Dietary Nitrate and Corresponding Gut Microbiota Prevent Cardiac Dysfunction in Obese Mice

The development of diabetic cardiomyopathy induced by obesity and over-nutrition (i.e., high-fat diet [HFD] feeding) is multifactorial involving dysregulation in numerous tissues. Hypertrophy, inflammation, and oxidative stress within adipose tissue results in systemic lipid oversupply, cytokine release, and ectopic lipid accumulation in peripheral tissues, events associated with insulin resistance and hyperglycemia. These impairments in adipose homeostasis can influence cardiac metabolism, as cardiac lipid oversupply is known to induce myocardial hypertrophy and fibrosis and impair contractility and, as a result, decrease cardiac function. Recently, the gut microbiome has also been implicated as a factor modulated by obesity and insulin resistance and capable of influencing cardiac function (1). Given the whole-body nature of diabetic cardiomyopathy, considerable interest has been placed on understanding the efficacy and cellular mechanisms of potential therapeutic approaches, particularly those that can influence several aspects of cardiovascular-related risk.

Dietary nitrate represents one such potential compound, as nitrate consumption has long been known to decrease blood pressure (BP) (1,2), protect against myocardial ischemia-reperfusion injury (3), and improve calcium (Ca²⁺) handling and cardiac contractility in healthy mice (4). Moreover, nitrate has also been shown to attenuate myocardial fibrosis associated with diastolic dysfunction in hypertensive rats (2) and improve several aspects of whole-body metabolism (adipose inflammation, hepatic lipid accumulation, redox imbalance, glucose intolerance) in HFD-induced mouse models of obesity (5–9). These ubiquitous effects of nitrate are thought to arise because of serial reduction of nitrate by bacteria in the oral cavity and intestines, producing the potent signaling molecule and vasodilator nitric oxide (NO) (10). Commensal bacteria are required for nitrate metabolism, which is supported by findings that nitrate consumption does not alter glucose tolerance or BP in germ-free mice (10). In addition to needing bacteria for the metabolism of nitrate, nitrate also may be capable of altering both the oral (11) and gut microbiomes (12,13). Given the influence of gut dysbiosis on cardiovascular function and the potential of dietary nitrate to alter the gut microbiome, the cardioprotective effects of nitrate could be due to either bioactive effects of NO itself or secondary to changes in microbial populations. Therefore, we examined whether nitrate could prevent HFD-induced cardiac abnormalities and whether this was a result of direct (dietary) or indirect (microbial) effects of nitrate. To do so, we first established the influence of dietary nitrate consumption on functional and cellular outcomes following 8 weeks of HFD-feeding in mice. We then transplanted fecal material from HFD or HFD+Nitrate donors to separate mice consuming the HFD to determine whether the beneficial effects of nitrate persisted in a fecal microbial transplantation (FMT) model.

Male C57Bl/6N mice were bred in-house and consumed a low-fat diet (LFD, 10% energy from fat; D12450J, Research Diets), HFD (60% energy from fat; D12492N), or HFD with dietary nitrate (HFD+Nitrate) (4 mmol/L NaNO₃ via drinking water) for 8 weeks. Average age following the dietary intervention was 25.7 ± 1.5 weeks LFD, 25.5 ± 2.0 weeks HFD, and 24.3 ± 2.1 weeks HFD+Nitrate (P = 0.85). A subset of mice were fed the identical HFD for 6 weeks, and FMT was performed via oral gavage on three separate days (days 1, 3, 5) (14,15) over the duration of week 6 using a fecal slurry (16) from original LFD, HFD, or HFD+Nitrate mice. Mice continually consumed the HFD prior to whole-body characterization and tissue collection in week 8 (all mice were 25 weeks of age). All animals were given access to ad libitum food and water and were kept in a 22°C temperature-controlled environment with a 12–12 h light-dark cycle. All animals were anesthetized using 2% isoflurane. The experimental procedures were approved by the University of Guelph Animal Care Committee.

Echocardiography of the left ventricle (LV) was performed on anesthetized (isoflurane) mice using Vevo 2100 imaging systems (VisualSonics) (17,18). B-mode and M-mode images of the LV were taken at the same time from the parasternal long axis of the heart (17). Doppler of the pulmonary trunk was performed from the parasternal short-axis view (19).

