SGLT2 inhibition modulates NLRP3 inflammasome activity via ketones and insulin in diabetes with cardiovascular disease

Sodium–glucose cotransporter 2 (SGLT2) inhibitors reduce cardiovascular events in humans with type 2 diabetes (T2D); however, the underlying mechanism remains unclear. Activation of the NLR family, pyrin domain-containing 3 (NLRP3) inflammasome and subsequent interleukin (IL)-1β release induces atherosclerosis and heart failure. Here we show the effect of SGLT2 inhibitor empagliflozin on NLRP3 inflammasome activity. Patients with T2D and high cardiovascular risk receive SGLT2 inhibitor or sulfonylurea for 30 days, with NLRP3 inflammasome activation analyzed in macrophages. While the SGLT2 inhibitor’s glucose-lowering capacity is similar to sulfonylurea, it shows a greater reduction in IL-1β secretion compared to sulfonylurea accompanied by increased serum β-hydroxybutyrate (BHB) and decreased serum insulin. Ex vivo experiments with macrophages verify the inhibitory effects of high BHB and low insulin levels on NLRP3 inflammasome activation. In conclusion, SGLT2 inhibitor attenuates NLRP3 inflammasome activation, which might help to explain its cardioprotective effects.

monocytes and neutrophils. Please discuss the mechanism of effects of SGLT2 inhibitors [REDACTED] on serum IL-18 and IL1B?
3. It is stated that some subjects were already taking Metformin and were allowed to continue this medication throughout the trial. However, it is not partivcularly clear who these subjects were and whether these subjects were used in any of the assays in which only a subset or participants were used (importantly RNAseq). This is critical given that Metformin likely has NLRP3-altering capabilities due to its AMPK-stimulating actions. 4. SGLT2 group lost more weight. Is weight-loss the biggest driver of observations of reduced inflammation? Please discuss.
Reviewer #3 (Remarks to the Author): In their paper «SGLT2 inhibition modulates NLRP3 inflammasome acitvitx in patients with type 2 diabetes and cardiovascular disease» Kim et al. evaluate in vivo in a randomized manner in a relatively small cohort of patients well matched patients the effects of empagliflocin and glimepiride for 30 days on inflammatory pathways.
This design and the hypothesis that interleukin-1beta might be involved is novel, interesting and clinically potentially important. The measured multiple parameters in plasma oft he patients and in monocytes obtained form the recruited subjects.
They found that both drugs similarly lowered plasma glucose, but empagliflocin led to a greater reduction of interleukin-1 beta.
Empagliflocin, but not glimepiride led to a marked increase in plasma levels of fasting betahydroxy butyrate and a significant decrease in urinc acid and fatsing insulin, whiel free fatty acids increased. Furthermore, empagliflocin improved insulin sensitivity, while glimepirides szimulated insulin secretion as expected.
This study is novel, interesting and downsides, howeverwell designed and written and illustrated with appropriate figures summarizing the data.
There are a few downsides, however: 1. The mechanism by which empagliflocin would activate the inflammasome NLRP3 is not addressed. NLRP3 is activated by cristals, among them cholesterol and uric acid. As uric acid changes, could this be involved? Additional data should be provided on this aspect. 2. In the ex vivo experiments NLRP3 only tended to be reduced by empagliflocin, while interleucin-1beat was significantly reduced -is there thus another mechanisms involved rather than through NLRP3? Although RNA sequencing provided insights, this is not properly follwoed up. 3. Would empagliflocin in vitro still reduce interleukin-1beta secretion or expression in macrophages, if NLRP3 expression would be silenced beforehand? 4. Although beta-hydroxybutyrate has been described to suppress NLRP3 in mice, what ist he mechanisms in human macrophages?

Response to Reviewers' Comments
We would like to thank the reviewers for their comments and criticisms, which helped to greatly improve the quality of this manuscript.

Reviewers' comments:
Reviewer #1 (Remarks to the Author): So Ra Kim et al SGLT2 inhibition modulates NLRP3 inflammasome activity in patients with type 2 diabetes and cardiovascular disease The authors assessed the effect of SGLT2 inhibitor empagliflozin on NLRP3 inflammasome activity in patients with type 2 diabetes and high cardiovascular risk who received either an SGLT2 (n=29) inhibitor or a sulfonylurea (n=32) for 30 days. The paper describes the many metabolic changes observed and particularly draws attention to the changes of the NLRP3 inflammasome activation analyzed in macrophages. [REDACTED].

Comment 1:
The paper is well-written, the statistics are robust and appropriate and the descriptions of the methods generally good. I have no major criticisms. Comment 2: I applaud the number and detail of the investigations in what is both a randomized trial and a physiological investigation.

