Chronic Kidney Disease-Associated Defect in Humoral Immune Response Is Driven by Inflammation

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript presents a well-designed translational study investigating the mechanisms underlying impaired humoral immunity in chronic kidney disease (CKD). The combination of clinical data, in vitro experiments, and murine models strengthens the validity of the conclusions. The findings challenging the initial hypothesis about uremic toxins' direct role and highlighting systemic inflammation as the primary driver are novel and clinically relevant. However, several limitations should be addressed to improve the manuscript's impact and clarity.

Major Weaknesses and Suggestions

1. While the study robustly links systemic inflammation (e.g., CRP, IL-6) to humoral defects, the specific inflammatory pathways or mediators disrupting lymphoid architecture or lymphocyte function remain unexplored. The manuscript does not elucidate how inflammation directly impairs germinal center formation or T-B cell interactions.
2. The human cohort data focus on bulk lymphocyte counts and Tfh frequencies but lack depth in profiling regulatory or exhausted immune populations (e.g., Tregs, PD-1+ T cells, myeloid-derived suppressor cells). This limits understanding of the inflammation-driven immunosuppressive mechanisms.
3. The study dismisses uremic toxins' direct effects but does not fully address their potential indirect contributions to inflammation (e.g., via monocyte activation). The positive correlations between some toxins and immune responses (Figure 3E) are unexplained and may reflect confounding factors.

Author Response

The manuscript presents a well-designed translational study investigating the mechanisms underlying impaired humoral immunity in chronic kidney disease (CKD). The combination of clinical data, in vitro experiments, and murine models strengthens the validity of the conclusions. The findings challenging the initial hypothesis about uremic toxins' direct role and highlighting systemic inflammation as the primary driver are novel and clinically relevant. […]

We thank the reviewer #1 for his positive appreciation of our work.

Major Weaknesses and Suggestions

1. While the study robustly links systemic inflammation (e.g., CRP, IL-6) to humoral defects, the specific inflammatory pathways or mediators disrupting lymphoid architecture or lymphocyte function remain unexplored. The manuscript does not elucidate how inflammation directly impairs germinal center formation or T-B cell interactions.

 

We agree with the reviewer that this represents a limitation of our study. However, we would like to emphasize that the primary objective of our work was not to elucidate the molecular mechanisms by which inflammation directly impairs germinal center formation or T–B cell interactions.

Rather, based on extensive literature suggesting that (i) the defect in immune responses associated with CKD results from the direct effects of uremic toxins (review in Espi M. et al, Toxins (Basel). 2020 May 6;12(5):300.), (ii) indoxyl sulfate (IS), a major protein-bound uremic solute, is an activating ligand of the aryl hydrocarbon receptor (AhR; see: Dou, L. et al, Kidney Int 2018, 93,986-999; Schroeder JC. et al, Biochemistry 2010;49:393-400.), and (iii) AhR signaling impairs B- and T-cell functions (see: Li, J. et al, J. Immunol. 2017, 199, 3504–3515.; Villa, M. et al, EMBO J. 2017, 36, 116–128; and Vaidyanathan, B. et al, J. Exp. Med. 2017, 214, 197–208; Piper, C. et al, Cell Rep. 2019, 29, 1878–1892.e7.; Gandhi, R. et al Nat. Immunol. 2010, 11, 846–853; Funatake, C.J. et al, J. Immunol. 2005, 175, 4184–4188), we initially sought to test the hypothesis that IS activates AhR in lymphocytes, thereby directly altering T- and/or B-cell function.

Our data clearly demonstrate that neither IS, nor the other tested PBURS directly impair T–B cell interactions. This unexpected finding prompted us to re-examine both our clinical and experimental data and to explore alternative mechanisms. This approach ultimately led us to identify inflammation as a key driver of the defective humoral response in CKD, which constitutes the central message of the manuscript.

We therefore consider that a detailed investigation of the molecular pathways through which inflammation impairs humoral immunity falls outside the scope of the present study and represents the focus of future work. Nevertheless, to more thoroughly address the reviewer’s concern, we have revised the manuscript to explicitly acknowledge this limitation and to expand the Discussion. In this revised section, we review recent literature from non-CKD contexts (including infection and experimental models) that provides plausible mechanistic hypotheses for how inflammation may impair humoral responses. These hypotheses will hopefully serve as a conceptual framework and starting point for future studies in CKD.

 

Changes:

Page 15: Another limitation is the fact that, while our work establishes a strong association between systemic inflammation and impaired humoral immunity, the specific molecular mechanisms by which inflammation disrupts lymphoid architecture and/or impairs lymphocyte function in CKD patients remain undefined. Recent studies have highlighted how lymphoid stromal networks and chemokine gradients, such as CXCL13 produced by reticular cells, are essential for B cell zone organization and effective humoral responses [37] Sustained inflammatory cytokine signaling (e.g., IL‑6) has been implicated in altering stromal cell homeostasis, which could disrupt fibroblastic reticular cell niches and chemokine gradients required for germinal center maintenance [38]. In line with this idea, the marked lymph node shrinkage in the inflammatory CKD mouse model, which was not observed in non-inflamed CKD mice, is reminiscent of the structural disruptions reported in other inflammatory states, including sepsis and sterile inflammation [39]. Such architectural disorganization has been shown to impair antigen retention and presentation, ultimately weakening B-cell responses in experimental models [39].

