Psyllium fiber protects mice against western diet-induced metabolic syndrome via the gut microbiota-dependent mechanism

ABSTRACT

Impacts of dietary fiber on intestinal inflammation are complex, but some specific semi-purified fibers, particularly psyllium, can protect humans and rodents against colitis. Mechanisms underlying such protection are not fully understood but may involve activation of the FXR bile acid receptor. Obesity and its associated consequences, referred to as metabolic syndrome, are associated with, and promoted by, low-grade inflammation in a variety of tissues including the intestine. Hence, we examined whether psyllium might ameliorate the low-grade intestinal inflammation that occurs in diet-induced obesity and, moreover, the extent to which it might ameliorate adiposity and/or dysglycemia in this disease model. We observed that enriching a high-fat diet with psyllium provided strong protection against the low-grade gut inflammation and metabolic consequences that were otherwise induced by the obesogenic diet. Such protection was fully maintained in FXR-deficient mice, indicating that distinct mechanisms mediate psyllium’s protection against colitis and metabolic syndrome. Nor did psyllium’s protection associate with, or require, fermentation or IL−22 production, both of which are key mediators of beneficial impacts of some other dietary fibers. Psyllium’s beneficial impacts were not evident in germfree mice but were observed in Altered Schaedler Flora mice, in which psyllium modestly altered relative and absolute abundance of the small number of taxa present in these gnotobiotic mice. Thus, psyllium protects mice against diet-induced obesity/metabolic syndrome by a mechanism independent of FXR and fermentation but nonetheless requires the presence of at least a minimal microbiota.

KEYWORDS:

• Obesity
• diabetes
• short-chain fatty acids

Introduction

Dietary fiber, the collective term for the chemically diverse array of non-easily digestible components of plant-based foods, is broadly health-promoting¹. The extent to which semi-purified fiber supplements can recapitulate such benefits is far from clear and thus remains an active area of research. Consumption is associated with reduced incidence of a variety of chronic inflammatory diseases, possibly reflecting the ability of dietary fiber to promote intestinal and metabolic health. Recent studies from us and others suggest particularly pronounced metabolic benefits from soluble fibers, which are readily fermentable by gut bacteria^2,3. However, while epidemiological studies suggest that fiber-rich foods can prevent development of inflammatory bowel disease (IBD), and persons afflicted with this disease frequently report that fiber-rich foods exacerbate their disease, perhaps reflecting that their dysbiotic microbiotas cannot metabolize them⁴. The potential of enriching diets with fiber, particularly fermentable fiber to promote gut inflammation, can be seen in the DSS colitis model wherein enriching diets with inulin or pectin exacerbates disease^5,6. This accords with observations that some IBD patients seek to avoid fiber-rich foods, especially those rich in fermentable fiber, thus depriving them its broad array of benefits. However, it is increasingly appreciated that irrespective of solubility and fermentability, specific fibers have a distinct impact on microbiota and, moreover, gut health. Most relevant to this study is that screening panels of semi-purified fibers in mouse models of colitis revealed the Plantago seed-derived semi-soluble fiber, and psyllium is unique in its ability to strongly suppress DSS and T-cell-mediated colitis^7,8. Such findings accord with a study that found psyllium helped maintain remission in ulcerative colitis patients⁹. Psyllium’s protection against experimental colitis is associated with its increasing serum bile acids (BA) and, furthermore, required the presence of the FXR bile acid receptor⁷. How psyllium elevates serum BA is not well understood but may involve its long-appreciated ability to sequester luminal BA, thus removing them from enterohepatic circulation. Such BA are replaced by synthesizing a new pool of BA from cholesterol, thereby lowering blood cholesterol levels, which, in addition to promoting bowel regularity, is a reported health benefit of psyllium consumption^10,11.

That metabolic syndrome can be viewed as a type of inflammatory disease, namely, one of low-grade inflammation, which led us to investigate the extent to which psyllium might confer protection against this disorder. We observed that psyllium’s beneficial metabolic impacts were not limited to lowering cholesterol. Rather, psyllium protected mice from the low-grade inflammation and array of obesity-associated metabolic abnormalities that otherwise resulted from consumption of a low-fiber high-fat diet. Such protection did not involve activation of FXR nor increased production of short-chain fatty acids (i.e. fermentation) but nonetheless required that at least a minimal microbiota be present.