Comment 3:
The risk, however, which is not outlined in the limitations, is that with very many metabolic measurements it is possible to conclude a multiplicity of 'pathways' which may be, among other possibilities, direct, indirect, or even agent-specific. So the hypothesis has to be tentative and in deference to the careful writing of the authors I acknowledge that they use the terms 'proof of principle' and their bottom line 'The present data suggest that these mechanisms might help to explain the cardioprotective effects of SGLT2 inhibitor in humans' is appropriate language.
Response: We deeply respect your comprehensive understanding of our research. As the reviewer commented, based on well-established evidences of the critical contribution of the NLRP3 inflammasome/IL-1β to cardiovascular (CV) diseases, our study is a 'proof-of-concept' randomized controlled trial aimed to investigate the potential mechanism of SGLT2 inhibitors regarding CV protection and regulation of inflammasome activity.
Based on your comments, the limitations have been revised as follows: "A limitation of this study is that we did not assess whether the changes in inflammasome activity by SGLT2 inhibitor could be linked to the improvement in CV outcomes .... The present study is a proof-of-concept RCT aimed to elucidate the glucose-independent mechanism of SGLT2 inhibitors regarding CV protection. ....

Besides our findings, it is important to bear in mind that there may be multiplicity of
other pathways or mechanisms which can be directly or indirectly involved in the protection of CVD by SGLT2 inhibitors. Nevertheless, this is the first study to provide evidence that SGLT2 inhibitor suppresses NLRP3 inflammasome activation in patients with T2D at high risk of CVD." The manuscript has been revised to include:

Reviewer #2 (Remarks to the Author):
This study by Lee and colleagues was initiated to investigate why SGLT2 inhibitors reduce adverse CVD events in T2D patients and whether this effect was independent of its glucoselowering effects. It's an interesting study that reveals an important finding that SGLT2 inhibitors may have additional clinical benefits lowering NLRP3 inflammasome activation, which itself has been linked to many chronic inflammatory diseases. Given SGLT inhibitors increase BHB the authors tested in the involvement of this metabolite in possible regulation of the inflammation via the inflammasome pathway. [REDACTED]. Overall this is an well done study that provides new insights on the clinical relevance of SGLt2 mediated ketogenesis on inflammasome activation.
Following points should be addressed - Response: Thank you for the valuable comments. Unfortunately, we did not have residual samples of myeloid cells from the trial to explore the cleavage of caspase-1. Thus, we obtained PBMCs from diabetic donors and examined caspase-1 activation using western blots that detect the enzymatically active p20 subunit of caspase-1. It was really difficult to demonstrate biochemical cleavage of caspase-1 using human samples. As a result, BHB (20 mM) inhibited both the ATP-induced cleavage of caspase-1 into p20 and the processing of the biologically active form of IL-1β, whereas treatment in vitro with SGLT2 inhibitor empagliflozin (10 μM) did not inhibit caspase-1 activation (as shown in the figure below).
Given that the maximum plasma concentration of empagliflozin is less than 1 μM when treated with 25 mg empagliflozin once daily in patients with type 2 diabetes 1 , it seems that empagliflozin does not directly inhibit NLRP3 activation. Isolation of PBMCs from whole blood and differentiation into macrophages were performed as described in the main manuscript. Immunoblots from PBMCs untreated (Unt) or treated with BHB (10 or 20 mM), empagliflozin (10 uM) and LPS (0.5 ug/mL) for 3 h, followed by treatment with ATP (3 mM) for 1 h, as indicated. Cell were lysed in buffer containing 20 mM HEPES (pH7.5), 0.5 % Nonidet P-40, 50 mM KCl, 150 mM NaCl, 1.5 mM MgCl2, 1 mM EGTA, and protease inhibitors. Samples were then centrifuged at 20,000 g for 12 min, the supernatant was collected and the protein contents were measured using the Bradford assay (Bio-Rad, 161-0158). Equivalent amounts of each protein extract were heat-denatured in 5 x sample buffer (0.3 M Tris-Cl (pH6.8), 500 mM DTT, 10 % SDS, 50 % Glycerol, 0.05 % Bromophenol blue). Cell culture supernatants were precipitated by the addition of an equal volume of methanol and 0.25 volumes of chloroform, then were vortexed and centrifuged for 10 min at 20,000 g. The upper phase was discarded and 500 ul methanol was added to the interphase. This mixture was centrifuged for 10 min at 20,000 g and the protein pellet was dried at 55 °C, resuspended in 2 x sample buffer and boiled for 8 min at 98 °C. Samples separated on 10% polyacrylamide gels, and electrophoretically transferred onto a polyvinylidenefluoride membrane (Bio-Rad, 162-0177). After blocking, membranes were treated with the following antibodies: anti-IL-1β (Cell Signaling Technology, 12703S) and anti-caspase-1 (Cell Signaling Technology, 3866S).
Immunostaining was performed using chemiluminescent reagents (Clarity Max Western ECL Substrate; Biorad, 170-5062) and Agfa medical X-ray film (CP-G plus). Actin protein levels were used as a loading control.
Taken together, SGLT2 inhibitor might inhibit caspase-1 and NLRP3 activation via several metabolic changes including BHB, rather than having an inhibitory effect in itself.