Mice were anesthetized with isoflurane, and body temperature was maintained at 37.5°C. An incision was made along the anterior region of the neck and a 1.2F pressure catheter (FTS-1211B-0018; Transonic Systems) was inserted into the LV via the right common carotid artery. Hemodynamic signals were digitized at a sampling rate of 2 kHz and analyzed using Spike2 v10 software (CED Spike2).

Mice were fasted 4–5 h and an intraperitoneal glucose tolerance test was performed with 2 g glucose/kg body wt (7). The area under the curve (AUC) was calculated after subtracting the baseline value (7).

Histology of epididymal white adipose tissue (eWAT) (20), LV (18), and intestinal tissues (duodenum and cecum) (21,22) was performed as previously described. Further information is in the Supplementary Methods.

Mitochondrial respiration (ADP, malonyl-CoA [M-CoA], l-carnitine titrations) and reactive oxygen species (ROS; succinate, ADP) emissions experiments were performed in LV permeabilized muscle fibers (17). Mitochondrial respiration was performed in eWAT and inguinal WAT (iWAT), and mitochondrial ROS experiments were performed in eWAT (7). Further information is in Supplementary Methods.

Fecal pellets from LFD, HFD, and HFD+Nitrate mice were homogenized using a glass mortar and pestle in sterile PBS (supplemented with 0.05% l-cysteine–HCl), at a ratio of 1 mL per fecal pellet, as previously described (14,16). Fecal pellets from each group were pooled together and passed through a cheese cloth strainer. Slurries were stored in aliquots at −80°C until the day of gavage. FMT was administrated via orogastric gavage (200 µL slurry per mouse) with a curved rounded needle on 3 separate days in the fed state.

Blood was collected via cardiac puncture and centrifuged at 2,300g for 10 min at 4°C. Serum and fecal slurries were analyzed for nitrate+nitrite (NOx) concentration fluorometrically using a commercially available kit (Cayman Chemicals) (17). Serum samples were analyzed for nonesterified fatty acid (NEFA) and triglycerides (TAG) using commercially available kits (Wako Diagnostics).

Fecal DNA was extracted using a commercially available kit (Zymo Quick-DNA Fecal/Soil Microbe Miniprep Kit). Thereafter, sample library preparation, normalization, and Illumina MiSeq sequencing were conducted by the Advanced Analysis Centre (University of Guelph), as previously described (23,24), and fecal metabolites were analyzed (25). Further information is in Supplementary Methods.

Lipid species were extracted and measured in LV, muscle, and eWAT tissue (26–28). Further information can be found in the Supplementary Methods.

LV and eWAT were homogenized in lysis buffer (17,20), diluted to 1 µg/µL, and loaded equally into standard SDS-PAGE gels. Target proteins are listed in Supplementary Table 1. Western blots were quantified using FluorChem HD2 imaging (Alpha Innotech).

Statistical analyses were completed using GraphPad Prism 9 software (GraphPad Software). One-way ANOVA was used between LFD, HFD, and HFD+Nitrate with Tukey multiple comparison post hoc analysis. Fecal metabolites and LV lipids were compared between HFD and HFD+Nitrate using unpaired two-tailed Student t tests. Unpaired two-tailed Student t tests were used to compare HFD-HFD FMT and HFD-HFD+Nitrate FMT, or HFD-HFD FMT and HFD-LFD FMT. ADP and l-carnitine titrations were analyzed using nonconstrained Michaelis-Menten kinetics. M-CoA inhibition was determined using one-phase decay analysis. Statistical significance was determined as P < 0.05. Data are expressed as mean ± SEM and depicted as bar and scatter plots. Appropriate statistical analysis details are listed in the respective figure legends.

The data generated during the current study are available from the corresponding author upon reasonable request.