Methods

Mice

C57BL/6 WT and FXR^−/− mice were purchased from Jackson Laboratory (Bar Harbor, ME). Gnotobiotic C57BL/6 germ-free mice (GF) were purchased from Taconic Biosciences Inc (Rensselaer, NY). Mice harboring Altered Schaedler Flora (ASF) were generated from these GF mice as previously described¹². All mice were bred, maintained, and experimentally studied at Georgia State University under an institutional animal care and use committee (IACUC protocol # A17047).

Diet-induced metabolic syndrome

All mice were maintained on standard grain-based chow (GBC, Purina 5001) until 6–8 weeks of age at which point they were further fed with this diet or a purified “open-source”, low-fiber high-fat diet (LF-HFD), LF-HFD enriched with indicated fiber, all ad libitum, for a 4-week study period. Precise diet compositions are indicated in Supplementary Table S1, but briefly, the LF-HFD is composed of 5% cellulose by weight, which is its only fiber source. The fiber-enriched LF-HFDs all contained 5% cellulose + indicated amount of additional fiber with the rationale that GBC is approximately 15–25% fiber by weight. Mice were monitored weekly for body weight. At day 28, mice were euthanized, and serum, colon length, washed colon weight, epididymal fat pad, cecum, and spleen weights were collected for downstream analysis.

Glucose and insulin tolerance test

Glucose tolerance (GTT) was measured 25 days following initiation of indicated diet. Mice were placed in a clean cage and provided with water but without food for 5 hours. Baseline blood glucose was then measured using a Nova Max plus Glucometer. Mice were intraperitoneally administered 2 mg of glucose/gm body weight, and blood glucose levels measured 30, 60, and 90 minutes later. Data are expressed as mg glucose/dL blood. 2 days later (i.e. day 27) to measure “fasting glucose levels”, mice were placed in a clean cage and provided with water but without food for 5 h, at which point, fasting glucose levels were determined and insulin tolerance testing (ITT) was initiated. Specifically, 5-h fasted mice were injected with 0.5 U insulin/kg body weight and blood glucose levels were measured at 30, 60, and 90 min after injection.

RNA extraction and real-time PCR

Total RNA was isolated from colon using TRIzol (Invitrogen, Carlsbad, CA); the expression level of IL−22 was analyzed by using quantitative real-time PCR according to the Biorad iScript One-Step RT-PCR Kit in a CFX96 apparatus (Bio-Rad, Hercules, CA) with the primers presented in Table S2. (F/R): IL−22: GTGCTCAACTTCACCCTGGA, TGGATGTTCTGGTCGTCACC; 36B4: TCCAGGCTTTGGGCATCA, CTTTATTCAGCTGCACATCACTCAGA. Differences in transcript levels were quantified by normalization of each amplicon to housekeeping gene 36B4.

Assessment of pro-inflammatory cytokines was performed with SYBR Green using the StepOnePlus PCR system (Applied Biosystem), and gene expression was normalized to Gapdh. qPCR primers are

Il6 F: 5’-GGCGGATCGGATGTTGTGAT−3’

Il6 R: 5’-GGACCCCAGACAATCGGTTG−3’

Tnf $\mathit{\alpha }$ F: 5’-CGAGTGACAAGCCTGTAGCC−3’

Tnf $\mathit{\alpha }$ R: 5’-CATGCCGTTGGCCAGGA−3’

Cxcl1 F: 5’-TTGTGCGAAAAGAAGTGCAG−3’

Cxcl1 R: 5’-TACAAACACAGCCTCCCACA−3’

Gapdh F: 5’-CAAATGGTGGAAGCACAGTT-GGCA−3’

Gapdh R: 5’-TTGTGTCCAGGTCCTCCATGA-TGT−3’

Quantification of the fecal bacterial load

To measure the total fecal bacterial load and the relative quantity of each Altered Schaedler Flora (ASF) strains, total DNA was isolated from the known amount of feces using the QIAamp DNA Stool MiniKit (Qiagen, Hilden, Germany). DNA was then subjected to qPCR using the QuantiFast SYBR Green PCR kit (Bio-Rad, Hercules, CA) with universal 16S rRNA primers to measure the total bacteria number and specific ASF primer (Table S2). Results are expressed as bacteria number per mg of stool using a standard curve for total bacteria and as relative quantity compared to total bacteria for the ASF strain.

