Submit to this Journal Review for this Journal Propose a Special Issue

Article Menu

Share Help Cite Discuss in SciProfiles

Open AccessReview

Peer-Review Record

Type 2 Diabetes and the Multifaceted Gut-X Axes

Nutrients 2025, 17(16), 2708; https://doi.org/10.3390/nu17162708

by Hezixian Guo^1,†, Liyi Pan^1,†, Qiuyi Wu¹, Linhao Wang¹

, Zongjian Huang¹, Jie Wang¹, Li Wang¹, Xiang Fang¹

, Sashuang Dong¹

, Yanhua Zhu^2,3,* and Zhenlin Liao^1,*

Reviewer 1: Anonymous

Reviewer 2: Anonymous

Reviewer 3:

Alexander E. Berezin

Nutrients 2025, 17(16), 2708; https://doi.org/10.3390/nu17162708

Submission received: 22 July 2025 / Revised: 14 August 2025 / Accepted: 17 August 2025 / Published: 21 August 2025

(This article belongs to the Special Issue Dietary Regulation of Glucose and Lipid Metabolism in Diabetes)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors review gut homeostasis and T2D pathogenesis and then discuss several “axes” : 1) Gut-Pancreas (2) Gut-Endo- Axis (a little bit unclear but defined as: enteroendocrine signals (e.g. PYY, ghrelin) for regulation of appetite, adipose tissue, and systemic metabolism; (3) Gut-Liver (NAFLD) and hepatic insulin resistance); (4) Gut-Kidney Axis: effect of gut-derived toxins and nutrients on diabetic kidney disease and effects of incretins and SGLT2 inhibitor therapies. In addition is discussed microbial SCFAs improving insulin sensitivity, LPS 42 driving inflammation via TLR4, and aryl hydrocarbon receptor ligands modulating immunity. Modulating the gut microbiome and its metabolites through diet, pharmaceuticals, or microbiota therapies are discussed as therapies. Gaps for translating these insights into clinical practice are discussed.

Comments: This is a very comprehensive review covering a huge area of research. It is very informative and well written. I actually enjoyed reading it. I think it will be useful for many people in many areas. Below I have goine into some details, where I could find inaccuracies , errors, misunderstandings etc. I have been careful to do this because I think the review might be useful for many and I would hate to see too many misunderstandings promoted.

100, while it is true that bariatric surgery can lead to diabetes remission this is not by “reshaping gut-hormone profiles and microbiota composition”, but by accelerated and abnormal exposure of distal gut segments to nutrients . Also, the success of the SGLT2inhibitors is partly due to a “plummer-like” and not normal effect on renal glucose excretion .
116. I think a new section was intended here?
130: Although this is an unusual review, I question the wisdom of also reviewing reviews. All reviews are biased. By reviewing reviews there is a serious risk of “doble bias”. I am also unhappy about exclusion of literature from before 2015. A lot of important literature in this area appeared before that time. Gastrointestinal physiology was certainly strongly developed throughout the 20^th century, particularly from1950 and onwards. I also question the use of the systematic review approach in the literature search – this is an overviewing narrative review, and the most important is to be sure to include the most important communications. I don’t know what to do about the deselection of earlier literature in the review, but the editors should serious consider whether this should be encouraged.

L.179: “gut microbial dysbiosis and increased intestinal permeability in obesity/T2D lead to 1metabolic endotoxemia, which can drive insulin resistance [28]”.- The importance of this is controversial – it should not be presented as an established fact

189: typo
195: an impaired incretin release is not established in T2DM (but impaired incretin action is)

Table 1: GLP-1 Receptor agonists are not the same as GLP-1 analogs – The term was chosen to include also exendin-derived peptides which are not GLP-1 analogs (the Gila monster has its own GLP-1 molecule which is more GLP-1 like). Regarding the A1c improvements they are said to be 0.9- 1.0 %. The relevance of this obviously depends on the baseline A1c. Decreases with tirzepatide and semaglutide are usually > 2.0 %, but even more important is the number of individuals reaching levels below 6.5 % or even lower (more than 50 % reaching <5.7 % in the case of tirzepatide 15 mg) .

Also the question is whether GIP inclusion really increases the antidiabetic activity of the co-agonists – the biggest difference is the dose! For example 7.2 mg semagluide and tirzepatide 10-15 mg have about the same activity, at least in people with diabetes.

Also the authors proport the notion that the SCFAs promote GLP-1 release via FFAR2/3 receptors – that is wrong and the cited consequences are also erroneous

The IPA story is still not confirmed

It is true that the mechanism of bariatric surgery is often referred to as the “hindgut effect”, but actually this is a misconception – its is not related to the hindgut but to the mid-gut (the hind gut starts at mid colon)

Also the fecal microbiota transplant results are very variable – in one of the best studies there was effect in half of the transplanted individuals – what does that show?

As noted the effects of probiotics are weak and variable - and it is highly doubtful whether this is a viable approach .

In conclusion, regarding table 1 – it is probably useful, but it should be more cautiously phrased - most of the approaches are overrated.

216 It is said that there isa release of GLP-1 from the coon – in fact it has never been demonstrated that GLP-1 from the colon contributes to the circulating levels – on the contrary people with total colectomy have completely normal both fasting and postprandial GLP-1 levels
225: it is a prerequisite for the effects of the GLP-1RAs on insulin secretion that there is sufficient residual beta cell function.

l.231. it is highly controversial whether artificial sweeteners stimulate GLP-1 secretion – in the most robust studies there was no such effect.

242. As already mentioned, the actual contribution by GIP to the actions of tirzepatide remains unclear – other potent GIP/GP-1 co-agonists do not have antidiabetic properties in excess of the GLP-1 part . The case for glucagon is better (probably because it also acts on the GL1 receptor )
252: the role of SCFA to stimulate GLp-1 is highly controversial (and wrong)
263-277: If the authors had written: “it has been proposed”, all of this would be acceptable. The actions of these metabolites are almost entirely hypothetical
305: the misconception regarding the hindgut hypothesis has been alluded to above.
306: The increased GLP-1 response is invariable – not just “often”
311: this ref is not the best regarding stool transplantation - the leading group is that of Max Niewdorf
349: the physiological role of oxyntomodulin is unclear - only after bariatric surgery are the levels sufficient for this peptide to have significant effects.
351: GLP-1 does not cross the Blood brain barrier – it interacts with neurons in the circumventricular organs. The most significant expression of the receptors is also found there .
357: it should be emphasized that ghrelin levels are deceased in obesity
361: RYGB does not restrict food intake! On the contrary it allows rapid unhindered passage of nutrients from the esophagus to the more distal small intestine. The decreased food intake is exclusively due to decreased appetite.
380: note that the GIP- GLP-1 studies referred to are produced in rodents – nothing is know regarding humans.
403: there is consensus that GLP-1 does not increase EE in people

Figure 2: the upper right circle indicates that PYY interacts with the anterior pituitary – where does that come from? The problem with this kind of graphic representation that it is often inaccurate.

Also, the interaction of GLP-1 in the arcuate nucleus – this is complicated and not well worked out; it seems there may some communication between the median eminence (to which GLP-1 has immediate access) and the neighboring arcuate - another (but less likely) hypothesis is that there is a communication from the ventricles via tanycytes.

449-451 “Bile acids, SCFAs, and intestinal leakage-derived LPS modulate adipogenesis, energy expenditure, and inflammation by promoting adipose browning and activating the TLR4-NF- 450 κB pathway in adipose macrophages, respectively” . The evidence that this occurs in humans is really weak!
469: NAFLD is now MAFLD (recognizing that truly non-alcoholic cases are very rare) – OK, mentioned later, but why not introduce it now? l. 505
532: it should be emphasized that the main reason for hepatic insulin resistance is steatosis which is very common in obesity and T2DM-. It also should be mentioned that one of the major factors regulating hepatic overproduction of glucose in T2DM is glucagon excess. On the whole, from a therapeutic standpoint glucagon is an extremely important factor in the regulation of hepatic metabolism .
579; refs 156 – 157 Any updates regarding this compound (JKB-121) ?

588: there is agreement that there are no GLP-1 receptors on hepatocytes (but , as stated possibly on stellate cells although this is also controversial)

In table 2, section about TGR5 agonists, the authors again mention activation of BAT - but I miss a statement on the importance of this in adult humans with very limited BAT, both here and elsewhere. Otherwise, I agree with the descriptions of the other treatments in the table.

663: there is only one of the common SGLT2inhibitors (namely canagliflozin) that also affects SGLT1 and therefore glucose absorption in the gut – the others do not have effect.