Eight weeks of HFD-feeding increased body weight (Supplementary Fig. 1A), heart weight/tibia length (Fig. 1A and B), serum NEFA and TAG (Fig. 1C), and impaired glucose tolerance (Supplementary Fig. 1B and C). In stark contrast, dietary nitrate attenuated the HFD-induced increases in serum lipids (Fig. 1C) and glucose intolerance (Supplementary Fig. 1B and C), in concert with a fivefold increase in serum NOx (Supplementary Fig. 1D). Caloric intake was increased in HFD-fed mice regardless of nitrate consumption, and water intake did not differ (Supplementary Fig. 1E–H). Moreover, while HFD-feeding decreased end-diastolic volume (EDV) (Fig. 1D, E, and F), stroke volume (SV) (Fig. 1G), and cardiac output (Fig. 1H) without affecting end-systolic volume (ESV) (Fig. 1I) or heart rate (Supplementary Fig. 1E), dietary nitrate completely prevented these deleterious changes (Fig. 1D–H). Similarly, nitrate prevented the HFD-induced impairments in right ventricle (RV) SV and cardiac output (Supplementary Fig. 1I and J). These functional outcomes appeared linked to anatomical changes since dietary nitrate increased pulmonary trunk diameter (Supplementary Fig. 1L) but not the LV outflow tract velocity time integral (Supplementary Fig. 1M). Dietary nitrate also prevented the increase in posterior wall thickness (Fig. 1J) and LV fibrosis (Fig. 1E and K) occurring with HFD, which combined with the preservation of heart mass (Fig. 1A and B) and end-diastolic diameter (Supplementary Fig. 1N) suggests dietary nitrate partially attenuated HFD-induced concentric hypertrophy. Extracellular signal–regulated kinase (ERK) phosphorylation has been linked to cardiac hypertrophy, and phosphorylation of ERK1 was increased only in HFD mice (Supplementary Fig. 2A and B). In addition, HFD-feeding tended to increase ERK2 phosphorylation compared with LFD mice (t test P = 0.058), while dietary nitrate prevented this response (Supplementary Fig. 2A and B). However, there were no differences in ejection fraction (63.7 ± 1.0% LFD vs. 64.0 ± 1.1% HFD vs. 69.3 ± 0.6% HFD+Nitrate; P = 0.35) or fractional shortening (34.8 ± 0.4% LFD vs. 32.0 ± 0.5% HFD vs. 35.8 ± 0.4% HFD+Nitrate; P = 0.65). Aortic systolic BP (SBP), diastolic BP (DBP), and mean arterial pressure (MAP) were increased in HFD-fed mice regardless of nitrate consumption (Fig. 2A and B). Peak LV pressure (LVP) (Fig. 2C), and end-diastolic pressure (EDP) (Fig. 2D) were increased following the HFD. Nitrate consumption attenuated the rise in EDP (Fig. 2D), suggesting preserved cardiac physiology. There were no differences in rate of change of LVP (dP/dt max, dP/dt min, or dP/dt at LVP of 40 mmHg) (Supplementary Fig. 1P–R). Myocardial lipids have been associated with impaired cardiac function (29), and nitrate supplementation decreased lipid content of various ceramide (Fig. 2E), diglyceride (DAG) (Fig. 2F), and TAG (Fig. 2G) species in the LV. Altogether, nitrate supplementation attenuated the phenotypic changes in cardiac structure and function caused by HFD-feeding.

As dietary nitrate consumption prevented HFD-induced cardiac dysfunction, we aimed to examine possible underlying mechanisms within the LV. Ca²⁺ handling and the regulation of sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA) are important for maintaining cardiac contractility; however, there were no differences in protein content of SERCA2 or calsequestrin 2 (CSQ2) between groups (Fig. 3A and B). While SERCA activity is highly regulated by phospholamban (PLN), the HFD-mediated changes in total (t) and phosphorylated (p) PLN were similar in the nitrate group (Fig. 3A–C), and therefore likely do not contribute to the improved cardiac function observed with dietary nitrate.

