Sex Differences in Dietary Copper-Fructose Interaction-Induced Alterations of Gut Microbial Activity are Not Correlated to Hepatic Steatosis

Background: Inadequate copper intake and increased fructose consumption represent two important nutritional problems in the US. Diet copper-fructose interactions alter gut microbial activity and contribute to the development of nonalcoholic fatty liver disease (NAFLD). The aim of this study is to determine whether dietary copper-fructose interactions alter gut microbial activity in a sex-differential manner, and whether sex differences in gut microbial activity are associated with sex differences in hepatic steatosis. Methods: Male and female weanling Sprague-Dawley (SD) rats were fed ad libitum with an AIN-93G puried rodent diet with dened copper content for 8 weeks. The copper content is 6mg/kg and 1.5mg/kg in adequate copper diet (CuA) and marginal copper diet (CuM), respectively. Animals had free access to either deionized water or deionized water containing 10% fructose (F) (w/v) as the only drink during the experiment. Body weight, calorie intake, plasma ALT, AST and liver histology were evaluated. Fecal microbial contents were analyzed by 16S rRNA sequencing. Fecal and cecal short chain fatty acids (SCFAs) were determined by gas chromatography-mass spectrometry (GC-MS). Results: Male and female rats exhibit similar trends of changes in the body weight, body weight gain and calorie intake in response to dietary copper and fructose, with a generally higher level in male rats. Several female rats in CuAF group developed mild steatosis, while no obvious steatosis was observed in male rats fed with CuAF or CuMF. Fecal 16S rRNA sequencing analysis revealed distinct alterations of the gut microbiome in male and female rats. Linear discriminant analysis (LDA) effect size (LEfSe) identied sex-specic abundant taxa in different groups. Further, total SCFAs, as well as, butyrate were decreased in a more pronounced manner in female CuMF rats than in male rats. Of note, the decreased SCFAs are concomitant with the reduced SCFA producers, but not correlated to hepatic steatosis. Conclusions: Our data demonstrated sex differences in the alterations of gut microbial activities and hepatic steatosis in response to dietary copper-fructose interaction in rats. Tissue-specic responses to dietary copper and fructose likely contribute to the sex differences in gut microbial activity and metabolic phenotype. two-way testing factors of copper (Cu), fructose (F) and interactions (Cu × F), followed by multiple comparison test. Statistical signicance was set to p ≤ 0.05. P values are displayed for the factors Cu, F, and Cu × F. NS, P> 0.05. * versus CuA; # versus CuAF; $ versus CuM. CuA, adequate copper diet; CuM, marginal copper decient diet; CuAF, adequate copper diet +10% fructose drinking; CuMF, marginal copper decient diet +10% fructose drinking. design of research, data analysis and interpretation, manuscript preparation; M.S., F.Y., X.M., X.Y., X.Z., data collection and analysis; X.L., E.C.R., data analysis, R.A.P., Z.D., data interpretation and intellectual contribution to manuscript preparation; C.J.M., overall research direction and support.


Introduction
The prevalence of nonalcoholic liver disease (NAFLD) in the United States has increased rapidly in the past two decades, from 19-24%, which is close to the global prevalence of 25.24% (1,2). Based on the epidemiological data from obesity and type 2 diabetes in adults, the estimated prevalence of NAFLD will continue to increase up to 33.5% in 2030; nonalcoholic steatohepatitis (NASH) will increase proportionately from 20% of NAFLD to 27%, ranking it as a top indication for liver transplantation (3,4).
Of note, NAFLD and NASH exhibit age and sex differences, with a higher prevalence in men than in premenopausal women. Conversely, a higher rate of NAFLD was found among the postmenopausal women (5-7). In agreement with this nding, sex differences also exist in the risk factors, such as obesity and type 2 diabetes (8,9). Biological sex differences are exhibited in many physiological phenomenon, including fat distribution, triglyceride storage in liver and muscle (10), fatty acid and glucose metabolism (11), etc. Therefore, understanding sex differences in physiology and pathophysiology is required for precision medicine.