Short-chain fatty acids (SCFA) quantification

Cecal content was collected immediately from non-fasted mice after euthanasia and then weighed before freezing. Levels of acetate, propionate, and butyrate were measured after extraction with ethyl acetate using Agilent 7890A gas chromatography (Santa Clara, CA) with a fused silica capillary column (Nukon SUPELCO No: 40369-03A, Bellefonte, PA) as described¹³. Heptanoic acid is used as the internal standard, and the results are presented as μmol/g or μg/g of cecal/fecal content.

Statistical analyses

Statistical significance of results was analyzed using GraphPad 9 software by analysis of ANOVA variance with a correction for the multiple analysis Dunnett test. When specified experiments were performed multiple times, each shows a similar pattern of results, which means that P values were similar between experiments (±1 log₁₀). Significance is expressed as *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Results

Most vivaria maintain mice on grain-based chow (GBC), which is a closed, proprietary formula composed of relatively unrefined, variable, and non-readily manipulable ingredients, thus stymying attempts to precisely and reproducibly manipulate the diet or report the diet composition. This limitation prompted development of “open-source” purified diets, which are composed of purified relatively invariant components, whose relative abundance can be formulated to investigate precise roles of specific macro- and micronutrients^14,15. With its variability notwithstanding, GBC is relatively low in fat (about 6% my weight) and rich in fiber (15–25% by weight), including soluble/fermentable fiber, which nourishes gut microbiota and promotes intestinal and metabolic health^3,16. In contrast, open-source control diets typically contain only 5% fiber and are thus considered low-fiber diets¹⁶. Furthermore, the fiber used is typically cellulose, which is insoluble and highly resistant to fermentation. Such lack of fermentable fiber causes stark changes to gut morphology that associate with modest increases in adiposity¹⁷, which becomes quite pronounced when such low-fiber diets are enriched in fats such as lard³. As an approach to study impacts of defined diets on metabolic syndrome, we purchased male 6- to 8-week-old C57 BL/6 mice from a large commercial supplier (Jackson Labs) that, like most rodent suppliers, maintains its mice on GBC. These mice were maintained in our vivaria on GBC for 5–10 additional days during which times mice were assigned to experimental groups (n = 5), which were then maintained on GBC or switched to an open-source low-fiber, high-fat diet (LF-HFD) or high-fat diets enriched with fiber as indicated in Table S1. At the time of diet switch (day 0), mean body mass was similar for all groups (range 20–23 gm) and with no statistically significant differences between any two groups of mice, irrespective of whether corrected for multiple comparisons (p-values for all individual comparisons > 0.1).

As expected, relative to mice maintained continuously on GBC, those subjected to a 4-week ad libitum LF-HFD feeding exhibited markedly increased adiposity as evidenced by increases in total body weight (Figure 1A) and the weight of their major (epididymal) fat pad (Figure 1B). Such increased adiposity was accompanied by dysglycemia as indicated by glucose and insulin tolerance testing (Figure 1C-D). Furthermore, as observed previously, LF-HFD-induced metabolic syndrome associated with loss of cecal/colon mass and colonic shortening, which is thought to reflect the low-grade inflammation that associates with and may promote characteristic of metabolic syndrome⁵ (Figure 1E-G). In further accord with previous studies, enriching LF-HFD with the fermentable fiber inulin ameliorated both the metabolic abnormalities and the indicator of low-grade inflammation that resulted from LF-HFD³. The semi-purified fibers pectin and cellulose also showed metabolic benefits not shared by Hi-Maize, which is a resistant corn starch metabolized by gut bacteria and thus considered a functional fiber, However, only psyllium shared inulin’s ability to both restore intestinal mass and metabolic health. The reduction in intestinal mass induced by LF-HFD is a previously observed microbiota-dependent consequence of this diet thought to reflect low-grade inflammation⁵. Low-grade inflammation can impair satiety signaling, leading to hyperphagia and promoting metabolic syndrome, all of which are ameliorated by inulin³. While the crumbly nature of high-fat diets makes measurements of food consumption challenging, our measurements of this parameter suggested that psyllium also ameliorated satiety signaling dysfunction in that it lowered caloric intake to a similar extent as inulin although differences between LF-HFD and either fiber-enriched diet were not statistically significant (daily mean ± SD caloric intake per mouse per day for GBC, inulin, and psyllium diets were 14.59 ± 4.46, 12.11 ± 2.50, and 13.03 ± 3.29, respectively).