Comment 2:
The prior papers have shown that BHB also inhibits the IL1B secretion in human monocytes and neutrophils. Please discuss the mechanism of effects of SGLT2 inhibitors and  levels might lead to a failure to detect differences in serum IL-1β levels before and after drug treatment.
Second, extensive use of aspirin or statin (more than 90 %) in both sulfonylurea and SGLT2 inhibitor groups might attenuate changes in serum proinflammatory cytokines. It is well known that aspirin or statin therapy can reduce proinflammatory cytokines and CRP 7-9 . [REDACTED].
Finally, IL-1β signaling in the immune cells of vessel tissues or of atherosclerotic lesions enriched for various endogenous danger signals, rather than circulating IL-1β levels, might be a major determinant of atheroma formation or heart failure 5,10,11 . IL-1β exerts its effect in an autocrine/paracrine fashion and, consequently, the detected levels of IL-1β in the plasma may be very low 12 . Furthermore, although most circulating cytokines are secreted from activated macrophages and lymphocytes, adipocytes and skeletal muscle are also a possible source of cytokines 13,14 . In addition, activation of the NLRP3 inflammasome requires two independent steps: priming and triggering 15 . Therefore, cytokine production capacity may not be related to serum concentration. IL-1β serum concentration probably represents only an indirect marker of chronic inflammation and a fraction of the real production capacity of activated monocytes/macrophages 16  It is stated that some subjects were already taking Metformin and were allowed to continue this medication throughout the trial. However, it is not partivcularly clear who these subjects were and whether these subjects were used in any of the assays in which only a subset or participants were used (importantly RNAseq). This is critical given that Metformin likely has NLRP3-altering capabilities due to its AMPK-stimulating actions.
Response: The proportion of participants who were administered metformin during the study period was as follows (shown in Table 1  As the use of metformin is first-line therapy, most participants had already been administered metformin and were included in all analyses of the present study. Since in each participant, preexisting metformin was maintained before and throughout the study period in dosage regimen, additional subgroup analyses according to metformin use were not performed.

Among 7 subjects of RNA seq analysis, 3 of 3 in sulfonylurea group and 3 of 4 in SGLT2
inhibitor group were treated with metformin before and throughout the study period.
The manuscript has been revised to include: "Baseline characteristics of study participants

.... As metformin is first-line therapy, most participants in both groups were taking it
and continued on the same dosage during the study period. ..."

(Methods)
"RNA extraction and sequencing Total RNA was isolated from macrophage lysates obtained from PBMCs of 7 individuals (3 from the sulfonylurea arm, 4 from the SGLT2 inhibitor arm) before and after treatment using a Qiagen RNA extraction kit (Qiagen, Valencia, CA, USA).

Among 7 individuals, 3 of 3 in sulfonylurea group and 3 of 4 in SGLT2 inhibitor group
were treated with metformin before and throughout the study period. Total RNA quality and quantity was verified on ..." The manuscript has been revised to include:
However, SGLT2 inhibitor showed a greater reduction in IL-1β secretion compared to sulfonylurea (time × group interaction P = 0.002), which remained significant after adjustment for body weight change (time × group interaction P = 0.02)." ".... To rule out the effect of body weight loss on changes in inflammasome activity, correlation analyses were conducted. As a result, there was no significant correlation between changes in body weight and changes in IL-1β or TNF-α release ( Supplementary Fig. 2)."