As an alternative target, mitochondria are central to energy metabolism. However, mitochondrial protein content (Fig. 3D and Supplementary Fig. 2C), maximal respiratory capacity (Fig. 3E), and mitochondrial ADP sensitivity (Fig. 3F) were not altered by the HFD or nitrate. HFD+Nitrate increased mitochondrial carnitine palmityl transferase I (CPT-I) sensitivity to l-carnitine (Fig. 3G), and maximal CPT-I–supported respiration was increased with HFD (palmitoyl-CoA + l-carnitine) (Fig. 3H), representing compensatory responses likely linked to cellular lipid overload associated with diabetic cardiomyopathy (30), as opposed to mechanisms reducing diastolic volume. There were no differences in the sensitivity of CPT-I to the inhibitory effects of M-CoA in either group (Fig. 3H and Supplementary Fig. 2D).

ROS has been linked to cardiac dysfunction (31,32) and shown to mediate an increase in ERK phosphorylation (33), indicating a signal that could connect LV mitochondrial function to fibrosis and hypertrophy. We observed that HFD consumption increased maximal succinate-supported mitochondrial ROS ∼50% (Fig. 3I), an effect that was completely mitigated by nitrate supplementation. Submaximal mitochondrial ROS (+100 μmol/L ADP) was also increased nearly threefold in the LV of HFD-fed mice compared with LFD and HFD+Nitrate mice (Fig. 3I). As a result, the ability of ADP to suppress mitochondrial ROS was impaired with the HFD and completely preserved with nitrate consumption. While catalase content was increased in the HFD+Nitrate mice, 4-hydroxynonenal (4HNE), a marker of lipid peroxidation, was not different between animals, indicating the absence of overt redox stress (Supplementary Fig. 2C and E).

We next investigated the possibility that nitrate supplementation altered the composition of the gut microbiome as an initiating event in the cardioprotective nature of dietary nitrate. At the phylum level, the relative abundance of Firmicutes and the ratio of Firmicutes to Bacteroidetes were increased following HFD-feeding (Fig. 4A–D), indicating a classic compositional change in the fecal microbiome with obesity (34). In contrast, dietary nitrate supplementation during the HFD prevented these shifts (Fig. 4B and C). There were no overt changes in fecal short-chain fatty acids (SCFA), while there were trends for nitrate to increase fecal formate and lactate concentrations compared with mice fed only the HFD (Fig. 4E). As whole-body inflammation is associated with intestinal permeability (35), we examined structural changes of the intestinal tract. As expected, the HFD decreased the number of goblet cells; however, we did not find differences in villus length, crypt depth, or goblet cell number within the duodenum and cecum between HFD and HFD+Nitrate mice (Supplementary Fig. 3).

In a separate subset of animals, 6 weeks of HFD-feeding did not alter ESV, EDV, SV, cardiac output, or LV ejection fraction (Supplementary Fig. 4) compared with LFD mice. We therefore chose this time point to administer FMT to HFD mice from HFD donors or HFD+Nitrate donors (Fig. 5) to examine a potential preventative effect of HFD+Nitrate FMT. At 6 weeks of HFD-feeding (prior to FMT), glucose tolerance was similar between groups (Supplementary Fig. 5A and B), and body mass was comparable at all time points (Supplementary Fig. 5C). As NOx concentrations were not increased in the feces of HFD+Nitrate mice (Fig. 5) or serum of HFD-HFD+Nitrate FMT mice (Fig. 6A), we confirm the absence of nitrate in FMT animals. Despite this, 9 days after performing FMT, HFD-HFD+Nitrate FMT animals presented with a strong trend toward lower serum TAG (P = 0.057) and a significant reduction in serum NEFA compared with HFD-HFD FMT (Fig. 6B). In addition, glucose intolerance was prevented in HFD-HFD+Nitrate FMT mice compared with HFD-HFD FMT mice (Fig. 6C and D), which combined, represent metabolic changes that may prove beneficial for metabolic syndrome and cardiac function.