Figure 1. Dietary psyllium provides substantial protection against HFD-induced metabolic syndrome. Male 6- to 8-week-old C57Bl/6 mice were fed with indicated diet for 28 days. (a) Relative body weight. (b) Epididymal fat pad weight. (c) Glucose tolerance test and respective areas under the curve. (d) Insulin tolerance test and respective areas under the curve. (e) Colon length. (f) Colon weight. (g) Cecum weight. (h) Spleen weight. Data are expressed as means ± SEM of n = 5 mice per group and are representative of three independent experiments. Significance was determined by the ANOVA variance test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Varying the amount of psyllium enrichment of LF-HFD from 2.5–17%, it is found that as little as 2.7% psyllium provides stark protection against DSS colitis, the extent of which does not increase significantly at higher concentrations of psyllium⁷. In contrast, psyllium’s protection against LF-HFD-induced adiposity appeared to be proportional to its concentration with significant protection requiring concentrations of at least 5.5%, while maximal protection was observed at the highest concentration tested (Figure 2). Restoration of glucose tolerance required LF-HFD is enriched with at least 11% psyllium although all concentrations of psyllium tested ameliorated insulin resistance as indicated by the insulin tolerance test (Figure 2C-D). All tested psyllium concentrations also significantly restored intestinal mass and, moreover, ameliorated the mild splenomegaly that was otherwise induced by LF-HFD (Figure 2E-H), suggesting that psyllium had ameliorated the low-grade inflammation induced by this diet. To more directly examine this notion, we measured psyllium’s impact on colonic expression of pro-inflammatory cytokines (TNF$\mathit{\alpha }$, IL6, and CXCL1) by qRT-PCR. In accord with other studies, LF-HFD led to moderate but significantly elevated expression of these cytokines relative to GBC-fed mice (Figure 2I). This increase in colonic proinflammatory gene expression was reduced, in a dose-dependent manner by psyllium, thus supporting the notion that psyllium’s ability had indeed reduced inflammation potentially accounting for some of its metabolic benefits although the approaches used here did not enable us to decipher the cause/effect nature of this relationship. Nonetheless, collectively, these data suggest that low doses of dietary psyllium can ameliorate inflammation, and perhaps metabolic abnormalities associated with this state, but significant reductions in adiposity require relatively high doses of this fiber.

Figure 2. Psyllium protects against HFD-induced metabolic syndrome in a dose-dependent manner. Male 6- to 8-week-old C57Bl/6 mice were fed with indicated diet for 28 days. (a) Relative body weight over time. (b) Epididymal fat pad weight. (c) Glucose tolerance test and respective areas under the curve. (d) Insulin tolerance test and respective areas under the curve. (e) Colon length. (f) Colon weight. (g) Cecum weight. (h) Spleen weight. (i) Colonic expression of pro-inflammatory cytokines by q-RT-PCR. Data are expressed as means ± SEM of n = 5 mice per group. Significance was determined by the ANOVA variance test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

We next turned our attention to investigating the mechanism by which psyllium protected against metabolic syndrome. Psyllium’s protection against DSS colitis is associated with an increase in serum bile acids (BA) and furthermore requires the BA receptor, FXR⁷. We reasoned that its protection against diet-induced obesity may also be FXR-mediated. In alignment with this possibility, enrichment of LF-HFD with psyllium indeed led to increased serum BA (serum BA was 13.8 ± 4.6 and 24.6 ± 3.7 $\mathit{\mu }$ M, respectively, for mice consuming LF-HFD vs Psyllium-enriched diet, P = 0.0519), analogous to its impacts on BA in low-fat diets. However, the psyllium’s protection against metabolic syndrome maintained in FXR-KO mice arguing against this receptor playing a role (Figure 3A-D). Psyllium is viewed as a semi-soluble and partially fermentable fiber¹⁸, suggesting that fermentation might contribute to its beneficial impacts. However, relative to LF-HFD, the psyllium-enriched diet did not increase cecal short-chain fatty acids (SCFA), (Figure 4A), arguing that this fiber was not fermented in this context. Furthermore, psyllium’s protection against metabolic syndrome was not prevented by blocking fermentation with hops $\mathit{\beta }$-acids (Figure 4B–F), under conditions, which we previously showed block fermentation³. These findings, which contrast with our recent studies on inulin and showed both increased cecal SCFA and improved glycemic control in a manner prevented by hops $\mathit{\beta }$-acids, argued against the notion that psyllium’s ability to ameliorate metabolic syndrome was mediated by fermentation. In addition to fermentation, a major means by which inulin protects against diet-induced metabolic syndrome is by increasing production of IL−22, which is required for inulin’s beneficial impacts on metabolic syndrome³. In contrast, enrichment of LF-HFD with psyllium lowered colonic IL−22 levels arguing against a role for this cytokine in mediating psyllium’s metabolic benefits (Figure 4G).