I miss a brief summary of the impressive effects of SGLT2 inhibitors on DKD in general! This is very important and has led to approval also for non-diabetic kidney disease.

680: Again, these idea that microbiota metabolites have important effects on intestinal GLP-1 production are not supported by studies in humans at least not yet.
683: the expression of GLP-1 receptors in the kidneys is again a controversial issue – the only consensus site is the afferent arterioles! The most important study on the clinical effects is the FLOW trial!! Which very convincing and specifically deals with DKD
739: the authors mention that secreted GLP-1 may maintain the gut’s barrier function - That appears to be true, but even more important is the simultaneous secretion of GLP-2 which has marked barrier protecting effects.
755: There hasn’t been much mentioning of the AhR? I see that it comes later , but should be referred to or mentioned here
759-768 Th authors return to the effects of SCFAs on GLP-1 secretion - and as previously mentioned the evidence for this in people is very weak (in fact there is a lot of evidence to the contrary). In addition, while it is true that butyrate is a fuel for colonocytes, much less can be said about proprionate, the effect of which is humans is unclear – please outline the real metabolism of proprionate in humans ( in ruminants it represents an important stimulus to gluconeogenesis but not in humans . Acetate can be metabolized in certain tissues but what is the predominant fate of it? Rather than hailing the SCFAs as “SCFAs are a unifying beneficial thread” it would be nice to get some facts regarding their metabolism in humans.
791: Define and explain the TLR4

Regarding Fig 4: This figures contains all the more or less hypothetical , erroneous or uncertain features discussed above – it must be possible the emphasize the hypothetical and/or controversial nature of many of these factors.- Regarding the main theme: the pathogeneses of Type2 diabetes , I accept that many of these factors may influence the development, but the main mechanisms behind T2DM have not been sufficiently emphasized, namnely: a) a genetic disposition (high heritability) impairing beta cell function and b) obesity with development of insulin resistance (ectopic fat, liver, heart, muscle)). When glucose tolerance is impaired because of b), the poor beta cell function becomes apparent and glucose intolerance develops. The importance of this is apparent in cases of weight loss, where as shown in the Direct trial, a 15 % weight loss led to diabetes remission in 85 % of cases.

1011: the efficacy of tirzepatide is unique and is not due to a simple combination of GIP and GLP-1 action – such co-agonists are not more effective than GLP-1 alone. The efficacy of tirzepatide was a serendipitous finding.
1026 sevelamer should have been mentioned together with colesevelam.

Author Response

Response:
Thank you for the positive feedback and kind comments. We are very pleased that you found our review informative and useful. We have carefully addressed all the detailed points raised below in order to eliminate any inaccuracies or misunderstandings, as we likewise want the review to be as useful and accurate as possible.

100, while it is true that bariatric surgery can lead to diabetes remission this is not by “reshaping gut-hormone profiles and microbiota composition”, but by accelerated and abnormal exposure of distal gut segments to nutrients . Also, the success of the SGLT2inhibitors is partly due to a “plummer-like” and not normal effect on renal glucose excretion .

Response:
Thank you for this valuable comment. We agree and have revised the corresponding sentence in the Introduction to clarify these mechanisms. Specifically, we now state that bariatric surgery induces T2D remission primarily by rerouting nutrients rapidly to the distal small intestine (leading to exaggerated incretin release), rather than simply by reshaping gut hormone profiles or microbiota composition. We have also clarified the mechanism of SGLT2 inhibitors, noting that their glucose-lowering effect is partly due to a non-physiological increase in urinary glucose excretion (the “plumber-like” effect mentioned), rather than a normal renal process. The revised text reads: “For example, bariatric surgery can rapidly induce T2D remission, primarily by accelerating nutrient delivery to the distal gut which abnormally amplifies incretin hormone release, rather than simply by reshaping gut-hormone profiles or microbiota composition [14]. Likewise, SGLT2 inhibitors lower blood glucose partly via an artificial enhancement of urinary glucose excretion – a ‘plumber-like’ effect – highlighting how harnessing an abnormal excretory route can improve glycemic control.”

116. I think a new section was intended here?

Response:
Thank you for catching this formatting issue. We agree that a new section should start at this point. We have inserted a proper section break and heading at the intended location. The “Methods: Literature Retrieval & Appraisal” section now begins on a new line as Section 2, rather than running on from the previous paragraph. This correction is reflected in the revised manuscript.

130: Although this is an unusual review, I question the wisdom of also reviewing reviews. All reviews are biased. By reviewing reviews there is a serious risk of “doble bias”. I am also unhappy about exclusion of literature from before 2015. A lot of important literature in this area appeared before that time. Gastrointestinal physiology was certainly strongly developed throughout the 20th century, particularly from1950 and onwards. I also question the use of the systematic review approach in the literature search – this is an overviewing narrative review, and the most important is to be sure to include the most important communications. I don’t know what to do about the deselection of earlier literature in the review, but the editors should serious consider whether this should be encouraged.

Response:
Thank you for this thoughtful comment. We understand your concerns regarding potential bias from “reviewing reviews” and the exclusion of pre-2015 literature. In response, we have adjusted our approach and the manuscript text as follows: - We have added several key references from before 2015 (particularly classic studies in gastrointestinal physiology and metabolism) to ensure that important historical findings are acknowledged. For example, we now cite a few seminal mid-20th-century studies and other pre-2015 work where relevant. (Revised Methods section, and throughout the text where appropriate.) - We have reduced reliance on secondary review articles and instead cite original research for critical points, to avoid compounding bias. In the Methods, we clarify that while we followed a structured search strategy, our review is narrative in nature with an emphasis on including the most relevant and seminal studies regardless of publication date. We explicitly note that older foundational work was considered (and has now been included) despite our initial 2015 cutoff. (Revised Methods, line 189–196) - We agree that the goal is to cover the most important communications. Accordingly, we broadened our literature scope and toned down the systematic-review phrasing. We now describe our literature search in a way that stresses comprehensiveness and quality of evidence over strict date limits or purely systematic methods. These changes should mitigate the “double bias” issue and ensure we have not overlooked key prior research. We appreciate this suggestion, as it helped us improve the scholarly balance of the review.

Response:
You are absolutely right that this point is still controversial. We have modified the wording to avoid presenting it as an established fact. In the revised text, we now say: “gut microbial dysbiosis and increased intestinal permeability in obesity/T2D have been associated with metabolic endotoxemia, which is proposed to contribute to insulin resistance”. By adding “associated” and “proposed to,” we indicate that this mechanism is suggested by evidence but not definitively proven, especially in humans. This change should make it clear that the importance of metabolic endotoxemia in insulin resistance is still under investigation and not a settled consensus.

189: typo

Response:
Thank you for pointing out the typographical error at line 189. We have reviewed the text and corrected the typo in the revised manuscript. The error has been fixed (approximately line 232 in the current version). We appreciate your attention to detail.

195: an impaired incretin release is not established in T2DM (but impaired incretin action is)

Response:
Thank you for this clarification. We agree and have corrected the description of the incretin defect in T2D to distinguish between secretion and action. In the revised text, we now emphasize that T2D is characterized by impaired incretin action (especially a reduced effectiveness of GIP), rather than a proven reduction in incretin secretion. Specifically, we changed the wording from “incretin secretion or responsiveness is impaired” to “incretin responsiveness is impaired, despite generally preserved hormone secretion”. This highlights that the loss of the incretin effect in T2D is primarily due to decreased hormone effectiveness (receptor/action level), not necessarily decreased release. We believe this addresses your point that impaired incretin release is not established, whereas impaired incretin action (in particular GIP resistance) is well documented.

Response:
Thank you for these important clarifications. We have updated Table 1 and the related text accordingly:
• Terminology: We now use the term “GLP-1 receptor agonists (GLP-1 RAs)” instead of “GLP-1 analogues” in Table 1. This change reflects the fact that some agents (e.g. exendin-4–based peptides like exenatide) are not direct analogs of human GLP-1 but are GLP-1 receptor agonists. The table heading now reads “GLP-1 receptor agonists” to be inclusive and accurate.
• HbA1c reduction: We have expanded our discussion of the glycemic efficacy of these therapies. In Table 1’s “Key Findings” column and the accompanying narrative, we note that the typical HbA1c reduction of ~0.9–1.0% is an average that depends on baseline A1c. We further mention that more potent agents (e.g. high-dose semaglutide or tirzepatide) often achieve >2.0% HbA1c reductions, especially in patients with higher starting A1c. More importantly, we added that a substantial proportion of patients can reach non-diabetic HbA1c levels with these agents. For instance, we cite that with tirzepatide 15 mg, over 50% of patients achieved HbA1c <5.7% (normal range) in trials. We have included this insight to illustrate the clinical significance beyond the average reductions.
These revisions are reflected in Table 1 (Gut–Pancreas Axis interventions) and the associated text (around lines 292–300 and in the Table 1 footnote). We believe this addresses the concerns by using correct terminology and providing a fuller context for HbA1c improvements, including baseline dependence and the profound efficacy of newer agents.

