Hoch

Nature Metabolism volume 4, pages 1713–1731 (2022)Cite this article

34k Accesses

4 Citations

290 Altmetric

Metrics details

Dietary restriction is widely used to reduce fat mass and lose weight in individuals with or without obesity; however, weight regain after dieting is still a big challenge, and the underlying mechanisms remain largely elusive. Here we show that refeeding after various types of dieting induces quick fat accumulation in mice and enhanced intestinal lipid absorption contributes to post-dieting fat mass increase. Moreover, refeeding after short-term dietary restriction is accompanied by an increase in intestinal Lactobacillus and its metabolites, which contributes to enhanced intestinal lipid absorption and post-dieting fat mass increase; however, refeeding a high-protein diet after short-term dietary restriction attenuates intestinal lipid absorption and represses fat accumulation by preventing Lactobacillus growth. Our results provide insight into the mechanisms underlying fat mass increase after dieting. We also propose that targeting intestinal Lactobacillus to inhibit intestinal lipid absorption via high-protein diet or antibiotics is likely an effective strategy to prevent obesity after dieting.

The global prevalence of obesity has tripled since the mid-1970s1. Obesity represents a major health challenge which substantially increases the risk of diseases such as type 2 diabetes mellitus, fatty liver disease, cardiovascular disease, several cancers and even severe coronavirus disease 2019 (COVID-19)1,2,3. The fundamental cause of obesity is the imbalance of energy homeostasis with too many calories incorporated and too few calories expended2,4. Both environmental and genetic factors that affect energy homeostasis contribute to the development of obesity5,6. For environmental factors, total calorie intake, physical activity, ambient temperature, processed foods, antibiotic usage, early-life influences, pollution, stress, depression, lack of sleep and circadian misalignment are all very important factors causing obesity2,7,8,9. Nevertheless, whether there is any neglected but important environmental factor leading to obesity remains to be investigated.

Dietary restriction (DR) has been considered as an important strategy to improve lipid metabolism and reduce fat mass in both lean and obese mammals10,11,12,13,14. There is an increasing prevalence of DR in healthy weight females and males15,16. Both calorie restriction and intermittent fasting as specific types of DR can significantly reduce body mass in lean and slightly overweight adults mainly due to fat loss17,18. Because long-term voluntary food restriction is impractical for most people, short-term dietary restriction (SDR) or even simple meal skipping become popular for weight management19,20,21. Acute fasts are frequently undertaken in many individuals for cosmetic, health, religious, medical or cultural reasons22,23,24 and meal skipping rates in young adults (18–30 years) ranged between 5 and 83%21. It has been reported that weight loss induced by DR in healthy weight males or females will be regained after refeeding25,26,27; however, whether refeeding after dieting contributes to obesity and the underlying mechanisms are still largely unclear and new strategies to prevent weight regain after dieting need to be investigated.

Diet emerges as a pivotal determinant of gut microbiota community structure and function28,29,30 and DR as a special diet modification prevalently applied to reduce body weight also alters gut microbiome composition31,32,33. Gut microbiota in response to diet has been shown to influence fat digestion and absorption in the small intestine34,35. Moreover, experiments that compared germ-free and colonized mice or analyzed the influence of nutrients and other environmental factors that qualitatively change the composition of gut microbiota show that gut microbiota influences the metabolism of cells in the liver and adipose tissues and thereby modulates lipid homeostasis36,37,38. Gut microbiota contributes to host metabolism, including food intake, digestion, absorption and energy expenditure through the production of a myriad of metabolites, such as short-chain fatty acids, bile acids, amino acids and their derivatives39,40,41,42. Gut microbiota-derived metabolites exert their effects within the host as signaling molecules and substrates for metabolic reactions39,43,44. Studies find that microbiota-derived metabolites such as acetate, lactate, succinate and myristoleic acid can regulate intestinal lipid absorption and lipid metabolism in adipose tissues45,46,47. Gut microbiota and its derived metabolites can act as a transducer of dietary cues, including DR to regulate host metabolism48. For example, intermittent fasting promotes white adipose browning by shaping gut microbiota and leading to the elevation of intestinal acetate and lactate49. In addition, a previous investigation identifies an intestinal microbiome signature that persists after normal chow feeding of obese mice induced by a high-fat diet, a model of recurrent obesity, diminishes the levels of intestinal metabolite flavonoids and contributes to faster weight regain upon re-exposure to a high-fat diet50; however, how gut microbiota and its derived metabolites contribute to obesity after weight loss induced by DR in different models is yet to be further elucidated.