 

Page 16: Chronic inflammation–associated impairment of adaptive immunity is not unique to CKD, and evidence from other disease contexts may inform the CKD setting. Inflammation–induced immunosuppression has been particularly well studied in sepsis[40], where it is recognized as a multifactorial process characterized by the expansion of regulatory immune populations [41–43] and by the depletion of effector lymphocytes. Consistent with this literature, we confirmed marked lymphopenia in our CKD cohort, the severity of which was associated with impaired vaccine responses. A recent large cohort study, including CKD patients, further established a strong positive correlation between the CRP/lymphocyte ratio and CKD [44]. Furthermore, although the retrospective nature of our study did not allow for assessing regulatory populations in ESKD patients, other reports have demonstrated elevated levels of myeloid-derived suppressor cells in this population [45,46].

 

Pages 16 and 17: Another thoroughly explored pathological setting that mirrors the coexistence of immune insufficiency and chronic inflammation is inflammaging [47]. Inflammaging promotes immune deficiency through persistent, low-grade inflammatory signaling that progressively erodes immune cell function and renewal. Chronic exposure to proinflammatory cytokines (IL-6, TNF, IL-1β) driven by senescent cells and their senescence-associated secretory phenotype (SASP) induces T-cell exhaustion, impairs antigen presentation and skews hematopoiesis toward dysfunctional myeloid lineages [48,49]. Sustained activation of NF-κB and cGAS–STING pathways by DNA damage and cytosolic nucleic acids leads to maladaptive innate immune activation and interferon desensitization, blunting antimicrobial responses [50]. Concurrent metabolic dysregulation, including mitochondrial dysfunction, NAD⁺ depletion and mTOR hyperactivation, compromises immune cell fitness and longevity [51] . Age-related impairment of macrophage efferocytosis and gut microbiota dysbiosis further perpetuate chronic inflammation while weakening immune resolution and immune education [52]. Together, these interconnected mechanisms create an immune system that is chronically inflamed yet functionally incompetent, resulting in increased susceptibility to infection, poor vaccine responses, mirroring the situation observed in CKD [53,54].

 

1. The human cohort data focus on bulk lymphocyte counts and Tfh frequencies but lack depth in profiling regulatory or exhausted immune populations (e.g., Tregs, PD-1+ T cells, myeloid-derived suppressor cells). This limits understanding of the inflammation-driven immunosuppressive mechanisms.

 

We agree with the reviewer. This limitation is now clearly highlighted in the revised version of our MS.

 

Changes:

Page 16 : Furthermore, although the retrospective nature of our study did not allow for assessing regulatory populations in ESKD patients, other reports have demonstrated elevated levels of myeloid-derived suppressor cells in this population [45,46].

1. The study dismisses uremic toxins' direct effects but does not fully address their potential indirect contributions to inflammation (e.g., via monocyte activation). The positive correlations between some toxins and immune responses (Figure 3E) are unexplained and may reflect confounding factors.

 

We agree with the reviewer that our data do not rule out a potential indirect contribution of uremic toxins to CKD-associated inflammation. Indeed, several independent studies have shown that monocytes exposed to uremic toxins, in particular IS, produce pro-inflammatory cytokines (e.g., Kim H.Y. et al., FASEB J. 2019, 33, 10844–10858; Nakano T. et al, Circulation. 2019;139(1):78–96; and Espi M. et al., Toxins (Basel) 2020, 12(5):300), which could contribute to CKD-associated inflammation. While this concept was already discussed in the original manuscript, we have expanded this section in the revised version to address the reviewer’s comment.

Of note however if this was the main mechanism driving CKD-associated inflammation we would expect to observe an (inverse) correlation between the level of PBURS and the parameter associated with the immune response, which was not the case. Based on this and the fact that many other mechanisms contribute to promoting inflammation in CKD patients, we concluded that inflammation rather than just the uremic milieu is the main driver of CKD-associated defect in humoral immune response

 

Changes:

Pages 14 and 15: In parallel, CKD is associated with elevated production of pro-inflammatory cytokines, notably through persistent monocyte activation by toxins such as IS, which signals via the AhR. Interestingly, although numerous studies show that IS activates AhR signaling in monocytes [26–29], our data suggest that this does not occur in lymphocytes. This difference may reflect the fact that IS is a negatively charged, hydrophilic molecule that cannot passively cross cell membranes and instead requires specific transporters for uptake. IS primarily enters cells via organic anion transporters (OATs); however, OAT expression in immune cells remains poorly characterized. Thus, the differential impact of IS on monocytes versus lymphocytes may be due to variations in OAT expression. Regardless, CKD-induced inflammation exacerbates kidney injury, perpetuating a vicious cycle. If CKD-associated inflammation was solely determined by uremic toxins accumulation paralleling eGFR decline, we would expect a perfect positive correlation between uremic toxins levels and defective vaccine responses in our ESKD cohort and CKD mouse model—yet this was not observed. This suggests that, beyond uremic toxins, additional factors contribute to CKD-associated inflammation.

 

Regarding the positive correlations observed between certain toxins and immune responses, we agree with the reviewer that these associations likely reflect confounding factors, and their statistical significance was lost after correction for multiple testing.

 

Changes:

See revised Figure 3E

 

Page 10: Correlation analyses between each toxin and four parameters of the vaccine-induced response (anti-RBD IgG, serum neutralization capacity, spike-specific CD4⁺ T cells, and spike-specific Tfh cells) yielded 48 associations in total. None of them reached statistical significance (Figure 3E).

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript examines a significant and timely issue concerning the mechanisms of impaired humoral immunity in end-stage renal disease (ESKD), combining patient data with in vitro and in vivo experimental models. The clinical cohort is well characterized, the data are clear, and the integration of data from humans and mice is an advantage. However, several mechanistic conclusions, particularly those that exclude the role of uremic toxins and suggest inflammation as the main cause of immune dysfunction, are currently overestimated in relation to the data presented. In order to strengthen impact of the study, it is necessary to clarify the experimental hypotheses, incorporate additional controls and adopt a more cautious interpretation of negative and correlative findings.