Response:
This is an excellent point. We have adjusted our discussion of dual GIP/GLP-1 co-agonists to reflect the uncertainty about GIP’s added benefit. In the revised text, we acknowledge that tirzepatide’s superior efficacy may not stem directly from GIP receptor agonism per se, but could largely reflect its higher effective dose or other pharmacological factors. For example, we now note that a high dose of a GLP-1 RA alone (e.g. 7.2 mg semaglutide weekly) can achieve glycemic control comparable to tirzepatide 10–15 mg, suggesting that the contribution of GIP is not clearly above and beyond what an optimized GLP-1 effect provides. We have cited clinical efficacy data to support this comparison. Consequently, we phrase the role of GIP in co-agonists as potential rather than assured – i.e., the added value of including GIP remains debated. This clarification appears in our discussion of incretin-based dual agonists (Revised manuscript, around lines 305–314 and 1228–1236). We appreciate the opportunity to temper that discussion and make it more scientifically nuanced.

Also the authors proport the notion that the SCFAs promote GLP-1 release via FFAR2/3 receptors – that is wrong and the cited consequences are also erroneous

Response:
We appreciate this correction. In light of your comment, we have removed or revised the claim that short-chain fatty acids (SCFAs) directly stimulate GLP-1 release via FFAR2/FFAR3. In Table 1 (SCFA row) and the corresponding text, we now present this idea with much more caution. We state that while SCFAs can engage FFAR2/3 receptors on L-cells in experimental settings, the evidence for a significant increase in GLP-1 secretion in vivo, especially in humans, is weak and inconclusive. We have adjusted the “Key Findings” for SCFAs accordingly and corrected any downstream implications that were based on the earlier, erroneous interpretation. In short, we no longer assert a direct SCFA → GLP-1 causal link as a fact. These changes are reflected in Table 1 and the Gut–Pancreas Axis discussion (lines 320–324 and 1303–1310). Thank you for alerting us to this – we believe the revised text is now scientifically accurate on this point.

The IPA story is still not confirmed

Response:
You are correct – the purported benefits of indole-3-propionic acid (IPA) remain unconfirmed at this time. We did note in the original manuscript that IPA’s mechanism is “not fully proven,” but we have now further emphasized the tentative nature of findings related to IPA. In Table 1 and the associated discussion (lines 333–340), we explicitly state that the link between higher circulating IPA and lower T2D risk is based on associative evidence (e.g., from the Finnish cohort study) and that the mechanistic explanation (antioxidant/anti-inflammatory protection of β-cells) is a hypothesis rather than established fact. By highlighting that IPA’s role is still speculative, we address the concern that the “IPA story” should not be over-interpreted. We will continue to treat IPA as an interesting but unconfirmed piece of the puzzle.

Response:
Thank you for pointing this out. We have corrected our terminology regarding the so-called “hindgut effect.” In the manuscript, wherever we described the mechanism of bariatric surgery, we now clarify that this commonly termed “hindgut effect” is actually a mid-gut phenomenon (involving the distal small intestine, not the anatomical hindgut/colon). For example, in Table 1 we originally mentioned an enhanced distal gut nutrient delivery “(‘hindgut’ effect).” We have put “hindgut effect” in quotes and added a note that this term is a misnomer, as it really refers to nutrient exposure in the midgut (ileum) rather than the hindgut (colon). This correction appears in the bariatric surgery entry of Table 1 (lines 415–419) and the related discussion. We believe this addresses the misconception and will prevent readers from being misled by the terminology.

Also the fecal microbiota transplant results are very variable – in one of the best studies there was effect in half of the transplanted individuals – what does that show?

Response:
We agree that the results of fecal microbiota transplantation (FMT) in metabolic disease have been quite variable. In the manuscript, we have tempered our discussion of FMT to acknowledge its inconsistent efficacy. Specifically, in Table 1 and the accompanying text (Gut–Pancreas Axis, FMT entry, lines 431–439), we now mention that FMT outcomes are mixed – for example, even in one of the most rigorous studies, only roughly half of the recipients showed a significant metabolic improvement. We added that this suggests a “responder vs. non-responder” phenomenon and indicates that FMT’s benefits are not uniform. We also note that factors such as the donor’s microbiome (“super-donor” effect) and the need for diet or other co-interventions likely influence success. By including these points, we aim to show readers what that variability implies: namely, that FMT can have positive effects in some individuals but fails in others, highlighting the complexity of this approach.

As noted the effects of probiotics are weak and variable - and it is highly doubtful whether this is a viable approach .

Response:
Thank you for underscoring this. We have revised the manuscript to reflect the weak and inconsistent effects of probiotics on metabolic outcomes. In the section on microbiome-targeted therapies (Gut interventions, around line 581–590), we now emphasize that probiotic supplementation has shown only modest, highly variable benefits in glycemic control and insulin sensitivity, often on the order of very small HbA1c reductions (~0.2% as per meta-analyses). We also explicitly state that these results are variable across studies and individuals, casting doubt on probiotics as a reliable or standalone therapy for T2D. In fact, we mention that due to the inconsistency, probiotics are not yet a proven viable approach for diabetes management. This change aligns with your comment and ensures readers do not come away with an overly optimistic view of probiotics. We share the cautious stance, and the text now communicates that clearly.

In conclusion, regarding table 1 – it is probably useful, but it should be more cautiously phrased - most of the approaches are overrated.

Response:
We appreciate this overall feedback on Table 1. We have revisited Table 1 and toned down the phrasing to adopt a more cautious, balanced tone. We agree that many gut-related intervention effects can be overstated if not carefully described. Thus, we have made the following changes in Table 1: we inserted qualifying words (e.g., “may”, “proposed”, “preliminary evidence for…”) where appropriate, and removed any language that implied an overly guaranteed benefit. For example, instead of saying an intervention “robustly” does X, we might say it “has been shown to…” with context. We also added caveats in the Relevance column noting limitations or variability (such as for probiotics and FMT, as discussed). These edits collectively ensure that Table 1 presents a more balanced view of the approaches – acknowledging potential benefits while also reflecting the limitations and the early-stage or mixed nature of some evidence. We believe the table is now useful but also appropriately cautious, as per the Nutrients standards and your advice.

216 It is said that there isa release of GLP-1 from the coon – in fact it has never been demonstrated that GLP-1 from the colon contributes to the circulating levels – on the contrary people with total colectomy have completely normal both fasting and postprandial GLP-1 levels

Response:
Thank you for pointing out this important nuance. We have modified the text to avoid overstating the role of colonic GLP-1 in circulation. In the section describing incretin hormone secretion (around line 463–468), we originally noted that GLP-1 is secreted by L-cells in the ileum and colon. We now add that colon-derived GLP-1 does not meaningfully contribute to circulating GLP-1 levels. Specifically, we included the observation that patients with a total colectomy still exhibit normal fasting and postprandial GLP-1 levels, indicating the colon’s contribution is negligible. Thus, our text clarifies that the physiologically relevant GLP-1 for metabolic effects comes from the small intestine (ileal L-cells), not the colon. We believe this change addresses the issue: readers will no longer be misled into thinking colonic GLP-1 output is significant systemically.

225: it is a prerequisite for the effects of the GLP-1RAs on insulin secretion that there is sufficient residual beta cell function.

Response:
We appreciate this suggestion, and we have added the recommended clarification. In our discussion of GLP-1 receptor agonists (GLP-1 RAs) as a therapy, we now explicitly note that these agents require adequate residual β-cell function to exert their insulinotropic effects. In practice, this means GLP-1 RAs will only stimulate insulin secretion if the patient’s pancreatic β-cells are still present and capable of responding (which is generally the case in type 2 diabetes, but not in absolute insulin-deficient states). We inserted this point in the section describing GLP-1 RA mechanisms (revised text around line 418–423, and also in Table 1 relevance column for GLP-1 RAs). This additional detail makes it clear that a prerequisite for GLP-1 RA efficacy is the presence of functional β-cells. Thank you for highlighting this fundamental concept, which is now reflected in the manuscript.

231. it is highly controversial whether artificial sweeteners stimulate GLP-1 secretion – in the most robust studies there was no such effect.