Sex hormones and sex chromosome are two major factors driving sex differences (7). The role of sex hormones has been demonstrated in both human and animal studies. For example, postmenopausal women with estrogen de ciency display a higher risk for NAFLD progression to brosis (12). In contrast, liver injury was improved by hormone replacement therapy in postmenopausal women with type 2 diabetes (13). Ovariectomized (OVX) female rats exhibit exacerbated hepatic steatosis when exposed to high-fat high-fructose diet (HFFD), which was reversed by estrogen replacement (14). A four core genotypes mouse model (XX gonadal male and female, XY gonadal male and female) allows for the identi cation of whether sex differences arise from the sex chromosome complement. Using this approach, it was revealed that XX mice are prone to developing obesity and fatty liver in response to high fat diet, regardless of sex hormones (15).
In addition to genetics and sex hormones, diet is a key environmental factor leading to sex differences in metabolic diseases (16). Copper and fructose are two dietary factors known to be critical in the pathogenesis of NAFLD (17)(18)(19)(20)(21)(22). Sex differences in the metabolic effects of fructose and/or copper de ciency have been noted in rodents (23)(24)(25)(26) as well as in humans (27,28), with more harmful effects reported in males and more protective effects in females, which is consistent with the sex differences in NAFLD (7). In fact, sex differences in fructose-induced metabolic effects are more complex and vary by tissue and organ (14,29,30). Although sex hormones are one of the factors leading to sex differences in copper-fructose interaction-induced metabolic disorders (26), the underlying mechanisms are largely unknown.
A growing body of evidence has shown that gut microbiota play a causal role in driving the development of obesity, diabetes and NAFLD (31)(32)(33)(34). Diet, as one of the most common environmental factors, shapes the gut microbiome (35). Interestingly, diet-induced alterations of gut microbiota exhibit a sex-dependent phenotype (36,37). Previous studies have shown that distinct alterations of the gut microbiome are linked to speci c metabolic traits (38) as well as to different stages of NAFLD (39,40), leading to the hypothesis that sex differences in the gut microbiota is linked to distinct metabolic phenotypes or disease severity. Our previous studies have shown that dietary copper-fructose interactions shifted gut microbiota and correlated to the development of hepatic steatosis in male rats (41,42). Given that diet shapes the gut microbiome in a sex-speci c manner (36), we aimed to determine whether dietary copper-fructose interaction alters gut microbiota and induces hepatic steatosis in a sex-dependent manner and whether sex differences in metabolic phenotype contribute to the distinct alterations of the gut microbiota.

Animals and diets
Male and female weanling Sprague-Dawley rats (35-45 g) from the Harlan Laboratories (Indianapolis, IN) were fed (ad lib) an AIN-93G puri ed rodent diet with a de ned copper content. The rats received either 1.5 ppm or 6.0 ppm of copper as marginal or adequate doses, respectively, for 8 weeks. Control animals were fed adequate copper with no added fructose. The animals were single housed in stainless steel cages without bedding in a temperature-and humidity-controlled room with a 12:12 h light-dark cycle. Animals had free access to either deionized water or deionized water containing 10% fructose (w/v).
Fructose enriched drinking water was changed twice a week. Food consumption and body weight were monitored on a weekly basis. After 2 hours fasting, all the animals were sacri ced under anesthesia with ketamine/xylazine (100/10 mg/kg I.P. injection). Blood was collected from the inferior vena cava, and citrated plasma was stored at -80 °C for further analysis. Portions of liver tissue were xed with 10% formalin for subsequent sectioning, while others were snap-frozen with liquid nitrogen. All studies were approved by the University of Louisville Institutional Animal Care and Use Committee, which is certi ed by the American Association of Accreditation of Laboratory Animal Care.
16S ribosomal RNA Gene (16S rRNA) Library Preparation and Sequencing on the Illumina MiSeq Fecal pellets were collected into sterile tubes at the end of the experiment and stored at -80 °C. Microbial genomic DNA was extracted from frozen fecal samples using DNeasy PowerSoil kit (Cat#:12888-100, Qiagen, Germantown, MD) according to the manufacturer's instructions.