Please find my comments below

Comment 1:

The mechanistic hypothesis set out in this manuscript is based on the activation of the aryl hydrocarbon receptor (AhR) signalling pathway by indoxyl sulfate (IS) in an experimental co-culture system. Despite the use of 40 μM IS and the quantification of free IS in the culture medium, as outlined in the 'Methods' section, no direct or indirect evidence of AhR activation is provided under these experimental conditions. Including information about AhR signalling would strengthen the conclusions, given its central importance in the proposed mechanism. One possible way to achieve this would be to demonstrate the involvement of AhR, for example by employing an AhR agonist or antagonist control, or by conducting an analysis  of AhR target gene. In the absence of evidence, it is unclear whether the lack of functional effects reflects the true strength of T–B cooperation in AhR activation or insufficient pathway engagement.

Comment 2:

In the co-culture experiments, only one concentration of indoxyl sulfate was tested, so the absence of a dose-response analysis limits the range of interpretation of the findings. Testing a range of IS concentrations, or providing stronger justification for the chosen concentration, would make the findings more credible.

Comment 3:

The conclusion that IS does not have a negative effect on T-B cooperation is mainly based on results evaluated by B-cell proliferation. However, there is no positive control to demonstrate that the co-culture system can detect functional the inhibition of T-B cooperation. Including a known inhibitory condition (e.g. blockade of CD40-CD40L interactions) would allow readers to assess the sensitivity and reability of the experimental system.

Comment 4:

Analyses of the correlations between uremic toxin levels and vaccine-induced immune system parameters involved a large number of comparisons. It is not clear whether correction for multiple testing was applied. Please clarify this issue.

 

Author Response

The manuscript examines a significant and timely issue concerning the mechanisms of impaired humoral immunity in end-stage renal disease (ESKD), combining patient data with in vitro and in vivo experimental models. The clinical cohort is well characterized, the data are clear, and the integration of data from humans and mice is an advantage. […]

We thank reviewer #2 for the encouraging feedback on our work

Comment 1:

The mechanistic hypothesis set out in this manuscript is based on the activation of the aryl hydrocarbon receptor (AhR) signalling pathway by indoxyl sulfate (IS) in an experimental co-culture system. Despite the use of 40 μM IS and the quantification of free IS in the culture medium, as outlined in the 'Methods' section, no direct or indirect evidence of AhR activation is provided under these experimental conditions. Including information about AhR signalling would strengthen the conclusions, given its central importance in the proposed mechanism. One possible way to achieve this would be to demonstrate the involvement of AhR, for example by employing an AhR agonist or antagonist control, or by conducting an analysis  of AhR target gene. In the absence of evidence, it is unclear whether the lack of functional effects reflects the true strength of T–B cooperation in AhR activation or insufficient pathway engagement.

We fully agree with the reviewer that we do not provide direct evidence of AhR signaling in human lymphocytes cultured with (high dose of) indoxyl sulfate (IS). The main reason the experiments suggested by the reviewer were not performed is that we did not observe any impact of IS on T–B cell interactions in our co-culture system. This effectively rules out our initial hypothesis that IS at an uremic rate directly alters T- and/or B-cell function (through AhR signaling or a different mechanism) as an explanation for the CKD-associated defect in humoral immune responses. This unexpected finding prompted us to re-examine both our clinical and experimental data and to explore alternative mechanisms. This approach ultimately led us to identify inflammation as a key driver of the defective humoral response in CKD, which constitutes the central message of the manuscript.

 

Importantly, and as suggested by the reviewer, the lack of effect of IS on T–B cell interactions in vitro does not exclude a potential role for IS-mediated AhR activation in the CKD-associated defect in humoral immunity in vivo. For instance, IS-induced AhR signaling in monocytes can trigger proinflammatory cytokine production (e.g., Kim H.Y. et al., FASEB J. 2019, 33, 10844–10858; Nakano T. et al, Circulation. 2019;139(1):78–96; and Espi M. et al., Toxins (Basel) 2020, 12(5):300), which contributes to the chronic inflammatory milieu observed in CKD—a recognized driver of humoral immune dysfunction. However, if monocyte AhR activation by IS was the main mechanism, one would expect an inverse correlation between circulating IS levels and measures of humoral immunity in patients, which was not observed. This suggests that while IS-mediated monocyte activation may contribute to CKD-associated inflammation, it cannot fully account for the observed defect in humoral immune responses. This element of discussion has been added to the revised version of the MS.

 

Finally, it should be noted that, to our knowledge there is currently no published evidence demonstrating direct activation of AhR by indoxyl sulfate (IS) in lymphocytes. One possible explanation is that IS is a negatively charged, hydrophilic molecule that cannot passively diffuse across cell membranes and instead requires specific transporters for cellular uptake. IS is known to enter cells primarily via organic anion transporters (OATs); however, OAT expression in immune cells, particularly in lymphocytes, remains poorly characterized. The differential effects of IS on monocytes versus lymphocytes may therefore reflect differences in OAT expression between these cell types. Although testing this hypothesis is beyond the scope of the present work, this discussion has now been incorporated into the revised manuscript.

 

 

Changes:

Pages 14 and 15: In parallel, CKD is associated with elevated production of pro-inflammatory cytokines, notably through persistent monocyte activation by toxins such as IS, which signals via the AhR. Interestingly, although numerous studies show that IS activates AhR signaling in monocytes [26–29], our data suggest that this does not occur in lymphocytes. This difference may reflect the fact that IS is a negatively charged, hydrophilic molecule that cannot passively cross cell membranes and instead requires specific transporters for uptake. IS primarily enters cells via organic anion transporters (OATs); however, OAT expression in immune cells remains poorly characterized. Thus, the differential impact of IS on monocytes versus lymphocytes may be due to variations in OAT expression. Regardless, CKD-induced inflammation exacerbates kidney injury, perpetuating a vicious cycle. If CKD-associated inflammation was solely determined by uremic toxins accumulation paralleling eGFR decline, we would expect a perfect positive correlation between uremic toxins levels and defective vaccine responses in our ESKD cohort and CKD mouse model—yet this was not observed. This suggests that, beyond uremic toxins, additional factors contribute to CKD-associated inflammation.