Response:
Thank you for bringing up this point. We have revised the text to reflect the controversy regarding artificial sweeteners and GLP-1 secretion. In the Gut–Pancreas Axis section (around line 476–478), we no longer imply that non-nutritive sweeteners definitively stimulate GLP-1. Instead, we now state that evidence is conflicting and, notably, the most rigorous human studies have found no significant GLP-1 release effect from artificial sweeteners. We reference that while some early or small studies suggested a possible effect, larger controlled trials did not support it – thereby framing it as an unresolved issue. This change makes it clear to the reader that the notion of sweeteners triggering incretin release is highly debated and not a settled fact.

242. As already mentioned, the actual contribution by GIP to the actions of tirzepatide remains unclear – other potent GIP/GP-1 co-agonists do not have antidiabetic properties in excess of the GLP-1 part . The case for glucagon is better (probably because it also acts on the GL1 receptor )

Response:
We acknowledge this reiteration and have incorporated this perspective into our revision. Consistent with our earlier changes (see Response to Comment 9), we now stress in the manuscript that the contribution of GIP receptor agonism in tirzepatide’s efficacy remains unclear, since other GLP-1+GIP co-agonists haven’t outperformed GLP-1 alone. We amended the text around line 307–314 to reflect that tirzepatide’s extraordinary results likely involve factors beyond a simple additive GIP effect (e.g., its molecular design or dosing). We also added a note that combining GLP-1 with glucagon agonism might offer more demonstrable synergy – indeed, GLP-1/glucagon dual agonists show distinct benefits (glucagon’s action on GLP-1 receptors and raising energy expenditure), which suggests a clearer complementary mechanism. This is mentioned in our discussion of emerging multi-agonist therapies (Section 4). In summary, we have made sure to present tirzepatide’s mechanism as complex and unique, and not to overstate GIP’s role without evidence. Thank you for reinforcing this point; our revised text now accurately conveys the uncertainty and contrasts GIP vs. glucagon co-agonism as you indicated.

252: the role of SCFA to stimulate GLp-1 is highly controversial (and wrong)

Response:
We acknowledge this comment, which echoes the earlier point about SCFAs and GLP-1. As noted in our response to Comment 10, we have removed the claim that SCFAs definitively stimulate GLP-1 release. Throughout the manuscript (including line 322 in Table 1 and the discussion around lines 1303–1310), we now present this as a controversial and unproven idea. We clearly state that evidence for SCFAs inducing GLP-1 secretion in humans is weak and contradictory, thereby addressing the concern. In summary, we agree with the reviewer and have corrected the text accordingly, so that the role of SCFAs in GLP-1 release is no longer portrayed inaccurately.

263-277: If the authors had written: “it has been proposed”, all of this would be acceptable. The actions of these metabolites are almost entirely hypothetical

Response:
Thank you for this suggestion. We have rephrased the relevant sentences in the manuscript to explicitly indicate the hypothetical nature of the metabolite actions discussed in that section. Instead of stating those mechanisms as if they were established, we now introduce them with qualifiers such as “it has been proposed that...” and “hypothesized to...”. For example, in the integrated discussion of microbial metabolites (revised lines 759–767), we changed assertions to conditional language (“has been proposed to improve X” rather than “improves X”). We also add phrases like “remains largely hypothetical” when summarizing those multi-factorial mechanisms. These changes ensure the reader understands that the actions of certain gut metabolites (short-chain fatty acids, etc.) on T2D pathophysiology are speculative and not confirmed. We agree that this is a more accurate way to present the content, and we have implemented it as recommended.

305: the misconception regarding the hindgut hypothesis has been alluded to above.

Response:
Yes, we have addressed the “hindgut hypothesis” misconception as noted above (see Response to Comment 12). To reiterate, we have corrected both instances in the manuscript where this term was used. By the time we mention it around line 305 (the gut–liver section), we had already updated the earlier context. In the revised version, we ensure consistency: whenever “hindgut hypothesis” is referenced, we clarify it’s a misnomer and actually refers to mid-gut mechanisms. Thus, the line 305 instance has been aligned with the correction made earlier, so that the misconception is resolved throughout the paper. Thank you for pointing out both occurrences; we have made sure they are handled identically.

306: The increased GLP-1 response is invariable – not just “often”

Response:
Thank you for the correction. We have adjusted our wording to convey that the post-bariatric surgery increase in GLP-1 response is consistently observed (essentially invariable). In the original text we had used “often” or similar phrasing, which understated the regularity of this phenomenon. We changed it to terms like “consistently” or “invariably” to reflect that virtually all RYGB patients show an exaggerated GLP-1 surge after meals. This edit is made in the bariatric surgery context (Table 1 and line 415–420) where we discuss the effects of Roux-en-Y gastric bypass. The revised description now accurately indicates that the enhanced GLP-1 response is a near-universal outcome of that procedure, rather than an occasional one.

311: this ref is not the best regarding stool transplantation - the leading group is that of Max Niewdorf

Response:
Thank you for this bibliographic suggestion. We have reconsidered our reference on fecal microbiota transplantation (FMT) outcomes. In the revised manuscript, we updated the citation to a more seminal study by a leading group (e.g., Prof. Max Nieuwdorp’s group) when discussing FMT’s effects on insulin sensitivity. We agree that referencing the most authoritative studies strengthens the manuscript. Specifically, we have now cited a prominent trial from Nieuwdorp’s team (as well as a recent follow-up if available) in the context of our FMT discussion (line 432–439 and reference list). This change ensures that our review points to the best available evidence on stool transplantation. Thank you for pointing this out; using the top-tier references indeed improves the credibility and quality of our review.

349: the physiological role of oxyntomodulin is unclear - only after bariatric surgery are the levels sufficient for this peptide to have significant effects.

Response:
We have taken this comment into account and modified our discussion of oxyntomodulin in the gut–brain axis section. We now explicitly note that the normal physiological role of oxyntomodulin remains unclear because circulating levels of this peptide are typically low, and only after interventions like bariatric surgery do levels rise high enough to potentially exert significant effects (such as appetite suppression or glucose regulation). In the revised text (gut endocrine hormones section), we mention that oxyntomodulin’s impact is mostly observed in post-bariatric physiology and that its function under usual conditions is still being investigated. By adding this context, we ensure readers understand that oxyntomodulin is not a major player under baseline conditions. Thank you for highlighting this; our manuscript now accurately reflects the uncertainty about oxyntomodulin’s role.

351: GLP-1 does not cross the Blood brain barrier – it interacts with neurons in the circumventricular organs. The most significant expression of the receptors is also found there .

Response:
Thank you for this clarification. We have corrected our description of central GLP-1 action to align with the current understanding. In the revised text, we state that peripherally derived GLP-1 does not readily cross the blood–brain barrier (BBB). Instead, GLP-1’s effects on the brain are mediated via areas that lack a full BBB, notably the circumventricular organs (such as the area postrema and other regions around the ventricles). We mention that GLP-1 receptors are highly expressed in these regions, allowing GLP-1 to influence neural activity by interacting with neurons there. This change is reflected in the gut–brain axis discussion (around line 548–553). We have removed any implication that GLP-1 freely enters the brain parenchyma. By emphasizing the role of circumventricular organs and vagal pathways, we now provide an accurate explanation: GLP-1 acts on the brain indirectly or via BBB-permeable zones, not by crossing into all brain areas.

357: it should be emphasized that ghrelin levels are deceased in obesity

Response:
We agree and have added this emphasis. In our discussion of gut hormones (gut–brain axis, appetite regulation), we now explicitly note that ghrelin levels are typically reduced in individuals with obesity. For instance, when mentioning ghrelin’s role, we include that patients with obesity (and thus many with T2D) tend to have chronically lower ghrelin concentrations compared to lean individuals. This piece of information puts our statements about post-bariatric ghrelin changes in context. The revised text (around line 637–644) now reads in part: “...ghrelin (the hunger hormone, which is paradoxically low in obesity)...”. By doing so, we underscore that baseline ghrelin is already suppressed in the obese state, aligning with your comment. This change ensures completeness and accuracy in our hormone profile description.

361: RYGB does not restrict food intake! On the contrary it allows rapid unhindered passage of nutrients from the esophagus to the more distal small intestine. The decreased food intake is exclusively due to decreased appetite.