Figure 3. Psyllium protective effect against HFD-induced metabolic syndrome is FXR independent. Male 6- to 8-week-old FXR-KO (C57bl/6 FXR^−/−) mice were fed with indicated diet for 28 days. (a) Relative body weight over time. (b) Epididymal fat pad weight. (c) Glucose tolerance test and respective areas under the curve. (d).5 hours fasting glucose level. Data are expressed as means ± SEM of n = 4 mice per group. Significance was determined by ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 4. Psyllium did not induce SCFA and did not require fermentation to protect against HFD-induced metabolic syndrome. Male 6- to 8-week-old C57Bl/6 mice were fed with indicated diet for 28 days. (a) Levels of SCFA in the cecum. (b) Relative body weight over time. (c) Epididymal fat pad weight. (d) 5 hours fasting glucose level. (e) Colon length. (f) Colon weight. (g) Relative expression of IL-22. Data are expressed as means ± SEM of n = 5 mice per group except for the following conditions, in which n = 4, due to a mouse being euthanized prior to completion of experiment for reasons not related to experimental conditions (4A: Psy group, 4C-F: LF-HFD group, and 4 G: GBC group). Significance was determined by ANOVA. *P < 0.01, **P < 0.0001.

The above-mentioned impacts of inulin are associated with, and may require, its ability to restore gut microbiota density, i.e. number of bacteria per gram feces or colonic content, which is decimated by low-fiber diets. In contrast, we found that, analogous to its impacts in low-fat diets, enrichment of LF-HFD with psyllium has the opposite impact on bacterial load (Figure 5A). Specifically, relative to GBC-fed mice, those consuming LF-HFD exhibited an 8-fold drop in fecal bacterial density. Enriching LF-HFD with psyllium resulted in a further 6-fold drop in this bacterial density, indicating that, further contrast to inulin, psyllium was not protecting against metabolic syndrome by broadly nourishing gut microbiota. Nonetheless, use of germ-free mice indicated that psyllium’s ability to prevent negative consequences of LF-HFD was indeed microbiota-dependent. Specifically, in accord with other studies, weight gain increases in fat pad mass and hyperglycemia induced by LF-HFD (measure of glucose and insulin tolerance testing is challenging in gnotobiotic conditions) was reduced in germ-free mice relative to conventional mice but nonetheless clearly evident (Figure 5B-E). However, while enrichment of LF-HFD with psyllium fully protected conventionally colonized mice from these indices of metabolic syndrome, it failed to provide any amelioration of these parameters in germfree mice. Nor did the psyllium-enriched diet increase the colon length in germ-free mice (Figure 5B-E). These results suggest that the gut bacteria in psyllium fed mice, although 50-fold less abundant than that of GBC mice, were nonetheless pivotal in mediating psyllium beneficial metabolic impacts.

Figure 5. Psyllium reduces bacterial density, and its metabolic benefits require the presence of a minimal gut microbiota. (a). Male 6- to 8-week-old C57Bl/6 mice (conventionally colonized) were fed indicated diet for 28 days and bacterial density was measured. (B−F). Male 6- to 8-week-old C57Bl/6 mice maintained in conventionally colonized (Conv), germ-free (GF), and altered Schaedler flora (ASF) state were fed with the indicated diet for 28 days. (b) Relative body weight over time. (c) Epididymal fat pad weight. (d) 5 hours fasting glucose level. (e) Colon length. Data are expressed as means ± SEM of n = 5 mice per group for conventional mice, N = 4 mice per group for the GF mice, and n = 5 mice per group for ASF mice. (f) Bacterial density in the feces of ASF mice. (g) Measure of abundance of ASF strain by qPCR. (h) Relative change in ASF strain in comparison with GBC. Significance was determined by ANOVA. *P < 0.05, **P < 0.01 ***P < 0.0001.