 

Comment 2:

In the co-culture experiments, only one concentration of indoxyl sulfate was tested, so the absence of a dose-response analysis limits the range of interpretation of the findings. Testing a range of IS concentrations, or providing stronger justification for the chosen concentration, would make the findings more credible.

The reviewer is correct in noting that a dose–response analysis would, in principle, strengthen the assessment of a relationship between indoxyl sulfate (IS) exposure and AhR activation in lymphocytes. This point is particularly relevant given the substantial inter-individual variability in free IS levels observed among patients with end-stage kidney disease (ESKD) (Figure 3B and S1D).

However, because we did not observe any effect on T/B-cell interactions when cells were exposed to an IS concentration carefully selected to reflect the upper range observed in patients undergoing chronic hemodialysis (see paragraph below), we did not pursue additional dose–response experiments. We considered such experiments unlikely to be informative, as effects observed at supraphysiological concentrations would not meaningfully reflect the clinical situation the model was designed to recapitulate. Conversely, we did not anticipate any effect at lower concentrations, a notion supported by the absence of correlation between free IS levels and parameters of the humoral immune response in the ESKD cohort (Figures 3B and 3C).

Regarding the selection of this single, physiologically relevant “high” IS concentration, we rigorously quantified free IS levels—since only the unbound fraction is biologically active—in serum samples from a large cohort of ESKD patients. We then performed dose-escalation experiments in culture medium, measuring free IS at increasing total IS concentrations. These experiments showed that, in standard culture conditions, supplementation with 40 µM total IS resulted in approximately 25 µM free IS, corresponding to the upper range observed in ESKD patients (Supplementary Figure S1E–G). Coculture experiments were therefore performed in the presence of indoxyl sulfate (IS group) or in its absence (controls: culture medium alone or an osmolar control with 40 µM KCl) (Figure 2B). This rationale has now been clearly detailed in the revised version of the manuscript.

Changes:

Page 9: Since only unbound IS exerts biological activity, we quantified both total and free IS levels in the serum of 74 ESKD patients using UPLC (Supplementary Figure S1C). Free IS levels correlated moderately with total IS concentrations (p<0.001; R² = 0.55; Supplementary Figure S1D). In standard culture medium, supplementation with 40 µM total IS yielded ~25 µM free IS that reflected the upper range observed in ESKD patients (Supplementary Figure S1E–G). Cocultures were therefore conducted in the presence of indoxyl sulfate (IS group, red) or in its absence (controls: culture medium alone or osmolar control with 40 µM KCl) (Figure 2B).

 

Comment 3:

The conclusion that IS does not have a negative effect on T-B cooperation is mainly based on results evaluated by B-cell proliferation. However, there is no positive control to demonstrate that the co-culture system can detect functional the inhibition of T-B cooperation. Including a known inhibitory condition (e.g. blockade of CD40-CD40L interactions) would allow readers to assess the sensitivity and reability of the experimental system.

The reviewer is absolutely right, and we thank him for pointing out this issue. There was a conceptual error in panel A of Supplementary Figure 1 that created confusion regarding the functioning of the T–B cooperation assay. This assay is a robust and well-established system that has been developed and used in our laboratory for several years (see Charmetant et al., Sci Transl Med, 2022).

The readout of the assay is B-cell proliferation, which requires two distinct activation signals. Signal 1 is BCR-dependent and is provided by crosslinking surface IgM using monoclonal antibodies added to the culture medium. Signal 2 is T-cell–dependent (see Supplementary Fig. 1B, second column). Because the CD4⁺ T cells introduced into the coculture are not pre-activated, they express low basal levels of CD40L and can only provide help to B cells after receiving TCR stimulation. To achieve this, allogeneic CD4⁺ T cells are used. Their TCR repertoire contains approximately 1–10% of alloreactive T cells whose TCRs directly recognize intact allogeneic HLA molecules. This allorecognition allows T-cell activation, leading to upregulation of CD40L and the delivery of the second activation signal to B cells. In contrast, coculture with syngeneic T cells does not induce B-cell proliferation, as these T cells are, by definition, non-reactive to self-HLA due to thymic negative selection (this control has been added to revised Sup Fig 1B).

Panel A of Supplementary Figure 1 has been revised to better illustrate the bidirectional nature of this assay: B cells activate allogeneic T cells through TCR engagement, and the activated T cells in turn provide costimulatory signals that drive B-cell proliferation.

Changes:

Revised Supplementary Figure 1A and 1B

Pages 8 and 9: Briefly, purified B cells were stimulated through BCR crosslinking using anti-IgM monoclonal antibodies (signal 1) and cocultured with non-preactivated allogeneic CD4⁺ T cells. B-cell proliferation was monitored over six days by CellTrace Violet dilution. In this system, allogeneic B cells activate a fraction of CD4⁺ T cells through direct TCR recognition of intact allogeneic HLA molecules, leading to T-cell activation and upregulation of costimulatory signals. This T-cell–derived help provides the second activation signal (signal 2), which is strictly required—together with BCR engagement—for efficient B-cell proliferation (Supplementary Figure S1A, B).

Comment 4:

Analyses of the correlations between uremic toxin levels and vaccine-induced immune system parameters involved a large number of comparisons. It is not clear whether correction for multiple testing was applied. Please clarify this issue.