Sequencing Data Analysis
Quality control of raw sequence les was performed using FastQC and further analyzed using QIIME 2 (43). Chimeric sequences were removed by using both reference and de novo approaches. The nonchimeric sequences were assigned to operational taxonomy units (OUTs) at 97% similarity to the clustered Greengenes database (44,45). Rarefaction curve using the observed OTU provided by QIIME was used as a metric of α-diversity (46). Principal coordinate analysis (PCoA) was performed to compare microbial community structure between samples (β-diversity) using both weighted and unweighted UniFrac (47). Heat map analysis of OTU abundance was performed using R software (https://www.rproject.org/). Linear discriminant analysis (LDA) effect size (LEfSe) method was used to nd the most differentially abundant enriched microbial taxa between the different diets. The analysis was performed on Galaxy platform. The data generated from LEfSe analysis was shown by cladogram and histogram with LDA score > 2 and a signi cance of α < 0.05, as determined by Wilcoxon rank-sum test (48)(49)(50). A venn diagram was used to show genus distribution between groups.
SCFAs measurement by gas chromatography-mass spectrometry (GC-MS) About 50 mg of cecal and fecal stool samples were weighed and polar metabolites were extracted of for GC-MS analysis using established methods as described previously (51) .

Statistical analysis
Data were expressed as mean ± SD (Standard Deviation) and analyzed using two-way ANOVA to test the factors of copper, fructose, and their interactions (copper × fructose), followed by Tuckey's Multiple Comparison Test. The Kruskal-Wallis test was used for pairwise comparison between treatment groups (α-diversity). Comparison of the mean distance matrix (β-diversity) between two treatment groups using PERMANOVA (a non-parametric method for multivariate analysis of variance) with permutation tests was based on Unifrac distance matrix (999 Monte Carlo permutations). Differences at P ≤ 0.05 were considered to be statistical signi cant.

Results
Characterization of dietary copper-fructose interaction on metabolic phenotypes in male and female rats Male and female rats exhibit similar trends of changes in the body weight and body weight gain in response to dietary copper and fructose, with a generally higher level in male rats ( Fig. 1, Table 1 and   Table 2). Two-way ANOVA analysis showed that the liver weight of female rats, but not male rats, was affected by dietary copper content within the eight-week period. The liver/body weight ratio was altered by both dietary copper and fructose. However, copper-fructose interaction was apparent only in female rats. While the variations of perigonadal white adipose tissue (WAT) weight as well as WAT/body weight ratios were related to dietary copper content in male rats, they were more likely to be affected by dietary fructose in female rats. The energy e ciency ratio (EER, %), i.e., the ratio of body weight gain and total energy intake (52,53), was decreased by dietary fructose in both male and female rats compared to their controls, suggesting the metabolic effects of fructose may not be contributed to the calorie intake. Ad libitum feeding of fructose via drinking water led to a signi cant increase in water intake and a decrease in pellet food intake. Although there was a trend toward an increase in the total energy intake in rats fed with fructose compared to those without; the difference did not reach the statistical signi cance in either males or females.  Hepatic manifestations in response to dietary copperfructose interaction in male and female rats Neither male nor female rats showed obvious liver injury in terms of plasma ALT and AST after being exposed to CuA or CuM diets with or without 10% fructose (w/v) for 8 weeks (Fig. 2A). Three of eight female rats fed with CuA plus fructose developed mild steatosis, characterized with macrosteatosis around portal area. Only very mild microsteatosis could be visualized in either CuMF female rats or male rats fed with marginal copper diet and/or fructose (Fig. 2B). Compared to our previous study with AIN-76 diet (containing 49% sucrose) and 30% fructose (w/v) in the drinking water (21), the extent of hepatic steatosis is mild and no apparent liver injury was detected. Despite mild steatosis induced under the current conditions, sex differences still were detected, with female CuAF rats showing hepatic steatosis.
Distinct alterations of fecal gut microbiota in response to dietary copper and fructose between male and female rats as analyzed by 16S rRNA sequencing To examine whether copper-fructose interaction alters the gut microbiome in a sex-speci c manner, we performed 16S rRNA sequencing of fecal stool DNA. In male rats, either fructose or CuM results in a trend of decrease in alpha-diversity in terms of the observed OTU. However, only the difference between CuA and CuAF reached statistical signi cant (CuA versus CuAF, P = 0.049), suggesting fructose feeding led to reduced species richness in male rats (54). There was no signi cant difference between groups of female rats in terms of observed OTU, suggesting neither fructose nor CuM alters the species richness of the gut microbiota in female rats. (Fig. 3A, supplementary Table 1). Beta-diversity was evaluated by UniFrac analysis (47). Unweighted UniFrac is a qualitative β-diversity measure, which detects the difference in the presence or absence of lineages of bacteria in different communities (55). Unweighted UniFrac analysis demonstrated that the mean distance between groups CuA and CuAF, and CuA and CuM were signi cantly different in male rats (p < 0.05), suggesting that the bacteria community was different between these groups. (Fig. 3B, top panel, supplementary Table 2). In female rats, unweighted UniFrac analysis showed that the signi cant differences were between groups CuM and CuMF (p < 0.05) (Fig. 3B, bottom panel, supplementary Table 2). Weighted UniFrac measure was used for detecting differences in abundance (55), and no signi cant differences were detected between the four treatment groups in male or female rats. These results suggested that CuA and CuM lead to different bacterial community in male rats. Fructose feeding altered bacterial community in CuA fed male rats, but in CuM fed female rats.