To better understand the role of gut microbiota in mediating psyllium’s protection against DIO, we next examined whether the presence of minimal microbiota might be sufficient to mediate psyllium’s metabolic benefits. Specifically, we administered GBC, LF-HFD, and psyllium-enriched diet to Altered Schaedler Flora (ASF) mice, which are associated with an 8-species consortium that restores relatively normal metabolic and immune function¹⁹. Analogous to studies in conventionally colonized mice, psyllium provided complete protection against LF-HFD-induced adiposity (Figure 5B-C). In accord with previous results, the extent of LF-HFD-induced dysglycemia was reduced in ASF mice²⁰, but, nonetheless, was ameliorated by psyllium (Figure 5D). Furthermore, the psyllium-enriched diet increased the colon length in ASF mice (Figure 5E). We next explored the extent to which psyllium altered bacterial density in ASF-colonized germ-free mice. Analogous to conventional mice, psyllium resulted in a clear, about 4-fold, reduction in bacterial density (Figure 5F). Measure of levels of the individual ASF species by q-PCR revealed that psyllium also resulted in some modest changes in relative abundances of ASF strains including a reduction of ASF 500 (a Pseudoflavonifactor species) and a restoration of ASF 361 (Lactobacillus murinus) (Figure 5G-H). While discerning the role of these taxa will require additional studies, collectively, these data accord with the notion that psyllium’s ability to ameliorate diet-induced obesity and its associated consequences requires at least a minimal bacterial community to be present to interact with this fiber.

Discussion

Numerous clinical studies have demonstrated abilities of a variety of semi-purified plant-derived fiber supplements to prevent and/or mitigate a broad range of disease states. Mechanisms underlying these various beneficial impacts have not been well defined but rather generally attributed to broad mechanisms that might be engaged by numerous types of fibers, i.e. essentially presuming that most fibers act similarly. This mechanistic vagueness is, in large part, a consequence of inherent limitations in human studies. However, use of tractable mouse models has begun to address this knowledge gap, leading to appreciation that mechanisms by which fibers act are not, in fact, universally generalizable. Rather, it is becoming increasingly apparent that, in accord with their chemical diversity, specific fibers utilize distinct mechanisms for phenotypic impacts. Such appreciation led us to investigate mechanisms underlying impacts of psyllium, a fiber of particular importance in that it is widely consumed and shows a broad array of beneficial impacts. Psyllium has long been used by humans to alleviate functional gastrointestinal disorders, i.e. constipation/irritable bowel syndrome, and is also well appreciated to alleviate some aspects of metabolic syndrome. Furthermore, psyllium has also shown promise in management of ulcerative colitis, which we recently reported is mediated by activation of the FXR, bile acid receptor⁷. Herein, we report elucidation, albeit not the molecular level of mechanisms by which this fiber alleviates metabolic syndrome. Specifically, we rule out a panel of mechanisms previously proposed to mediate actions of psyllium, and other fibers, while finding that psyllium’s protection against diet-induced obesity (DIO) requires the presence of at least a minimal gut microbiota.

The notion that psyllium’s ability to alleviate DIO is mediated by gut microbiota comports with the notion that psyllium is semi-fermentable and that fermentable fibers are thought to play a critical role in maintaining a stable health-promoting gut microbiota. Indeed, switching mice from a GBC diet, which is naturally rich in fiber when compared to a low-fiber diet or diet containing only the non-fermentable fiber cellulose, results in a 10-fold reduction in the total number of bacteria present in the cecum and colon³. This reduction in bacterial density is accompanied by a stark reduction in intestinal mass, encroachment of bacteria that remain, and low-grade inflammation, which is proposed to promote insulin resistance and other metabolic abnormalities that comprise metabolic syndrome, especially when the low-fiber diet is also high in saturated fat content. Enriching low-fiber diets with the fermentable fiber inulin restored intestinal mass and alleviated insulin resistance. Accordingly, upon appreciating that psyllium also restored intestinal mass and glycemic control, we presumed that it would also restore bacterial density. However, to our surprise, we found that, irrespective of fat content, enriching low-fiber diets with psyllium did not restore bacterial density but rather further lowered this parameter by an additional 6-fold. Psyllium’s reduction in gut bacterial loads correlated with reductions in levels of cecal SCFA and colonic IL−22 expression, both of which are known to be microbiota-dependent, increased by inulin, and help ameliorate insulin resistance^2,3. Our observation that psyllium did not increase SCFA levels accorded with fermentation blockade not reducing psyllium’s metabolic benefits but contrasts with findings by Llewellyn, who also observed psyllium lowered bacterial density but yet found that it increased SCFA⁸. We do not know the reason for this discrepancy. IL−22 is required for inulin’s ability to restore gut mass, prevent low-grade inflammation³, and ameliorate metabolic syndrome^2,21. Thus, the apparent lack of role for IL−22 and SCFA highlights that psyllium uses distinct mechanism(s) than inulin to both restore gut mass and ameliorate DIO.