We fully agree with the reviewer’s comment. The statistical analyses examining the relationship between uremic toxin levels and parameters of the humoral immune response have now been appropriately adjusted for multiple testing.

As a result, the few previously observed “unexpected” positive correlations (in the old version of Figure 3E), likely representing false-positive findings, are no longer present, which further reinforces the main message of the manuscript.

These corrected results have been incorporated throughout the revised version, replacing the previous analyses. We also specified the statistical analysis made in the statistical paragraph of the material and methods section.

Changes:

See revised Figure 3E

 

Page 10 : Correlation analyses between each toxin and four parameters of the vaccine-induced response (anti-RBD IgG, serum neutralization capacity, spike-specific CD4⁺ T cells, and spike-specific Tfh cells) yielded 48 associations in total. None of them reached statistical significance (Figure 3E).

Page 24 : For multiple comparisons of linear regression data made in Figure 3E, we applied a Bonferroni correction.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This paper systematically explores the mechanism of defective humoral immune responses in patients with chronic kidney disease (CKD) by analyzing clinical and biological data and combining in vitro experiments with animal models. It points out that systemic inflammation is the main driving factor, rather than the direct effect of uremic toxins. The research design is relatively complete and has certain academic value. However, there are some problems in the article, and it is recommended to revise it and then review it. The specific comments are as follows:

1. The literature [1] cited for the global prevalence data of CKD was published in 2022. It is recommended to supplement epidemiological literature from the past 2 - 3 years.
2.  In the second paragraph, “40 - fold higher risk” does not specify which population it is compared with. It is recommended to provide supplementary explanations. 
3. In the animal experiment section, only the name of the institution is mentioned, and the animal ethics approval number or license number is not provided. 
4. In the patient cohort, there are 80 cases in the ESKD group and 26 cases in the HV group. The sample size is relatively small, especially for the healthy control group. It is recommended to supplement the impact of sample size limitations on the conclusions in the discussion section.
5. In Figure 1, the r² values in panel D are described as “0.727” and “0.787” in the text, but are labeled as “R²” in the figure. It is recommended to unify the symbols. 
6. The elaboration of the mechanism of “inflammation leading to lymphatic structure destruction” is relatively brief. It is recommended to elaborate in detail in combination with relevant recent studies. 
7. In the Materials and Methods section: In the description of the UPLC method in Section 4.4, key chromatographic conditions such as column temperature and flow rate are not mentioned. In the animal model section, the specific number of mice in each group is not specified.
8. The format of the references is not unified, and the proportion of literature from the past five years is relatively low. It is recommended to supplement key literature published after 2021.

Author Response

This paper systematically explores the mechanism of defective humoral immune responses in patients with chronic kidney disease (CKD) by analyzing clinical and biological data and combining in vitro experiments with animal models. It points out that systemic inflammation is the main driving factor, rather than the direct effect of uremic toxins. The research design is relatively complete and has certain academic value. However, there are some problems in the article, and it is recommended to revise it and then review it. The specific comments are as follows:

1. The literature [1] cited for the global prevalence data of CKD was published in 2022. It is recommended to supplement epidemiological literature from the past 2 - 3 years.

 

We thank the reviewer for this constructive suggestion. Accordingly, we have added the recently published data from the Global Burden of Disease Study 2023 (Lancet. 2025 Nov 22;406(10518):2461–2482) to the Introduction of the revised manuscript.

 

Changes:

 

Page 4: In a recent analysis focused on adults aged 20 years and older over the period 1990 to 2023, from 204 countries and territories, global age-standardized prevalence of CKD in adults was 14·2%, a relative rise of 3·5 from 1990 [1].

 

1.  In the second paragraph, “40 - fold higher risk” does not specify which population it is compared with. It is recommended to provide supplementary explanations. 

We apologize for this lack of precision. The paragraph in the introduction has been revised, and the requested information are now provided.

 

Changes:

Page 4 : Multiple large epidemiologic studies have shown that patients with end-stage kidney disease (ESKD) on dialysis experience higher sepsis-related mortality than the general population. In U.S. registry data, annual sepsis-attributable death rates in dialysis patients were approximately 100–300 fold greater than in the general population, even after age, race, and diabetes mellitus stratification, where rates remained nearly 50-fold higher [4]. These disparities persisted when accounting for competing causes of death, with sepsis mortality still 30–45-fold above expected general population rates [4].

 

1. In the animal experiment section, only the name of the institution is mentioned, and the animal ethics approval number or license number is not provided. 

We apologize for this lack of precision. The ethics approval number for the animal experiments has been added in the M&M section in the revised MS.

 

Changes:

Page 21: The ethics approval number for this protocol is APAFIS#24971-2020022410571420.

 

1. In the patient cohort, there are 80 cases in the ESKD group and 26 cases in the HV group. The sample size is relatively small, especially for the healthy control group. It is recommended to supplement the impact of sample size limitations on the conclusions in the discussion section.

While we respectfully note that Reviewer #4 considered the cohort size appropriate, we have nonetheless explicitly addressed the potential impact of the limited sample size on our conclusions in the revised Discussion section to respond to the reviewer’s comment.

 

Changes:

Page 15: Among the limitations of our study is the relatively small cohort size, which negatively impacts the statistical power and could explain the absence of a detectable correlation between uremic toxin levels and parameters of the humoral immune response in ESKD patients. However, it should be emphasized that this lack of association was consistently observed across both in vitro and in vivo models. Moreover, despite its size, the cohort was sufficient to clearly demonstrate a robust relationship between systemic inflammation and impaired humoral immunity.

 

1. In Figure 1, the r² values in panel D are described as “0.727” and “0.787” in the text, but are labeled as “R²” in the figure. It is recommended to unify the symbols.

 

We apologize for this lack of precision. The consistency of the revised manuscript has been improved, and the R² symbol is now used uniformly throughout the text and figures in the revised version.