At the phylum level, fructose feeding led to a remarkable increase in the abundance of Bacteroidetes, Proteobacteria, and a decrease in Firmicutes independent of dietary copper content. In male rats, only the abundance of Bacteroidetes and Proteobacteria was altered by dietary fructose and the effect was less pronounced compared to female rats (Fig. 3C, supplementary Table 3 and Table 4). In agreement with this, more families and genera under the phylum Bacteroidetes, Firmicutes and Proteobacteria were altered in female rats compared to male rats. For example, Bacteroidaceae, Bacteroides, Lachnospiraceae, Erysipelotrichaceae, Allobaculum, Alcaligenaceae, and Sutterella were markedly shifted in female rats, but not in male rats. Even among the commonly changed taxa, such as Porphyromonadaceae, Parabacteroides and Blautia, the factors leading to such changes are different between males and females, as shown by two-way ANOVA (supplementary Table 3-6 and Fig. 4). In addition to the sex differences in response to dietary fructose and marginal copper, the composition of gut microbiota is also different between male and female rats when exposed to adequate copper diet, which was considered as a normal control. A higher abundance of Firmicutes and a lower abundance of Bacteroidetes were observed in female rats than in male rats, leading to a higher Firmicutes/Bacteroidetes ratio in females rats ( Collectively, female rats exhibit more pronounced alterations of gut microbiota and fructose plays a dominant role.
LEfSe identi ed microbiota signature associated with dietary copper and fructose To further identify more speci c taxa changes in gut microbiome by dietary copper and fructose, LEfSe analysis was performed using 16S rRNA metagenomic data (48). 15 and 26 differentially abundant taxa were identi ed with LDA score higher than 2 in male and female rats, respectively ( Fig. 5A and 5B). The Proteobacteria and Bacteroidetes were enriched in CuAF and CuMF group, respectively, in both male and female rats. No speci c taxa were identi ed to be enriched in CuM male rats. A highest number of abundant taxa was in CuMF group (7 of 15 in male and 12 of 26 in female). Sex differences in abundance also existed in CuA rats, which were considered as normal controls. Female CuA rats were characterized by enriched Firmicutes, particularly, Lachnospiraceae. Of note, while Porphyromonadaceae and Parabacteroides were enriched in CuMF male rats, they were also enriched in female CuAF rats, which is consistent with the mean abundance data analysis (supplementary Table 3 and Table 4). Particularly, abundant beta-Proteobacteria and Erysipelotrichi in CuMF rats as well as abundant alpha-Proteobacteria in CuAF rats were identi ed in female rats. Thus, distinct abundant taxa were identi ed by LEfSe analysis between male and females. The Venn diagram plot showed 51 shared genera by four groups in both male and female rats. There are total 65 and 56 detected genera in male and female rats, respectively. Fructose and marginal copper led to reduced genera in male rats, but an increase in female rats. Six genera were not altered by fructose or marginal copper diet in male rats, but only two were not altered in female rats (Fig. 5C), suggesting more genera changes occur in female rats.