HFD+Nitrate FMT did not alter the content of mitochondrial or redox proteins (Fig. 6E and Supplementary Fig. 5D), mitochondrial respiratory function (Fig. 6F and G), or Ca²⁺-handling proteins (Fig. 6H and Supplementary Fig. 5E and F). HFD+Nitrate FMT decreased maximal succinate-supported mitochondrial ROS emission ∼40% (Fig. 6I), in the absence of changing submaximal ROS (Fig. 6I) or ADP suppression of ROS (Fig. 6J). While LV fibrosis was not altered (Fig. 7A and B), EDV (Fig. 7C), SV (Fig. 7D), and cardiac output (Fig. 7E and F) were all ∼35% higher in HFD-HFD+Nitrate FMT mice. In addition, there were trends for ejection fraction (P = 0.10) and fractional shortening (P = 0.08) to be higher in HFD-HFD+Nitrate FMT mice (Supplementary Table 2). Similar to dietary nitrate, this occurred in the absence of changes in ESV (18.1 ± 3.9 µL HFD-HFD FMT vs. 22.0 ± 4.0 µL HFD-HFD+Nitrate FMT, P = 0.51). The changes in cardiac function with HFD+Nitrate FMT were not due to any differences in aortic SBP (Fig. 7G and H) or DBP (Fig. 7G and I) between groups. In addition, end-systolic LVP, dP/dt max, dP/dt min, and heart rate were not altered by HFD+Nitrate FMT (Supplementary Fig. 5G and H and Supplementary Table 2). While there were no differences in heart weight between HFD-HFD FMT and HFD-HFD+Nitrate FMT mice (Fig. 7J), LV posterior wall thickness was lower with HFD+Nitrate FMT (Fig. 7K), indicating prevention of hypertrophy markers.

To further understand whether these effects were due to the gut microbiome, we performed FMT in a separate group of HFD-fed animals using HFD feces or LFD feces, since the fecal microbiome composition was similar between LFD and HFD+Nitrate mice (Fig. 4A–C). We report that echocardiography parameters were similar between groups prior to FMT (Supplementary Fig. 6). However, FMT from LFD donors prevented the HFD-induced decrease in cardiac morphology (SV, cardiac output, ejection fraction, fractional shortening) and the HFD-induced increase in posterior wall thickness (Supplementary Fig. 7A–H). In addition, a trend toward greater glucose tolerance was evident in HFD-LFD FMT compared with HFD-HFD FMT mice (Supplementary Fig. 7I and J), responses similar to HFD-HFD+Nitrate FMT mice, suggesting the beneficial effects are microbiome linked.

Given the improvement in serum lipids we observed with both dietary nitrate and HFD+Nitrate FMT, we aimed to determine whether adipose tissue was associated with the cardioprotective outcomes in response to nitrate. In HFD-consuming animals, oral nitrate supplementation did not alter lipid content in skeletal muscle, or DAG and TAG content in eWAT; however, nitrate decreased ceramide species within eWAT (Supplementary Fig. 8D). While dietary nitrate did not alter adipocyte hypertrophy (cross-sectional area) (Fig. 8A and B), oral nitrate consumption blunted HFD-induced leukocyte infiltration measured by crown-like structures within eWAT (Fig. 8C), in line with our findings of reduced ceramide species associated with lower inflammation. In addition, we previously reported in the same animals that nitrate consumption prevented HFD-induced increases in eWAT mitochondrial ROS, c-Jun N-terminal kinases (JNK) phosphorylation, and 4HNE (7), indicating improvements in adipose homeostasis. However, HFD+Nitrate FMT did not alter eWAT hypertrophy (Fig. 8D and E), leukocyte infiltration (Fig. 8F), and inflammatory (pJNK, pERK) or oxidative stress (4HNE) protein markers (Fig. 8G and H). In addition, there were no differences in lipid-supported ROS emission in eWAT following HFD+Nitrate FMT (Fig. 8I), while respiration was higher in both eWAT (Fig. 8J) and iWAT (Fig. 8K) of HFD-HFD+Nitrate FMT mice. This indicates that ingestion of dietary nitrate can improve markers of adipose inflammation in HFD-fed mice, but FMT from HFD+Nitrate donors did not, suggesting changes in adipose homeostasis are not required for the cardioprotective effects of dietary nitrate.