 

Changes:

Page 7: As previously reported, anti-RBD IgG titers strongly correlated with neutralizing capacity, following an identical relationship in both groups (both p<0.0001, R² = 0.727 for HV; R² = 0.787 for ESKD; Figure 1D).

 

6. The elaboration of the mechanism of “inflammation leading to lymphatic structure destruction” is relatively brief. It is recommended to elaborate in detail in combination with relevant recent studies. 

We thank the reviewer for this constructive suggestion. The Discussion has been revised to provide a more detailed account of how inflammation leads to lymphatic structure destruction, integrating additional recent relevant studies.

 

Changes:

Pages 15 and 16: Sustained inflammatory cytokine signaling (e.g., IL‑6) has been implicated in altering stromal cell homeostasis, which could disrupt fibroblastic reticular cell niches and chemokine gradients required for germinal center maintenance [38]. In line with this idea, the marked lymph node shrinkage in the inflammatory CKD mouse model, which was not observed in non-inflamed CKD mice, is reminiscent of the structural disruptions reported in other inflammatory states, including sepsis and sterile inflammation [39]. Such architectural disorganization has been shown to impair antigen retention and presentation, ultimately weakening B-cell responses in experimental models [39].

 

1. In the Materials and Methods section: In the description of the UPLC method in Section 4.4, key chromatographic conditions such as column temperature and flow rate are not mentioned. In the animal model section, the specific number of mice in each group is not specified.

 

We apologize for this lack of precision. We specified these points in the revised version of our manuscript:

 

Changes:

Pages 20 and 21: Chromatographic separation was performed at 26 °C on a Waters Acquity UPLC BEH C18 column (1.7 µm, 100 x 2.1 mm) with a Waters Acquity UPLC BEH C18 Van Guard column (1.7 µm, 5 x 2.1 mm). The mobile phase consisted of a 50 mM ammonium formate buffer (mobile phase A, pH 3.0) and methanol (mobile phase B). A linear gradient elution was used to separate the compounds at a flow rate of 0.3 mL/min, and started at 98% A, followed by a composition change to 90% A in 7 min. In the next 9 minutes, the mobile phase changed to 100% B and was held for 3 min.

 

 

1. The format of the references is not unified, and the proportion of literature from the past five years is relatively low. It is recommended to supplement key literature published after 2021.

We apologize for this lack of precision. The reference section has been reformatted according to the instructions to the authors from Toxins.

 

Changes:

See the revised references section

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

General Comments

Authors propose a solid translational study aimed at elucidating the mechanisms underlying defective humoral immunity in end-stage chronic kidney disease, challenging the widely accepted hypothesis that uremic toxins,particularly indoxyl sulfate,directly impair adaptive immune responses.

The manuscript addresses a clinically relevant and timely topic and is strengthened by an integrated design combining human data, in vitro assays, and murine models. A major strength of the work is the rigorous testing of a biologically plausible hypothesis, ultimately demonstrating that indoxyl sulfate does not directly impair T–B cell cooperation and redirecting the focus toward systemic inflammation as a key determinant of humoral immune dysfunction. The study is generally well structured, follows an appropriate IMRaD format, and ethical aspects appear adequately addressed.

Nevertheless, while the overall quality is high, several issues limit the strength of the conclusions, particularly regarding statistical rigor, handling of clinical confounders, and depth of mechanistic interpretation. Addressing these points would substantially strengthen the manuscript.

Specific Comments

The clinical cohort is appropriately sized and the demonstration of impaired quantitative but preserved qualitative humoral responses is convincing. However, the stratification into responders and non-responders requires clarification. The absolute size of each subgroup should be explicitly reported, and the criteria used to define response should be better justified. Given the potential influence of age, diabetes, dialysis vintage, and comorbidities on vaccine responsiveness, the absence of a multivariable analysis limits the claim that inflammation, reflected by CRP levels and lymphocyte counts,is independently associated with impaired humoral immunity. Incorporating a regression model would substantially reinforce this conclusion.

The extensive analysis of uremic toxins is a strength, but the statistical handling of correlation analyses is insufficient. Multiple correlations are tested without correction for multiple comparisons, increasing the risk of false-positive findings. Appropriate adjustment or explicit discussion of this limitation is required.

The in vitro coculture experiments convincingly show that indoxyl sulfate does not impair T–B cooperation. However, this does not fully exclude a role for AhR signaling in CKD-associated immune dysfunction, as other uremic or inflammation-related ligands may activate this pathway. The conclusions regarding AhR should therefore be more circumscribed.

The murine models are well chosen and support the conclusion that inflammation, rather than uremia alone, impairs T-dependent humoral responses. Nonetheless, the inflammatory state induced is relatively acute compared with the chronic low-grade inflammation observed in human CKD, and this translational limitation should be more explicitly discussed. In addition, mechanistic links between inflammation and immunosuppression remain largely inferential, as no direct data are provided on regulatory or suppressive immune populations.

Finally, the conclusion that inflammation is the “primary driver” of humoral dysfunction may be overstated. Other contributing mechanisms, such as immunosenescence or metabolic dysfunction, cannot be excluded and should be acknowledged. Similarly, translational implications regarding anti-inflammatory strategies should be framed more cautiously.

 

Author Response

Authors propose a solid translational study aimed at elucidating the mechanisms underlying defective humoral immunity in end-stage chronic kidney disease, challenging the widely accepted hypothesis that uremic toxins,particularly indoxyl sulfate,directly impair adaptive immune responses.