Reviewer #3 (Remarks to the Author):
In their paper «SGLT2 inhibition modulates NLRP3 inflammasome acitvitx in patients with This design and the hypothesis that interleukin-1beta might be involved is novel, interesting and clinically potentially important. The measured multiple parameters in plasma oft he patients and in monocytes obtained form the recruited subjects.
They found that both drugs similarly lowered plasma glucose, but empagliflocin led to a greater reduction of interleukin-1 beta.
Empagliflocin, but not glimepiride led to a marked increase in plasma levels of fasting betahydroxy butyrate and a significant decrease in urinc acid and fatsing insulin, whiel free fatty acids increased. Furthermore, empagliflocin improved insulin sensitivity, while glimepirides szimulated insulin secretion as expected.
This study is novel, interesting and downsides, howeverwell designed and written and illustrated with appropriate figures summarizing the data.
There are a few downsides, however: [redacted] Consisted with these findings, we demonstrated that both ex vivo experiments [REDACTED] that higher BHB levels and lower levels of insulin and glucose (metabolic conditions after the treatment of empagliflozin) significantly suppressed the activation of the NLRP3 inflammasome ( Fig. 4 and 5 in the main manuscript).
As the reviewer commented, it is also established that uric acid crystals activate the NLRP3 inflammasome. SGLT2 inhibitors have been shown to reduce serum uric acid levels through increasing renal clearance of uric acid 19 .
We performed additional correlation analyses of the changes in serum uric acid, insulin, and BHB levels and changes in IL-1β release in all participants (N=61).

Fig. Rev. 2. Correlation between changes in serum uric acid (a), insulin (b), or BHB (c) levels and changes in IL-1β release (n = 61).
Pearson's or Spearman's correlation coefficient. Some variables were log transformed for analysis. Therefore, although SGLT2 inhibitor caused a significant decrease in serum uric acid in the present study (Fig. 2c in the main manuscript), whether this reduction directly contributed to the suppression of NLRP3 inflammasome is not conclusive.

Comment 2:
In the ex vivo experiments NLRP3 only tended to be reduced by empagliflocin, while interleucin-1beat was significantly reducedis there thus another mechanisms involved rather than through NLRP3? Although RNA sequencing provided insights, this is not properly follwoed up.
Response: Thank you for the valuable comments. As you mentioned, the mRNA and protein levels of NLRP3 were not significantly decreased by SGLT2 inhibitor treatment (Fig. 3g,  [redacted]  In the present study, we concluded that SGLT2 inhibitor treatment suppresses not NLRP3 inflammasome itself, but the activation of the NLRP3 inflammasome, via combined metabolic effects of increased serum BHB levels and decreased serum levels of glucose and insulin. In a previous paper 17  regardless of NLRP3 inflammasome expression, suggesting that BHB also exhibit an inhibitory effect on constitutive NLRP3 inflammasome activation. [redacted] As the reviewer mentioned regarding the mechanism, additional text has been added in Results, as follows: [redacted] Therefore, SGLT2 inhibitor treatment might suppress signaling pathway of NLRP3 inflammasome activation, rather than reduce NLRP3 itself.  We agree with the reviewer's concern about RNA seq data. Our GO data confirmed that patients treated with SGLT2 inhibitor had lower expression of immune process-related genes compared to those treated with sulfonylurea. However, because of the small number of samples available for RNA seq and the heterogeneity of study population due to the nature of human trial, there is a limitation to conduct in-depth analysis of upstream inflammasome-mediated IL-1β secretion). Activation of the NLRP3 inflammasome (= secretion of IL-1β) generally requires two signals: the priming (by LPS in our study) and the triggering (by ATP or palmitate in our study). In previous studies, DAMPs-induced IL-1β secretion was totally dependent on the NLRP3 (as shown in the figures below). 24,25 [redacted] [redacted] Therefore, NLRP3-silenced cells cannot induce IL-1β secretion under signal1/2 stimulation. It is impossible to test whether empagliflozin treatment will further reduce IL-1β secretion from this condition (NLRP3 silenced cells). Although IL-1β might be secreted by other activated inflammasomes (e.g. AIM-2) in macrophages, those inflammasome signalings are not our questions to be solved and not closely related to metabolic diseases and ketones.

Comment 4:
Although beta-hydroxybutyrate has been described to suppress NLRP3 in mice, what is the mechanisms in human macrophages?
Response: In a paper published in Nature Medicine 17 , BHB dose-dependently inhibited IL-1β secretion not only in mice but also in human monocytes (as a figure below), and which was reproducible in our study (Fig. 4a in the main manuscript). [redacted] As mentioned above (response to comment 2), the Nature Medicine's paper showed that BHB blocks NLRP3 inflammasome activation both by controlling an unknown upstream event that reduces K + efflux from macrophages and by inhibiting ASC polymerization, speck formation and assembly of the inflammasome. Although these mechanisms have been elucidated through experiments mainly on mice models, the mechanisms might be same in human macrophages as well, since humans and mice belong to mammals and share similar machinery related to NLRP3 inflammasome.