In mice with diet-induced obesity, we report that dietary nitrate attenuated LV mitochondrial ROS and pathological cardiac remodelling, in addition to improving adipose tissue inflammation, lipid homeostasis, and glucose tolerance. Oral nitrate feeding also mitigated compositional changes in the microbiome during the HFD. However, these effects do not appear to be entirely dependent on increasing serum nitrate, as transfer of fecal material via FMT (from both HFD+Nitrate and LFD donors) was sufficient to prevent impairments in cardiac morphology and function. As a result, our findings suggest the cardioprotective effects of nitrate may be linked to changes in the gut microbiome and can occur without an increase in serum nitrate concentrations.

The consumption of dietary nitrate can elicit many effects on whole-body metabolism, such as improvements in glucose tolerance, adipose tissue redox balance (6,7,9), BP (5), and liver steatosis (5) in mouse models of HFD-induced obesity. In addition, dietary nitrate is associated with changes in intestinal microbial richness in models of dysbiosis (12) and has been shown to influence the oral microbiome (11). This nitrate-microbiome interaction is bidirectional, as nitrate-mediated improvements in glucose tolerance and BP are absent in germ-free mice lacking commensal bacteria (10), which is related to the obligatory role of bacteria in the metabolism of nitrate to NO (36). Here, we also report that nitrate prevented the characteristic change in the fecal Firmicutes to Bacteroidetes ratio associated with the HFD. The gut microbiome represents a nexus between inflammation, lipid homeostasis, and cardiovascular function. While we did not report differences in SCFA, microbial metabolites during obesity have been associated with cardiac hypertrophy, fibrosis, and hypertension (reviewed in ref. 37), in addition to the activation of inflammatory and immune pathways (35), adipose proliferation, and macrophage infiltration (38). However, it is possible that metabolites beyond SCFA, such as trimethylamine-N-oxide and tryptophan-derived metabolites, could be mediating this relationship following nitrate feeding and FMT. Regardless, there is evidence to suggest the gut-heart axis may represent a viable therapeutic target for improving cardiometabolic health.

Proof-of-principle experiments have clearly shown that manipulating the gut microbiome can influence host metabolism, and FMT from resveratrol-fed mice was shown to improve glycemia, decrease inflammatory markers, and lower BP in mice fed an HFD for 8 weeks (14,15). These effects were similar to that of dietary resveratrol, suggesting a microbial mechanism of action for the beneficial effects of resveratrol in HFD-induced obesity. Similarly, we determined that FMT from HFD+Nitrate donors prevented the diastolic dysfunction, glucose intolerance, and serum dyslipidemia occurring in HFD-HFD FMT mice. While we did not report any influence of HFD+Nitrate FMT on WAT inflammation, our measured indices (leukocyte infiltration, pJNK, and pERK) may not be sensitive enough to detect a change 9 days following FMT, particularly as leukocyte infiltration is a prolonged structural change (20). Regardless, serum NEFA and TAG were lower in HFD-HFD+Nitrate FMT mice. This reduction in systemic lipids could be beneficial for heart function as the accumulation of cardiac reactive lipids has been shown to cause lipotoxic cardiomyopathy (29), and we report that oral nitrate attenuated the HFD-induced increase in LV lipid species.

Within the LV, impairments in mitochondrial respiration have been associated with cardiomyopathy in humans (31) and rodents (39) with diabetes. In addition, supplementation with mitochondrial-targeted antioxidant coenzyme Q (40) and overexpression of the cardiac mitochondrial antioxidant manganese-dependent superoxide dismutase (41) attenuate cardiac dysfunction and hypertrophy in diabetic mice. While we did not detect impairments in mitochondrial respiration in any condition, both dietary nitrate and HFD+Nitrate FMT attenuated maximal LV mitochondrial ROS emission. Dietary nitrate consumption has also been shown to blunt maximal LV ROS production in isolated mitochondria of doxorubicin-induced models of cardiotoxicity (42). However, whether this occurs secondary to reductions in serum NEFA remains unknown, as lipids are known to be particularly susceptible to the production of mitochondrial ROS, and several lipid species were decreased in the heart of HFD+Nitrate mice. Alternatively, posttranslational modifications induced by hyperglycemia have been shown to increase ROS in isolated cardiomyocytes (43), high glucose has also been shown to promote ERK-mediated cardiomyocyte apoptosis (44), and hyperglycemia independently increases the risk of cardiac dysfunction (45). This may indicate a mechanism of action in which hyperglycemia increases intracellular stress-signaling events that result in pathological transcriptional programs and decreased diastolic function, all of which were attenuated by nitrate.