Response:
Thank you for this correction. We have revised our description of Roux-en-Y gastric bypass (RYGB) to avoid calling it a “restrictive” procedure and to clarify the true mechanism of reduced intake. In the gut–brain axis section (appetite regulation), we now explain that RYGB does not mechanically restrict how much food can be eaten; rather, it allows food to pass quickly to the distal small intestine, and any reduction in food intake is almost entirely due to greatly diminished appetite and early satiety (driven by hormonal changes like high GLP-1/PYY and low ghrelin). We removed language implying that RYGB physically limits stomach capacity (since the gastric pouch is smaller but food transit is accelerated, not “restricted” in a conventional sense). The revised text (around line 631–636) emphasizes that RYGB’s effects on intake are through appetite/satiety signals, not a mechanical blockade. We appreciate this clarification – the manuscript now accurately reflects the current understanding of how RYGB works.

380: note that the GIP- GLP-1 studies referred to are produced in rodents – nothing is know regarding humans.

Response:
You’re correct, and we have added a clarifying note about the scope of evidence. In the section where we discuss combined GIP/GLP-1 agonism (this appears in our Table 1 and possibly an animal study reference), we now explicitly mention that those particular studies were in rodent models. For example, if we referred to research showing some GIP+GLP-1 effect, we append a phrase like “(demonstrated in rodents; human relevance unknown)”. This lets readers know that comparable evidence in humans is not yet available. We felt this was important to incorporate, so as not to over-extrapolate preclinical findings. This clarification is included in the revised text (likely near line 308–313 and a footnote to the reference). In essence, any discussion of a GIP+GLP synergy now comes with the caveat that it’s based on animal data and needs human confirmation.

403: there is consensus that GLP-1 does not increase EE in people

Response:
We appreciate this reminder. We have removed any suggestion that GLP-1 (or GLP-1 RAs) increase energy expenditure in humans. In the initial manuscript, we might have implied a slight effect of GLP-1 on energy expenditure (EE), but as you point out, the consensus is that GLP-1 does not significantly increase EE in people. In the revised text, wherever we discuss GLP-1’s effects (e.g., on weight loss or metabolism), we make it clear that weight loss from GLP-1 RAs is due to appetite reduction and decreased intake, not increased metabolic rate. If any phrasing hinted at an EE increase, it has been corrected or deleted. By aligning with the consensus, we avoid any inaccurate claims. The final version of the manuscript reflects that GLP-1’s impact on energy expenditure in humans is negligible or absent (e.g., see lines 403 and 129–130, which now do not attribute EE changes to GLP-1). Thank you for ensuring our statements stay accurate.

Response:
Thank you for the feedback on Figure 2. We have closely reviewed the figure and identified the inaccuracy you pointed out. The diagram had an element suggesting that PYY (peptide YY) directly interacts with the anterior pituitary, which is not a recognized physiological pathway. To address this: - We have removed or revised that part of the figure. In the updated Figure 2, we no longer depict PYY as acting on the anterior pituitary. Instead, PYY’s effects are limited to its known targets (e.g., brainstem and hypothalamic circuits related to appetite). The anterior pituitary reference has been deleted to avoid confusion. - Additionally, we have clarified the figure legend to caution that the figure is a schematic with simplified interactions, and we explicitly note that some connections are putative. This sets the proper context and avoids implying unsupported mechanisms. We agree that such graphics must be precise. The revised Figure 2 (and legend) should now be accurate and not suggest any “mystery” PYY–pituitary interaction. We appreciate your sharp eye in catching this; it helped us improve the figure’s fidelity.

Response:
You raise a very pertinent point about the complexity of GLP-1’s central effects, particularly concerning the arcuate nucleus (ARC). We have updated our description to avoid oversimplification. In the figure and text referring to GLP-1’s action in the brain, we now clarify that GLP-1 does not directly penetrate into the arcuate nucleus due to the blood–brain barrier. Instead, we mention the likely scenario: GLP-1 acts via the median eminence (a circumventricular organ), which is adjacent to the ARC and accessible to circulating GLP-1. We also allude to the possible role of tanycytes or other connecting cells that might convey signals from the CSF/ventricular system to ARC neurons, though this is speculative. Essentially, our Figure 2 legend and corresponding text note that GLP-1’s effect on arcuate POMC/NPY neurons is indirect, possibly mediated by neurons in areas lacking BBB (like median eminence) or via neuronal circuit loops. We refrain from depicting a direct GLP-1 → ARC pathway. By doing so, we incorporate the nuances you highlighted: acknowledging the communication between median eminence and ARC, and mentioning the hypothesis of tanycyte mediation (while labeling it as less likely). These clarifications ensure that our figure and narrative convey the complexity and uncertainty of GLP-1’s central pathways, rather than a misleading simple arrow. Thank you for guiding us to refine this aspect – it makes our discussion more accurate.

449-451 “Bile acids, SCFAs, and intestinal leakage-derived LPS modulate adipogenesis, energy expenditure, and inflammation by promoting adipose browning and activating the TLR4-NF- 450 κB pathway in adipose macrophages, respectively” . The evidence that this occurs in humans is really weak!

Response:
Thank you for drawing attention to this. We have softened the statement and added context to indicate the preliminary nature of this evidence in humans. The lines in question (now around 741–748 in the revised manuscript) have been rephrased as follows: instead of declaring that these factors do modulate adipose tissue in humans, we now say “have been suggested to modulate...”, and we explicitly note that this comes mainly from animal models or in vitro studies. We then add, “…however, direct evidence of significant adipose browning or macrophage activation by these pathways in humans is limited.” In other words, we present it as a hypothesis with weak human support. We also clarify which parts are specific to animal findings – e.g., SCFA-induced browning is mostly rodent data, LPS→TLR4 in adipose macrophages is inferred from mechanistic studies. This revision should alert readers that the idea of bile acids/SCFAs/LPS driving adipose browning and inflammation is not yet well-substantiated in human studies. We agree with your assessment and have adjusted the tone to be appropriately cautious here.

469: NAFLD is now MAFLD (recognizing that truly non-alcoholic cases are very rare) – OK, mentioned later, but why not introduce it now?

Response:
We appreciate this suggestion and have made the nomenclature update at first mention. In the revised manuscript, the first time we introduce NAFLD (non-alcoholic fatty liver disease) around line 667–670, we now immediately note that it is also called MAFLD (Metabolic Dysfunction-Associated Fatty Liver Disease). We explain briefly that the term MAFLD has been proposed to more accurately reflect the metabolic etiology, acknowledging that “truly non-alcoholic” cases are rare. We did have MAFLD mentioned later (around line 505 originally), but we agree it’s clearer to define it upfront. So now the text reads, for example: “...non-alcoholic fatty liver disease (NAFLD, also recently termed metabolic dysfunction–associated fatty liver disease, MAFLD)...”. This way, readers are aware of the new term from the outset. Later in the text when we use MAFLD, it will make immediate sense. Thank you for pointing this out; the early introduction of MAFLD should prevent confusion and is aligned with current trends in the literature.

532: it should be emphasized that the main reason for hepatic insulin resistance is steatosis which is very common in obesity and T2DM-. It also should be mentioned that one of the major factors regulating hepatic overproduction of glucose in T2DM is glucagon excess. On the whole, from a therapeutic standpoint glucagon is an extremely important factor in the regulation of hepatic metabolism .

Response:
Thank you for these excellent points. We have revised the Gut–Liver Axis section to emphasize two key aspects as suggested:
1. Hepatic steatosis as a driver of insulin resistance: We now explicitly state that the accumulation of fat in the liver (steatosis) is a primary cause of hepatic insulin resistance in obesity and T2D. We mention that fatty liver is extremely common in these conditions and directly contributes to the liver’s reduced response to insulin. This addition appears around line 777–781, reinforcing the central role of steatosis.
2. Glucagon excess in hepatic glucose overproduction: We have added a sentence noting that hyperglucagonemia is a major contributor to elevated hepatic glucose output in T2D. We explain that in T2D, inappropriate glucagon secretion by α-cells stimulates the liver to overproduce glucose. We further agree that from a therapeutic standpoint, glucagon’s role is critical – we briefly note that targeting glucagon signaling (e.g., using glucagon receptor antagonists or the effect of GLP-1 RAs on glucagon) is an area of interest. These points are inserted around line 781–787.
By incorporating these, we acknowledge that genetic predisposition aside, the two major pathophysiological mechanisms in T2D are indeed obesity-induced steatosis/insulin resistance and excessive glucagon driving hepatic glucose production. This has been highlighted in both the narrative and in concluding remarks. Thank you for pointing out their importance; our manuscript now clearly reflects the primacy of steatosis and glucagon in T2D liver pathology.

579; refs 156 – 157 Any updates regarding this compound (JKB-121) ?

