by innate immune mechanisms. Further, experiments using Ccr2^-/- mice underline the necessity of monocyte recruitment in mediating this protection, highlighting CCR2 as a key factor in the mobilization of β-glucan-trained monocytes to inflamed tissues. Transcriptomic profiling reveals that β-glucan training upregulates genes associated with pattern recognition, antimicrobial defense, immunomodulation, and interferon signaling pathways, suggesting broad functional reprogramming of the innate immune compartment. In addition, β-glucan training induces a distinct monocyte subpopulation with enhanced activation and phagocytic capacity. These monocytes exhibit an increased ability to infiltrate inflamed colonic tissue and differentiate into macrophages, marked by increased expression of Cx3cr1. Moreover, among these trained monocyte and macrophage subsets, other gene expression signatures are associated with tissue and mucosal repair, suggesting a role in promoting resolution and regeneration following inflammatory insult.

Strengths:

(1) Overall, the authors present a mechanistically insightful investigation that advances our understanding of trained immunity in IBD.

(2) By employing a range of well-characterized murine models, the authors investigate specific mechanisms involved in the effects of β-glucan training.

(3) Furthermore, the study provides functional evidence that the protection conferred by the trained cells persists within the hematopoietic progenitors and can be transferred to naïve recipients. The integration of transcriptomic profiling allows the identification of changes in key genes and molecular pathways underlying the trained immune phenotype.

(4) This is an important study that demonstrates that β-glucan-trained innate cells confer protection against colitis and promote mucosal repair, and these findings underscore the potential of harnessing innate immune memory as a therapeutic approach for chronic inflammatory diseases.

We thank the reviewer for their positive evaluation and constructive feedback on our manuscript.

Weaknesses:

However, FPKM is not ideal for between-sample comparisons due to its within-sample normalization approach. Best practices recommend using raw counts (with DESeq2) for more robust statistical inference.

We appreciate the reminder about best practices for RNA-seq analysis. We apologize for the inaccurate description in the Materials and Methods section. For all differential expression analyses, we have in fact used raw count data as input for DESeq2. FPKM values were only used for visualization purposes, such as in heatmaps and clustering analyses. We will correct this description in the revised manuscript to accurately reflect our analysis workflow.