Oxidative stress is also an important event contributing to myocardial fibrosis. However, while myocardial fibrosis has classically been associated with heart failure, recent evidence challenges the necessity of this structural occurrence (46). Our findings that diastolic dysfunction was prevented in HFD-HFD+Nitrate FMT mice in the absence of overt fibrotic changes further suggests that myocardial fibrosis may not be required in the development of cardiomyopathy. However, mechanisms characterizing the cardiac phenotype and associated markers in response to nitrate supplementation (fibrotic changes, protein expression, cardiac hypertrophy, and cardiomyocyte changes) remain to be fully understood, which is a limitation of our current understanding. On a cellular level, HFD-induced increases in superoxide production may decrease NO bioavailability (O₂⁻ + NO → ONOO⁻), increasing titin stiffness through secondary phosphorylation events (reviewed in ref. 47) and contributing to the observed decrease in EDV. While the consumption of dietary nitrate could increase NO bioavailability to rectify this, serum from HFD-HFD+Nitrate FMT mice did not contain additional NOx and the half-life of NO is ∼2 ms (48), suggesting increased NO bioavailability is not a key mechanism of action for the FMT-mediated preservation of cardiac morphology. Regardless, as HFD+Nitrate FMT did not reproduce all the effects of dietary nitrate, some of the beneficial changes with oral nitrate feeding may be beyond the gut and dependent on increasing serum nitrate levels, NO-related signaling, or structural changes not captured in the 2-week FMT model. Additionally, a limitation of our work is that we only performed experiments in male mice due to the greater susceptibility to diet-induced obesity and its consequences. Sexual dimorphisms in cardiac metabolism exist (49); therefore, the effects of nitrate in females warrants further investigation.

Overall, we report that impairments in glucose tolerance, blood lipids, diastolic function, and LV mitochondrial ROS emission were prevented following both dietary nitrate consumption and HFD+Nitrate FMT. While adipose tissue dysfunction is linked to cardiac outcomes, markers of overt adipose inflammation and cardiac fibrosis were only improved following dietary nitrate consumption but not HFD+Nitrate FMT. It is therefore likely that nitrate acts through multiple mechanisms of action, which are not solely dependent on increasing nitrate concentrations to exert a cardioprotective effect. One such mechanism may be linked to the gut microbiome, particularly as FMT (either LFD or HFD+Nitrate) also prevented several cardiometabolic impairments. These findings provide important insight into the growing body of literature indicating the therapeutic potential of nitrate in states of obesity and insulin resistance as well as highlight the gut microbiome as a modifiable cardiometabolic risk factor.

Funding. This work was funded by the Natural Sciences and Engineering Research Council of Canada (400362 to G.P.H. and Graduate Student Scholarship to H.L.P.). J.D.S. is a Canada Research Chair in Metabolic Inflammation.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. H.L.P., L.M.O., H.S.B., A.R., A.J.K., P.-A.B., R.M.H., B.C.-A., C.G.-H., K.M.J.H.D., A.C., E.A.-V., J.A.S., and G.P.H. organized and performed experiments. H.L.P., L.M.O., H.S.B., A.R., A.J.K., P.-A.B., R.M.H., B.C.-A., L.J.C.v.L., A.C., J.D.S., E.A.-V., J.A.S., and G.P.H. analyzed and interpreted the data. H.L.P., H.S.B., J.D.S., and G.P.H. designed the study. H.L.P. and G.P.H. drafted the manuscript, and all authors edited and approved of the final version. G.P.H. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented orally at the 80th Scientific Sessions of the American Diabetes Association, virtual meeting, 12–16 June 2020, and the abstract was published online at Diabetes 2020;69(Suppl. 1):75-OR (https://doi.org/10.2337/db20-75-OR). Parts of this study were accepted for presentation at the 30th European Congress on Obesity, Dublin, Ireland, 17–20 May 2023.