Response:
We appreciate this query about JKB-121. JKB-121 is a TLR4 (Toll-like receptor 4) antagonist that we had mentioned in context of potential NAFLD/NASH therapies. We have researched the latest information on JKB-121 and updated the manuscript accordingly. In the Gut–Liver therapeutic section, we now note that JKB-121 unfortunately did not show significant efficacy in clinical trials. Specifically, a phase II study in NASH patients (Diehl et al., 2018) found no difference between JKB-121 and placebo in improving liver enzymes or histology, in part due to an unexpected high placebo response[3][4]. We have added this outcome to indicate that despite promising preclinical data (TLR4 being a logical target to reduce inflammation), JKB-121 has not proven effective in human studies to date. This update is reflected around line 813–818, and we have cited the study/report documenting the trial results. In summary, we now inform the reader that JKB-121’s development has not yielded positive clinical results. Thank you for prompting us to include this important update, which brings the review fully up to date on this compound.

588: there is agreement that there are no GLP-1 receptors on hepatocytes (but , as stated possibly on stellate cells although this is also controversial)

Response:
You are correct. We have clarified our text to reflect the consensus that hepatocytes do not express GLP-1 receptors. In the Gut–Liver Axis section (line 799–804), when discussing GLP-1’s effects on the liver, we now state that GLP-1 receptors are not found on hepatocytes according to current evidence. We then mention parenthetically that some studies have suggested GLP-1R might be present on hepatic stellate cells (which could mediate certain effects like anti-fibrotic actions), but we also note that this is controversial and not definitively proven. By including this, we cover the nuance that if GLP-1 has direct hepatic effects, it’s likely via non-parenchymal cells (stellate cells or via neural innervation), not hepatocytes. We have removed any implication that GLP-1 directly acts on hepatocyte receptors. This revision ensures our description aligns with the established understanding. Thank you for highlighting this detail – our manuscript now accurately conveys where GLP-1 receptors are (and aren’t) in the liver.

Response:
Thank you for this important reminder. We have updated Table 2 (Gut–Liver therapies) and the related text to address the role of brown adipose tissue (BAT) activation by TGR5 agonists in humans. In the TGR5 agonist entry, we now add a note that while TGR5 activation can induce BAT activity in animal studies, adult humans have relatively little BAT, so this mechanism may have limited impact in clinical practice. Specifically, we mention that adult BAT depots are small and often inactive, which means the therapeutic relevance of BAT activation is uncertain. We included a similar caveat in the text narrative (around line 929–935) when discussing bile acid-based treatments, highlighting that claims of “increased energy expenditure via BAT” should be viewed cautiously for humans. By doing so, we temper the implication that BAT activation plays a major role in adult human metabolism. The rest of Table 2’s descriptions remain as they were (for other treatments, which you found acceptable). We believe this addition satisfies the need for context about BAT’s importance (or lack thereof) in adults. Thank you for pointing this out – it helps ensure we don’t inadvertently overstate that aspect.

663: there is only one of the common SGLT2inhibitors (namely canagliflozin) that also affects SGLT1 and therefore glucose absorption in the gut – the others do not have effect.

Response:
Thank you for this clarification. We have added the specified detail to our Gut–Kidney Axis section. When discussing SGLT2 inhibitors, we now note that among the common SGLT2 inhibitors, only canagliflozin has substantial SGLT1 inhibitory activity in the intestine (and thus can slow intestinal glucose absorption). The other widely used SGLT2 inhibitors (such as empagliflozin, dapagliflozin, etc.) are highly selective for SGLT2 and have negligible effect on SGLT1 at therapeutic doses, so they do not influence gut glucose uptake. This information is added around line 1063–1070. We also mention the relevance: canagliflozin’s dual SGLT1/2 action might contribute to some of its postprandial glucose effects, whereas the others work almost exclusively via the kidney. By including this, we ensure our review accurately differentiates canagliflozin’s profile from other SGLT2 inhibitors. We appreciate the opportunity to clarify this point, as it is clinically noteworthy.

I miss a brief summary of the impressive effects of SGLT2 inhibitors on DKD in general! This is very important and has led to approval also for non-diabetic kidney disease.

Response:
You are absolutely right. SGLT2 inhibitors have shown remarkably positive effects on diabetic kidney disease (DKD), and this deserved mention. We have now added a short summary highlighting the renoprotective effects of SGLT2 inhibitors in our Gut–Kidney Axis section. Around line 1080–1088, we write that SGLT2 inhibitors substantially slow DKD progression and reduce the risk of renal failure, as evidenced by large outcome trials. We also note that these benefits have been so robust that SGLT2 inhibitors (e.g., dapagliflozin, canagliflozin) have been approved for chronic kidney disease even in non-diabetic patients, due to their kidney-protective effects (e.g., reduction in proteinuria and preservation of GFR). We cite the relevant landmark trials (such as CREDENCE, DAPA-CKD) in support. This new addition ensures readers appreciate that beyond glycemic control, SGLT2 inhibitors confer significant kidney benefits – a major development in diabetes care. Thank you for pointing out this omission; including it indeed strengthens our review’s completeness regarding clinical implications.

680: Again, these idea that microbiota metabolites have important effects on intestinal GLP-1 production are not supported by studies in humans at least not yet.

Response:
We understand and have already adjusted related statements in response to your earlier comments. To specifically address this: in the latter part of our discussion (Gut–Kidney Axis and integrated model), we ensure that any suggestion of microbiota-derived metabolites significantly enhancing GLP-1 production is presented as unproven in humans. For example, around line 1098–1104, when talking about gut microbiota and L-cells, we use cautious phrasing (“preclinical studies suggest... but human evidence is lacking”). We had similarly revised the gut–pancreas section. So throughout the manuscript, it should now be clear that while gut bacteria metabolites (like SCFAs, bile acids) could influence GLP-1 in animal or cell models, we do not have solid human data confirming large effects on GLP-1 secretion. In summary, we fully concur with the need for caution, and we have consistently reflected that stance. Thank you for reiterating it; our text now uniformly conveys the tentative status of this idea.

683: the expression of GLP-1 receptors in the kidneys is again a controversial issue – the only consensus site is the afferent arterioles! The most important study on the clinical effects is the FLOW trial!! Which very convincing and specifically deals with DKD

Response:
We have addressed the kidney GLP-1 receptor issue in part (in line with Comment 37), but let us expand here. In discussing GLP-1 effects on the kidney (Gut–Kidney Axis), we now note that GLP-1 receptor expression in the kidney is limited and controversial – the only well-established location is on the afferent arteriole of the glomerulus. We mention that other purported sites (tubules, etc.) are not clearly confirmed. Thus, any direct renal effects of GLP-1 are likely mediated through hemodynamic changes (e.g., afferent arteriole dilation) or indirect systemic effects. Furthermore, we took your strong hint and added a reference to the FLOW trial – the first dedicated renal outcomes trial with a GLP-1 RA (semaglutide) in diabetes. In the revised text (around line 1085–1092), we briefly state that the FLOW trial demonstrated significant renal benefits of semaglutide in T2D patients with CKD, underscoring GLP-1 RA’s role in protecting kidney function (the trial was stopped early for efficacy, as widely reported). We cite the NEJM publication of FLOW (2022/2023) as evidence. This directly ties into our prior addition about SGLT2 inhibitors – together highlighting that both GLP-1 RAs and SGLT2 inhibitors have reshaped DKD management. In summary, we now caution that GLP-1R presence in the kidney is only conclusively in afferent arterioles, and we reinforce the clinical importance of GLP-1 RAs in DKD by mentioning the FLOW trial results. Thank you for prompting these valuable additions; they improve the accuracy and relevance of our kidney section.

739: the authors mention that secreted GLP-1 may maintain the gut’s barrier function - That appears to be true, but even more important is the simultaneous secretion of GLP-2 which has marked barrier protecting effects.

Response:
Thank you, we have added this important point. In our discussion of intestinal barrier function (Gut–Immune axis, around line 1136–1144), we now note that while GLP-1 may have some beneficial effects on gut barrier integrity, it is the co-secreted hormone GLP-2 that plays a major role in maintaining and strengthening the intestinal barrier. We explain that GLP-2 is released from L-cells alongside GLP-1 and has well-documented trophic and barrier-protective actions on the gut epithelium (e.g. enhancing mucosal growth and tightening junctions). We inserted a sentence highlighting GLP-2’s contribution: e.g., “Notably, GLP-2 is co-secreted with GLP-1 and is a key factor in preserving gut barrier function, even more so than GLP-1 itself.” By including GLP-2, we ensure readers are aware that GLP-2 is largely responsible for the gut barrier effects sometimes attributed generally to “incretins.” We agree with your point and believe the text now reflects the correct emphasis on GLP-2 in this context.