Sex differences in Fecal short chain fatty acids in response to dietary copper-fructose interaction
To better understand the sex differences in microbial activities induced by dietary copper and fructose, we measured SCFAs by GC-MS in cecal and fecal contents. Acetate, propionate, and butyrate are the predominant SCFAs in cecal and fecal contents. Overall, the levels of total as well as individual SCFAs were higher in cecal contents than that in fecal contents in both male and female rats. While the level of total cecal SCFAs is higher in males, the level of total fecal SCFAs are comparable between male and female rats. Fructose feeding resulted in a decrease of total SCFAs in both cecal and fecal contents in CuA-and CuM-fed rats; however, a signi cant decrease was found in female CuMF rats. A similar trend of alterations in SCFAs, but to a lesser extent, was observed in male rats, as shown in Fig. 6A. Consistently, acetate, propionate and butyrate were all markedly decreased in female CuMF rats (Fig. 6B). In addition, decreased total SCFAs was associated with the relatively increased proportion of acetate and decreased  (Fig. 6C). Collectively, a substantial decrease of SCFAs was seen in female rats, and profoundly so in CuMF group. Two-way ANOVA showed that the altered SCFAs was most likely due to the additive effect of copper and fructose in female rats, but the decrease in SCFAs in male rats was only attributable to copper.

Discussion
Copper-fructose interaction-induced metabolic effects exhibit sex dimorphism (23,25). Sex speci c alterations of gut microbiota in response to a speci c diets have been demonstrated in a variety of studies (57)(58)(59). Given that the gut microbiota play a causal role in driving the development of metabolic diseases, we aimed to determine whether sex-speci c alterations of the gut microbiota are linked to hepatic steatosis. Our data showed that sex differences do exist in the gut microbiota, gut microbiota metabolites such as SCFAs, and hepatic steatosis following dietary copper and fructose exposure. Female rats exhibited more pronounced alterations in the abundance of various taxa than that did male rats at multiple taxa levels, including phylum, family and genus. The number of distinct abundant taxa identi ed by LEfSe was also higher in female rats than in male rats. In addition, SCFAs were decreased to a greater extent in female rats compared to male rats, particularly in CuMF group. Moreover, female rats with an adequate copper diet developed mild, but apparent steatosis after 8 weeks of added fructose feeding (CuAF), but female CuMF rats, which showed the most signi cantly altered gut microbial activity, did not. Therefore, the altered gut microbial activity does not correlate with the hepatic fat accumulation.
SCFAs are the end products of microbial fermentation of indigestible ber, and they play a critical role in energy homeostasis and metabolism (60). In our study, we found signi cantly decreased SCFAs, particularly butyrate, concomitant with the reduced butyrate producers, Lachnospiraceae and Ruminococcaceae (61), in CuMF female rats, implying the most signi cantly altered gut microbial activities in this group. We found mild hepatic steatosis in CuAF female rats; thus, it is unlikely that this hepatic steatosis is attributable to the metabolic effects of gut microbiota. Accelerated de novo lipogenesis (DNL) is known to contribute to fructose-induced hepatic steatosis (62,63). However, the underlying mechanisms are unclear. A recent study demonstrated a two-point mechanism leading to fructose-induced hepatic steatosis. One part is gut bacteria derived acetate which serves as a substrate for acetyl-CoA synthesis via acyl-CoA synthetase short chain family member 2 (ACSS2) in the liver. Second, fructose metabolism in hepatocytes activates a signal leading to lipogenic gene expression (64).
Interestingly, the most signi cantly changed SCFAs occurred in CuMF rats, in which exacerbated liver injury and steatosis were seen in our previous study when rats were exposed to a high fructose diet via 30% fructose (w/v) in the drinking water and sucrose-enriched diet (AIN-76) (21). This nding suggests that hepatic steatosis may be related to the amount of fructose intake. In support of this, a recent study demonstrated that dietary fructose is primarily metabolized in the small intestine and only excess fructose intake spills over to the colon microbiota and liver (65). Previous studies showed that either inhibition of fructose metabolism in the liver (66) or elimination of gut microbiota by antibiotics (67) protected against fructose induced hepatic steatosis, indicating that fructose metabolism in both liver and gut microbiota is required to facilitate the development of steatosis. When a large amount of fructose intake saturates the capacity of the small intestine metabolism, presumably excess fructose will proceed to the colon, the gut microbiota and liver. However, the priority of excess fructose to be distributed and metabolized in colon microbiota or liver or other tissues is unclear when a modest amount of fructose was ingested. It has been shown that dietary copper-fructose interaction exacerbates copper de ciency-induced metabolic syndrome, likely due to impaired intestinal copper absorption because of excess fructose ingestion (21,68). Whether the extent of interaction relates to the relative amounts of copper and/or fructose, and subsequent metabolic effects remain largely unknown and warrant further study.