The manuscript addresses a clinically relevant and timely topic and is strengthened by an integrated design combining human data, in vitro assays, and murine models. A major strength of the work is the rigorous testing of a biologically plausible hypothesis, ultimately demonstrating that indoxyl sulfate does not directly impair T–B cell cooperation and redirecting the focus toward systemic inflammation as a key determinant of humoral immune dysfunction. The study is generally well structured, follows an appropriate IMRaD format, and ethical aspects appear adequately addressed.

Nevertheless, while the overall quality is high, several issues limit the strength of the conclusions, particularly regarding statistical rigor, handling of clinical confounders, and depth of mechanistic interpretation. Addressing these points would substantially strengthen the manuscript.

We thank the reviewer #4 for his positive appreciation of our work and for his constructive suggestions.

Specific Comments

The clinical cohort is appropriately sized and the demonstration of impaired quantitative but preserved qualitative humoral responses is convincing. However, the stratification into responders and non-responders requires clarification. The absolute size of each subgroup should be explicitly reported, and the criteria used to define response should be better justified. Given the potential influence of age, diabetes, dialysis vintage, and comorbidities on vaccine responsiveness, the absence of a multivariable analysis limits the claim that inflammation, reflected by CRP levels and lymphocyte counts, is independently associated with impaired humoral immunity. Incorporating a regression model would substantially reinforce this conclusion.

We fully agree with the reviewer and thank him for this constructive and insightful suggestion. In response, we performed an additional analysis of the clinical cohort that now includes a multivariable regression model assessing the impact of key patient characteristics on the response to COVID-19 vaccination in ESKD, defined by the presence of neutralizing antibodies. The absolute numbers of responders and non-responders, together with a clearer justification of the response criteria, have been explicitly reported in the revised manuscript. Importantly, the results of the multivariable analysis, now included in the revised version of the MS, demonstrate that CRP level is the only variable independently associated with vaccine responsiveness, whereas age, diabetes, dialysis vintage, and comorbidities were not. These findings substantially strengthen the manuscript and further support its central conclusion that the chronic kidney disease–associated defect in humoral immune responses is primarily driven by inflammation.

Changes:

See revised Figure 5

See new revised TableS2

 

Page 11: To identify the features associated with the humoral defect in ESKD patients, we stratified the latter into “responders” (n=43/73, 59%) and “non-responders” (n=30/73, 41%) according to their serum neutralization capacity after two vaccine doses (Figure 5A)

Pages 11 and 12: In exploratory univariate analyses, seven variables—age, BMI, presence of cardiac disease, and levels of CRP, prealbumin, lymphocyte count, and PTH—differed between responders and non-responders (p < 0.1; Supplementary Table S2, Figure 5B and Figure 5C). These variables were subsequently entered into a multivariate model (Supplementary Table S2), which identified CRP level as the only parameter independently associated with the humoral response to vaccination in ESKD patients (OR=0.86 [0.76-0.96], p=0.015).

The extensive analysis of uremic toxins is a strength, but the statistical handling of correlation analyses is insufficient. Multiple correlations are tested without correction for multiple comparisons, increasing the risk of false-positive findings. Appropriate adjustment or explicit discussion of this limitation is required.

We fully agree with the reviewer’s comment. The statistical analyses examining the relationship between uremic toxin levels and parameters of the humoral immune response have now been appropriately adjusted for multiple testing.

As a result, the few previously observed “unexpected” positive correlations (in the old version of Figure 3E), likely representing false-positive findings, are no longer present, which further reinforces the main message of the manuscript.

These corrected results have been incorporated throughout the revised version, replacing the previous analyses.

Changes:

See revised Figure 3E

Page 10: Correlation analyses between each toxin and four parameters of the vaccine-induced response (anti-RBD IgG, serum neutralization capacity, spike-specific CD4⁺ T cells, and spike-specific Tfh cells) yielded 48 associations in total. None of them reached statistical significance (Figure 3E).

Page 24: For multiple comparisons of linear regression data made in Figure 3E, we applied a Bonferroni correction

The in vitro coculture experiments convincingly show that indoxyl sulfate does not impair T–B cooperation. However, this does not fully exclude a role for AhR signaling in CKD-associated immune dysfunction, as other uremic or inflammation-related ligands may activate this pathway. The conclusions regarding AhR should therefore be more circumscribed.

We fully agree with the reviewer that, the lack of effect of IS on T–B cell interactions in vitro does not exclude a potential role for IS-mediated AhR activation in the CKD-associated defect in humoral immunity in vivo.

In fact, several studies have reported that, IS-induced AhR signaling in monocytes can trigger proinflammatory cytokine production (e.g., Kim H.Y. et al., FASEB J. 2019, 33, 10844–10858; Nakano T. et al, Circulation. 2019;139(1):78–96; and Espi M. et al., Toxins (Basel) 2020, 12(5):300), which contributes to the chronic inflammatory milieu observed in CKD—a recognized driver of humoral immune dysfunction. However, if monocyte AhR activation by IS was the main mechanism, one would expect an inverse correlation between circulating IS levels and measures of humoral immunity in patients, which was not observed. This suggests that while IS-mediated monocyte activation may contribute to CKD-associated inflammation, it cannot fully account for the observed defect in humoral immune responses. This element of discussion has been added to the revised version of the MS.

 

Changes:

Pages 14 and 15: In parallel, CKD is associated with elevated production of pro-inflammatory cytokines, notably through persistent monocyte activation by toxins such as IS, which signals via the AhR. Interestingly, although numerous studies show that IS activates AhR signaling in monocytes [26–29], our data suggest that this does not occur in lymphocytes. This difference may reflect the fact that IS is a negatively charged, hydrophilic molecule that cannot passively cross cell membranes and instead requires specific transporters for uptake. IS primarily enters cells via organic anion transporters (OATs); however, OAT expression in immune cells remains poorly characterized. Thus, the differential impact of IS on monocytes versus lymphocytes may be due to variations in OAT expression. Regardless, CKD-induced inflammation exacerbates kidney injury, perpetuating a vicious cycle. If CKD-associated inflammation was solely determined by uremic toxins accumulation paralleling eGFR decline, we would expect a perfect positive correlation between uremic toxins levels and defective vaccine responses in our ESKD cohort and CKD mouse model—yet this was not observed. This suggests that, beyond uremic toxins, additional factors contribute to CKD-associated inflammation.