755: There hasn’t been much mentioning of the AhR? I see that it comes later , but should be referred to or mentioned here

Response:
You’re right. The aryl hydrocarbon receptor (AhR) was introduced later in our review (in context of microbial metabolites like indole derivatives affecting immunity), but to improve flow, we have added an earlier mention of AhR in the section discussing gut microbiota metabolites and inflammation. Around line 903–909, when summarizing microbial signals, we now include a brief introduction: e.g., “Certain microbiota-derived molecules can activate host receptors (for instance, the aryl hydrocarbon receptor, AhR) that influence gut immunity…”. This way, readers see the term “AhR” earlier and are primed for the more detailed discussion that follows (where we talk about indole-3-propionic acid activating AhR and inducing IL-22, etc., later in the manuscript). By seeding the concept of AhR in advance, we ensure the later references to it (in Figures or text) are not jarring. We agree that this makes the narrative smoother. The Chinese translation of AhR (芳香烃受体) is also provided at first mention for clarity. Thank you for this suggestion; the text now introduces AhR in a logical place earlier in the review.

759-768 Th authors return to the effects of SCFAs on GLP-1 secretion - and as previously mentioned the evidence for this in people is very weak (in fact there is a lot of evidence to the contrary). In addition, while it is true that butyrate is a fuel for colonocytes, much less can be said about proprionate, the effect of which is humans is unclear – please outline the real metabolism of proprionate in humans ( in ruminants it represents an important stimulus to gluconeogenesis but not in humans . Acetate can be metabolized in certain tissues but what is the predominant fate of it? Rather than hailing the SCFAs as “SCFAs are a unifying beneficial thread” it would be nice to get some facts regarding their metabolism in humans.

Response:
Thank you for this very detailed and constructive critique. We have significantly reworked the section on SCFAs in the integrated discussion to address all these points: - SCFAs and GLP-1 secretion: We reiterate (as before) that evidence of SCFAs boosting GLP-1 in humans is weak and contradictory. At lines 759–767 (in revision), any statements about SCFAs increasing GLP-1 are heavily qualified or removed. We now say “it has been proposed” and cite that robust human studies do not confirm this effect (even mentioning evidence to the contrary, as you indicated). - Metabolism of propionate: We added a concise explanation of propionate’s fate in humans. We note that propionate is largely absorbed by the liver, where it enters the TCA cycle (as propionyl-CoA to succinyl-CoA) – it can be used for gluconeogenesis in the liver, but in healthy humans, propionate typically constitutes only a minor substrate for gluconeogenesis (unlike in ruminants, where it’s a major glucose source). We clarify that human studies suggest propionate is mostly metabolized in hepatocytes and does not dramatically raise glucose production under normal dietary conditions. - Metabolism of acetate: We also included what happens to acetate. We explain that acetate is the most abundant SCFA in circulation; it is taken up by many tissues (muscle, heart, brain) and utilized as an energy source by being converted to acetyl-CoA. We mention that a significant portion of acetate is also taken up by the liver for lipogenesis or oxidation to CO₂. In essence, we describe acetate’s predominant fate as a general metabolic fuel. - Toning down the “unifying beneficial thread”: We removed the phrase “SCFAs are a unifying beneficial thread” which was overly broad. In its place, we provide a more factual summary: e.g., “SCFAs have multiple effects – serving as energy sources and signaling molecules – that generally promote metabolic health (e.g., improved insulin sensitivity and gut integrity), though their net impact in humans is complex.” We incorporate the facts about metabolism to ground the discussion. These changes are reflected around lines 1299–1310 and 1315–1325 in the revised manuscript. We believe this addresses your concerns by giving readers concrete information about SCFA metabolism in humans and removing any misleading over-generalization. We acknowledge the controversies and make the discussion more factual and nuanced. Thank you for guiding us to strengthen this section; it is now much more aligned with the actual human physiology of SCFAs.

791: Define and explain the TLR4

Response:
Apologies for the oversight. We have ensured that TLR4 (Toll-like receptor 4) is defined at first mention in the manuscript. In the introduction where TLR4 first appears (line 117–119)[5], we now write “Toll-like receptor 4 (TLR4)” and briefly explain its role in the context: e.g., “TLR4, an immune receptor that recognizes endotoxin, triggering NF-κB inflammation…”. This provides readers the necessary understanding of the abbreviation. Subsequent uses of “TLR4” will then be clear. We double-checked and found that in one instance the “-like” was inadvertently omitted (as “Toll receptor 4”), which we corrected to “Toll-like receptor 4”. All appearances of TLR4 should now be clear and correctly formatted. Thank you for flagging this; the term is now properly introduced and explained.

Response:
We have taken a careful approach to revise Figure 4 (the integrated model figure) and its legend to address this concern. We agree that several elements in Figure 4 are based on hypotheses or controversial data, so we have done the following: - In the figure legend, we explicitly added a statement that “Many pathways illustrated (e.g., microbial metabolite effects on host metabolism) are hypothetical or proposed, and remain to be confirmed, especially in humans.” This sentence is prominently placed so readers immediately know to interpret the figure with caution. - We went through each arrow and label in the figure to ensure none implies an interaction that we have disclaimed in the text. For example, any arrow suggesting SCFAs → GLP-1 has been either removed or marked with a symbol (like a question mark) to denote uncertainty. We similarly marked other contentious links (like “IPA → β-cell protection via AhR”) with an asterisk that corresponds to a note about preliminary evidence. - We also adjusted figure callouts in the text (around line 1420–1430) to reiterate that Figure 4 is a conceptual framework including speculative connections. With these changes, Figure 4 now clearly emphasizes the hypothetical/controversial nature of many factors depicted. We believe this addresses your comment: readers will be alerted that the figure is a synthesis of emerging ideas and not all components are definitively proven. Thank you for this suggestion; it helped us improve the figure’s clarity and reliability.

1011: the efficacy of tirzepatide is unique and is not due to a simple combination of GIP and GLP-1 action – such co-agonists are not more effective than GLP-1 alone. The efficacy of tirzepatide was a serendipitous finding.

Response:
We have incorporated this insight into our discussion of tirzepatide and other co-agonists. In the manuscript, we now explicitly acknowledge that tirzepatide’s efficacy is unique and was somewhat unexpected (a serendipitous finding) rather than a predictable outcome of GIP+GLP-1 dual action. We mention that earlier attempts at combining GIP with GLP-1 did not yield superior results to GLP-1 alone, highlighting that tirzepatide’s dramatic success likely involves unique properties of that molecule (e.g., its specific receptor affinity profile, dosing, etc.) beyond just the dual agonism. This point is made in the context of our treatment section (around line 1230–1237). We used phrasing like: “Interestingly, tirzepatide’s remarkable efficacy was not simply the sum of GLP-1 and GIP effects – other GIP/GLP-1 co-agonists didn’t outperform GLP-1 alone – indicating tirzepatide’s benefits were somewhat unexpected and may reflect other optimized features.” We also use the word “serendipitous” to describe the discovery of its potent effect. This change ensures readers understand that tirzepatide’s result was novel and not easily generalizable just by adding GIP. It underscores that the field was surprised by how effective tirzepatide is, aligning with your comment.

1026 sevelamer should have been mentioned together with colesevelam.

Response:
Thank you for this suggestion. In the revised manuscript, we have added sevelamer alongside colesevelam when discussing bile acid sequestrants in the Gut–Liver therapy context. Originally, we only mentioned colesevelam (a bile-acid binding resin used in T2D trials), but we now include sevelamer as well, noting that it too is a bile acid-binding polymer that has been investigated for metabolic effects (and is used clinically for phosphate binding in CKD, with some glycemic and lipid benefits reported). We insert sevelamer in the same sentence as colesevelam (line 1254–1258), so it reads in effect: “…bile acid sequestrants (colesevelam, and sevelamer) that bind intestinal bile acids…have shown modest improvements in glycemic control…”. By doing so, we acknowledge both agents. We also ensure that any reference or footnote that was specific to colesevelam is adjusted if needed to cover sevelamer (though mechanistically similar, sevelamer’s data might be less extensive, but worth the mention as you indicated). This change makes that part of our review more comprehensive. We appreciate the opportunity to include sevelamer so readers get the full picture of resin-based therapies.

Reviewer 2 Report

Comments and Suggestions for Authors

The principal strengths of this review lie in its breadth of literature coverage, inclusion of high-impact and up-to-date references, and the clarity with which complex mechanistic pathways are communicated. The inclusion of summary tables and well-annotated figures enhances accessibility and aids in synthesizing a substantial body of evidence. Additionally, the discussion of shared mechanisms and an integrated multi-organ model is particularly valuable in presenting an up-to-date paradigm shift in T2D pathophysiology.