Despite signi cantly changed gut microbiota and SCFAs in CuMF rats, only a few of the female rats developed modest steatosis in CuAF group, suggesting the altered gut microbial activities were not su cient to lead to a signi cant phenotype change in the current study. At this point, it seems that hepatic steatosis and the shifts in gut microbial activity are not correlated. Of note, Porphyromonadaceae and Parabacteroides are two of the microbiota signatures associated with CuAF in female rats, although with relative low abundance (1.52%), which is different from male rats identi ed by LEfSe. Whether increased abundance of Porphyromonadaceae and Parabacteroides plays a causal role in fructoseinduced hepatic steatosis needs to be examined.
Sex difference in fructose-induced metabolic effects are mixed (24,69,70). In contrast to previous studies on copper-fructose interactions (23,25,26), our results showed that female rats are relative sensitive to fructose-induced hepatic steatosis. The discrepancy may be attributed to several factors.
First is the dose of copper and fructose. A lower dose of copper (0.6 ppm) and a higher dose of fructose (30-62%) were used in Field's as well as in Morrell's studies (23,26). It appeared that males are more sensitive to the deleterious effects of copper de ciency. In our study, marginal copper diet (1.5 ppm) and 10% fructose (w/v) in the drinking water were used, presumably leading to less pronounced copperfructose interactions and metabolic effects than previous studies (23,26). Second, the activities of fructose-metabolizing enzymes and intermediate metabolites differed by sex and copper level (71). In fact, the activities of liver enzymes involved in lipogenesis was affected not only by the type of carbohydrate, but also by the quantity (72). Lastly, differences in facilities, diet components and species as well as experimental durations may all contribute to discrepancy (25,73,74).
In support of our results, a previous study demonstrated that weanling female rats exhibit a higher rate of acetate incorporation into lipids in liver compared to male rats (75), suggesting a higher lipogenic capacity in female rats. However, there is species difference in driving the lipogenic enzyme activity in response to carbohydrate (72). In human studies, fructose-induced increase in hepatic DNL and decrease in fatty acid oxidation was more pronounced in men and premenopausal women than in postmenopausal women (28,63,76,77). Sex hormones are the known factors regulating sex dimorphism of fructose-related metabolic effects (7). However, the molecular underpinnings remain elusive. Recent studies showed that GLUT8 mediates distinct metabolic effects between males and females in response to dietary fructose (29,30,78). GLUT8 is a dual-speci city glucose and fructose transporter, which was found to be abundantly expressed in both murine and human liver and intestine (30,78,79). Interestingly, while GLUT8 mutation does not alter intestinal fructose absorption in male mice (29), it enhances intestinal fructose absorption in female mice, which was associated exacerbated hypertension, hyperinsulinemia, and hyperlipidemia when fed with high-fructose diet (30). Conversely, GLUT8-de cient male mice are protection from high-fructose diet-induced dyslipidemia, glucose intolerance and hypertension (29). These studies revealed an important molecular mechanism underlying the tissuespeci c and sex-speci c divergence in response to fructose.
A potential limitation of the current study is the one time analysis of gut microbiota and hepatic steatosis. Although female rats displayed earlier development of steatosis, it is di cult to predict the ultimate severity of steatosis and disease progression. Since male rats exhibit decreased diversity of gut microbiome, and given that the microbial gene richness is associated with in ammation, insulin resistance and dyslipidaemia (80,81), it is plausible that male rats develop steatosis with a prolonged duration on experimental regime. Thus, long-term and multiple time points evaluation will provide more accurate pro les of disease progression in the context of sex difference. However, sex differences observed in animal studies are under strictly de ned experimental conditions. Therefore, a caveat must be noted when extrapolating animal data to human, as humans have much more complex genetic and environmental factors than experimental animals.

Perspectives and Signi cance
In summary, our current study provides evidence of sex-speci c alterations in gut microbial activities and hepatic steatosis in response to dietary copper-fructose interaction in a rat model. However, sex differences in the liver and gut do not seem to be related. Future studies deciphering the molecular mechanisms would help us better understand sex-speci c responses to dietary copper-fructose interactions.

Conclusions
Our data demonstrated sex differences in the alterations of gut microbial activities and hepatic steatosis in response to dietary copper-fructose interaction in rats. Tissue-speci c responses to dietary copper and fructose likely contribute to the sex differences in gut microbial activity and metabolic phenotype. Availability of data and materials