The murine models are well chosen and support the conclusion that inflammation, rather than uremia alone, impairs T-dependent humoral responses. Nonetheless, the inflammatory state induced is relatively acute compared with the chronic low-grade inflammation observed in human CKD, and this translational limitation should be more explicitly discussed. In addition, mechanistic links between inflammation and immunosuppression remain largely inferential, as no direct data are provided on regulatory or suppressive immune populations.

We fully agree with the reviewer’s comment. These limitations are now explicitly discussed in the revised version of the MS.

Changes:

Page 12: To investigate the role of systemic inflammation in the defective humoral response observed in ESKD, we exploited the transient inflammation that occurs at the end of the adenine phase and resolves during the subsequent resting period, while the animals remain in a CKD state. Although the intensity and dynamics of inflammation in this murine model differ from the chronic low-grade profile seen in human CKD—potentially limiting the direct extrapolation of our conclusions—the model uniquely allows us to dissect the effects of inflammation independently of uremic toxins, providing causal insights that are not readily attainable in patients.

Pages 15 and 16: Another limitation is the fact that, while our work establishes a strong association between systemic inflammation and impaired humoral immunity, the specific molecular mechanisms by which inflammation disrupts lymphoid architecture and/or impairs lymphocyte function in CKD patients remain undefined. Recent studies have highlighted how lymphoid stromal networks and chemokine gradients, such as CXCL13 produced by reticular cells, are essential for B cell zone organization and effective humoral responses [37]. Sustained inflammatory cytokine signaling (e.g., IL‑6) has been implicated in altering stromal cell homeostasis, which could disrupt fibroblastic reticular cell niches and chemokine gradients required for germinal center maintenance [38]. In line with this idea, the marked lymph node shrinkage in the inflammatory CKD mouse model, which was not observed in non-inflamed CKD mice, is reminiscent of the structural disruptions reported in other inflammatory states, including sepsis and sterile inflammation [39]. Such architectural disorganization has been shown to impair antigen retention and presentation, ultimately weakening B-cell responses in experimental models [39].

Page 16: Furthermore, although the retrospective nature of our study did not allow for assessing regulatory populations in ESKD patients, other reports have demonstrated elevated levels of myeloid-derived suppressor cells in this population [45,46].

Finally, the conclusion that inflammation is the “primary driver” of humoral dysfunction may be overstated. Other contributing mechanisms, such as immunosenescence or metabolic dysfunction, cannot be excluded and should be acknowledged. Similarly, translational implications regarding anti-inflammatory strategies should be framed more cautiously.

We thank the reviewer for this insightful comment. We have moderated our conclusion regarding inflammation as the “primary driver” of CKD-associated humoral dysfunction and are now more cautious in discussing potential anti-inflammatory strategies in the revised manuscript. We also appreciate the suggestion to consider additional contributing mechanisms, such as immunesenescence or metabolic dysfunction, which may indeed interact with inflammation in the complex pathophysiology of CKD-associated humoral defects.

Changes:

Pages 16 and 17: Another thoroughly explored pathological setting that mirrors the coexistence of immune insufficiency and chronic inflammation is inflammaging [47]. Inflammaging promotes immune deficiency through persistent, low-grade inflammatory signaling that progressively erodes immune cell function and renewal. Chronic exposure to proinflammatory cytokines (IL-6, TNF, IL-1β) driven by senescent cells and their senescence-associated secretory phenotype (SASP) induces T-cell exhaustion, impairs antigen presentation and skews hematopoiesis toward dysfunctional myeloid lineages [48,49]. Sustained activation of NF-κB and cGAS–STING pathways by DNA damage and cytosolic nucleic acids leads to maladaptive innate immune activation and interferon desensitization, blunting antimicrobial responses [50]. Concurrent metabolic dysregulation, including mitochondrial dysfunction, NAD⁺ depletion and mTOR hyperactivation, compromises immune cell fitness and longevity [51] . Age-related impairment of macrophage efferocytosis and gut microbiota dysbiosis further perpetuate chronic inflammation while weakening immune resolution and immune education [52]. Together, these interconnected mechanisms create an immune system that is chronically inflamed yet functionally incompetent, resulting in increased susceptibility to infection, poor vaccine responses, mirroring the situation observed in CKD [53,54].

Page 17: In conclusion, our data suggest that the humoral defect in CKD patients is primarily a consequence of chronic inflammation—driven in part by uremic toxins acting on innate immunity (rather than directly on adaptive immune cells, as we initially suspected), but also by additional patient- and treatment-related factors. […]. Whether targeting inflammation in CKD patients -as tested to reduced inflammation induced-cardiovascular mortality in dialysis patients [55]- would also help restoring immune competence, improving vaccine responsiveness and ultimately reducing infectious complications, is an attractive hypothesis that remain to be confirmed in future interventional studies.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for your excellent, detailed responses. I agree completely with the authors' arguments and am very impressed with the way the comments have been addressed.

Reviewer 4 Report

Comments and Suggestions for Authors

I thank the authors for their careful and thorough revision of the manuscript.

The revised version adequately addresses all major points raised during the previous review. In particular, the addition of the multivariable analysis substantially strengthens the clinical interpretation of the data, the correction for multiple testing in correlation analyzes improves statistical rigor, and the discussion has been appropriately refined to circumscribe mechanistic claims and acknowledge translational limitations.