There are, however, several areas where minor revision would strengthen the manuscript:

Several core concepts—such as short-chain fatty acids (SCFAs), GLP-1 actions, and inflammatory signaling—are discussed in multiple sections, occasionally leading to redundancy. Streamlining these overlapping areas would improve coherence and conciseness.
Some discussion points on microbiome-targeted interventions (e.g., probiotics, FMT) may inadvertently overstate the strength of current clinical evidence. It is recommended to more clearly differentiate between established therapies, promising adjuncts, and experimental strategies, especially where evidence is still emerging or largely preclinical.
The clinical implications section might benefit from further elaboration on practical considerations such as patient stratification, cost-effectiveness, and feasibility of gut-directed therapies in real-world settings.

In summary, this is an up-to-date and informative review that makes a valuable contribution to the field. With minor revisions to address the points above, it will serve as an good resource for researchers and clinicians interested in this field.

Author Response

The submitted manuscript presents a scholarly and comprehensive review of the multifaceted roles of the gut and its “Gut-X axes” in the pathogenesis and management of type 2 diabetes (T2D). The work is timely and addresses an area of increasing relevance as the integration of metabolic, immune, and neuroendocrine pathways in diabetes comes to the fore. The manuscript’s organization is logical, progressing from basic concepts of gut homeostasis to the roles of gut-pancreas, gut-endocrine, gut-liver, and gut-kidney axes, culminating in clinical and translational perspectives.The principal strengths of this review lie in its breadth of literature coverage, inclusion of high-impact and up-to-date references, and the clarity with which complex mechanistic pathways are communicated. The inclusion of summary tables and well-annotated figures enhances accessibility and aids in synthesizing a substantial body of evidence. Additionally, the discussion of shared mechanisms and an integrated multi-organ model is particularly valuable in presenting an up-to-date paradigm shift in T2D pathophysiology.There are, however, several areas where minor revision would strengthen the manuscript:

1.Several core concepts—such as short-chain fatty acids (SCFAs), GLP-1 actions, and inflammatory signaling—are discussed in multiple sections, occasionally leading to redundancy. Streamlining these overlapping areas would improve coherence and conciseness.

Response: Thank you for highlighting the issue of redundancy. We agree with the reviewer and have adopted this suggestion by streamlining overlapping discussions of SCFAs, GLP-1, and inflammatory signaling throughout the manuscript. Repetitive background information has been consolidated to improve coherence and avoid unnecessary repetition. For example, the detailed explanation of GLP-1’s incretin effect is now provided only once (under the Gut–Pancreas Axis section), and later sections reference that explanation instead of restating it. Similarly, discussions of SCFAs and LPS/TLR4 inflammatory pathways that previously appeared in multiple sections have been tightened and cross-referenced. These changes make the text more concise and focused. We also added a note in the Discussion to acknowledge and minimize any remaining overlap. The revisions are located on Discussion, Limitations of the revised manuscript.

2.Some discussion points on microbiome-targeted interventions (e.g., probiotics, FMT) may inadvertently overstate the strength of current clinical evidence. It is recommended to more clearly differentiate between established therapies, promising adjuncts, and experimental strategies, especially where evidence is still emerging or largely preclinical.

Response: We appreciate this important suggestion. In response, we have modified the manuscript to temper the language around microbiome-targeted interventions and clearly distinguish the level of evidence for each approach. Specifically, in the Clinical Implications and Future Directions sections, we now differentiate between well-established treatments versus emerging or experimental strategies. For example, we explicitly state that certain interventions like fecal microbiota transplantation (FMT) for T2D are still experimental and not part of standard care, despite their promise. We have also emphasized that probiotics and similar adjunct therapies have modest effects and are considered complementary rather than standalone treatments. These clarifications ensure we do not overstate the current clinical evidence. The revised text now provides a more balanced view, highlighting which strategies are supported by robust clinical data and which remain investigational.

3.The clinical implications section might benefit from further elaboration on practical considerations such as patient stratification, cost-effectiveness, and feasibility of gut-directed therapies in real-world settings.

Response: We agree with the reviewer’s suggestion and have expanded the discussion of clinical and practical considerations in the Clinical Implications portion of the manuscript. In the revised version, we added text to address: (a) patient stratification – acknowledging that not all T2D patients will respond equally to gut-targeted interventions and that biomarkers or microbiome profiles may be needed to identify those most likely to benefit; (b) cost-effectiveness – noting the cost and standardization challenges of implementing microbiome analyses or advanced therapies (e.g., the expense and infrastructure needed for routine microbiome sequencing or FMT, and the importance of weighing costs versus benefits); and (c) feasibility – discussing real-world factors such as the complexity of fecal transplants (donor screening, regulatory oversight), patient adherence to diet or probiotic regimens, and multidisciplinary coordination. We have thereby provided a more comprehensive commentary on the translational aspects of Gut-X axis therapies, as recommended. These additions aim to give readers a sense of the practical real-world challenges and considerations before such therapies can be widely adopted.

4.In summary, this is an up-to-date and informative review that makes a valuable contribution to the field. With minor revisions to address the points above, it will serve as an good resource for researchers and clinicians interested in this field.

Response: We thank the reviewer for this positive overall assessment. We are delighted that the reviewer found our review “up-to-date and informative” and deemed it a valuable contribution. We have carefully made the minor revisions as recommended in the comments above to further improve the manuscript. Thank you again for your encouraging remarks.

Reviewer 3 Report

Comments and Suggestions for Authors

In the narrative review, the authors clearly elucidated gut-mediated crosstalk with the pancreas, endocrine system, liver, and kidneys in T2DM. The article is well structured and well written. The concept Gut-X axes and initial hypothesis are clearly presented. The authors used the PRISMA guideline to prepare the review. Nevertheless, I would like to suggest that the authors expand the subsection ‘Microbial inflammatory factor’ a little in order to provide a more comprehensive view of TLR recognition dysfunction and the persistence of intestinal antigens in T2DM. In addition, the authors could discuss the mechanisms of metabolic memory in diabetes mellitus in the context of Gut-X axes regulation.

Author Response

Response:

Comment 1 “Expand the subsection ‘Microbial inflammatory factor’ to provide a more comprehensive view of TLR (Toll-like receptor) recognition dysfunction and the persistence of intestinal antigens in T2DM.”

Response: Thank you for this insightful suggestion. We have implemented the recommendation by expanding the subsection on “Microbial inflammatory factors” to discuss Toll-like receptor (TLR) dysfunction and the persistence of gut-derived antigens in type 2 diabetes. In the revised manuscript, on page 8 (Gut–Pancreas Axis section), we added new content explaining how a compromised intestinal barrier in T2D leads to chronic exposure of the immune system to microbial products, resulting in sustained TLR activation. We describe that in T2D, TLR4 signaling may be over-activated or insufficiently regulated, so that even low-grade, persistent presence of endotoxin (LPS) and other intestinal antigens continuously triggers inflammation. This persistent TLR stimulation is now highlighted as a key mechanism linking gut dysbiosis to chronic systemic inflammation and β-cell stress in T2D. We also cited relevant studies to support this mechanism.

Comment 2: “Discuss the mechanisms of metabolic memory in diabetes mellitus in the context of Gut-X axes regulation.”

Response: We appreciate this suggestion and have added a new discussion about metabolic memory in the context of Gut-X axes. In the revised manuscript, we introduce the concept of “metabolic memory” – the phenomenon wherein early metabolic dysregulation (e.g., periods of hyperglycemia or inflammation) has lasting effects even after glycemic control is improved – and relate it to gut-mediated pathways (see page 16, last paragraph of Discussion). We propose that chronic gut-derived inflammation and other long-term alterations in the Gut-X axes might contribute to metabolic memory in diabetes. For example, we discuss how an initial dysbiotic state or gut barrier dysfunction could induce epigenetic changes or persistent immune activation that continue to drive insulin resistance and tissue damage even after blood glucose is brought under control. Conversely, early interventions targeting the gut microbiome and intestinal health might help mitigate the metabolic memory effect by addressing one root cause of chronic inflammation. We have thus integrated the concept of metabolic memory into our discussion, noting it as an important consideration for the long-term impact of Gut-X axis dysregulation in T2D.

Article Menu

Type 2 Diabetes and the Multifaceted Gut-X Axes

Further Information

Guidelines

MDPI Initiatives

Follow MDPI