Currently, how psyllium reduces gut bacterial density is not clear, as well as the extent to which its “antibiotic-like” action might contribute to its ability to ameliorate DIO. On the one hand, if the very low microbiota density of psyllium-fed mice was mimicking a germ-free state, we would have expected these mice to have a level of adiposity similar to that of germ-free mice fed with the psyllium-enriched obesogenic diet, but, in fact, both conventional and ASF mice fed with the psyllium-enriched diet had adiposity levels far below that of germfree mice fed with this diet. But, on the other hand, that psyllium also lowered bacterial density in ASF mice would be consistent with the notion that psyllium’s ability to reduce levels of one or more taxa might well contribute to its suppression of obesity. Yet, it is also quite plausible that one or more bacteria present in ASF mice and perhaps numerous taxa present in conventional mice metabolize psyllium or produce a metabolite(s), which subsequently drive its beneficial effects. In any case, we submit that psyllium complete inability to ameliorate DIO in germ-free mice argues against a mechanism that had been long proposed to mediate its metabolic benefits, namely, its bulk increased satiety/decreased appetite, as we do not see any reason why such a mechanism would not be operative in the absence of a gut microbiota. Rather, we submit that, like inulin, psyllium reduces hyperphagia by a microbiota-dependent mechanism, perhaps reflecting its ability to ameliorate low-grade inflammation, which can dysregulate satiety signaling. Our results also argue against a central role for psyllium’s well-established ability to sequester BA impacts on BA metabolism⁷ in that psyllium’s metabolic benefits were FXR-independent; moreover, it is difficult to envision that such sequestration would be an essential component of a microbiota-mediated mechanism by which this fiber ameliorates metabolic syndrome.

In conclusion, our finding that psyllium’s amelioration of metabolic syndrome requires some microbiota to be present but yet is not mediated fermentation suggests that this fiber’s beneficial impacts cannot be fully explained by well-established mechanisms by which other fibers have been shown to act. Defining microbiota-dependent mechanisms by which psyllium acts will require future studies, which should consider extra-intestinal impacts including those on liver and/or brain. Given the seemingly unique impacts of psyllium relative to other fibers, particularly its reduction of bacterial density that is associated with lowering of SCFA, which are widely recognized to be beneficial, make it hard to relate our findings to consumption of foods naturally rich in fiber. Nonetheless, psyllium is widely consumed as a dietary supplement and our results support the notion that consumption of psyllium supplements may provide an array of health benefits. However, the most striking impacts of psyllium on adiposity resulted from doses that would be difficult to achieve in humans. Yet some benefits, especially those most associated with ameliorating inflammation, were observed in response to lower doses tested. While relating such levels of psyllium consumption to humans eating a diverse diet that likely already contains various types of fiber is inherently difficult, we submit the existing body of work that shows that some, albeit often modest, beneficial metabolic impacts of psyllium can be achieved^22–25. Yet, we argue that a better understanding of mechanism, which we hope to be achieved in future studies, may allow a means to harness the mechanism of psyllium’s action potentially achieving benefits of greater magnitude. Thus, defining the microbiota-mediated mechanism by which psyllium prevents DIO remains an important research challenge.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All data supporting our conclusion herein can be found within this manuscript.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/19490976.2023.2221095

Additional information

Funding

This work was supported by NIH grants DK099071 and DK083890 to A.T.G. and CA219144 to M.V.K. A.B. was supported by CCF RFA 663306.