In this study, we designed ten types of dieting to investigate the effect of refeeding after dieting on body fat mass and explored the preventive effect of refeeding with a high-protein (HP) diet after dieting. The intestinal lipid absorption and lipid metabolism in white adipose tissue (WAT) were analyzed during refeeding after dieting. Moreover, gut microbiota and its derived metabolites were also studied during refeeding after dieting. Our data revealed that refeeding after dieting leads to HP diet-repressible accumulation of fat mass by increase in intestinal lipid absorption via enrichment of Lactobacillus and its metabolites.

To study the effect of refeeding after dieting on fat mass, we fed mice with 1-d food in 3 d or 4-d food in 6 d as models of SDR and then fed ad libitum (Fig. 1a). Typically, feeding with 10%, 25% and 65% food from day 1 to day 3, respectively led to a quick decrease in fat mass as expected (Fig. 1b). Notably, refeeding ad libitum after feeding with 10–25–65% food for 3 d led to quick fat mass accumulation (Fig. 1b and Extended Data Fig. 1a–d). Moreover, refeeding after feeding with 65–25–10% or 33.3% food daily for 3 d or 66.7% food daily for 6 d, also led to similar quick fat mass accumulation (Fig. 1c–e and Extended Data Fig. 1b–d). To further confirm the effect of refeeding after SDR on fat mass, we killed and dissected the mice. Refeeding after feeding with 10–25–65% food increased the size and weight of inguinal WAT (iWAT) and epididymal WAT (eWAT) (Fig. 1f,g and Extended Data Fig. 1f–j). Hematoxylin and eosin (H&E) staining revealed that refeeding after the SDR markedly increased the size of adipocytes in iWAT and eWAT (Fig. 1h and Extended Data Fig. 1k–m). Consistently, refeeding after the SDR led to the hypertrophy of interscapular brown adipose tissue (iBAT) and the accumulation of enlarged lipid droplets in iBAT cells (Extended Data Fig. 2a–c) and induced hepatic steatosis (Extended Data Fig. 2d,e).

a, Experimental design of SDR in mice feeding with 1-d food in 3 d or 4-d food in 6 d. AL, ad libitum; SDR(33.3% × 3), 33.3% food daily provided from day 1 to 3; SDR(66.7% × 6), 66.7% food daily provided from day 1 to 6. b–e, Refeeding after feeding with 1-d food in 3 d (b–d) or 4-d food in 6 d (e) markedly increased body fat mass, n = 9 biologically independent mice for AL, SDR(10–25–65%) and SDR(66.7% × 6), n = 7 for SDR(65–25–10%), and n = 8 for SDR(33.3% × 3). f, Representative images of isolated iWAT from the SDR(10–25–65%) group. The size for each square is 0.5 cm × 0.5 cm. D4D, day 4 at dark phase. g, The weight of iWAT in f (n = 12 biologically independent mice per group). h, Representative images of H&E-stained sections of iWAT in f. Scale bar, 50 μm. i–k, Refeeding after feeding with 2-d food in 3 d (i,j) or 1-d food in 2 d (k) significantly increased body fat mass; n = 8 biologically independent mice for AL, n = 9 for other groups. l,m, Refeeding after feeding with 66.7% food daily for 12 d (l) or 24 d (m) markedly increased body fat mass (n = 9 biologically independent mice per group). n, Refeeding after alternate-day fasting for 15 cycles markedly increased body fat mass (n = 9 biologically independent mice per group). The data shown in b–e, i–k or l–n were performed simultaneously with a single control experiment. Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test. a, P < 0.05; b, P < 0.01; c, ***P < 0.001; versus AL or D0 group. Exact P values can be found in Source Data Fig. 1.

Source data

To sufficiently investigate the effect of refeeding after DR on fat mass, we fed mice with 2-d food in 3 d, 1-d food in 2 d or 66.7% food daily for 12 d or as long as 24 d and then fed ad libitum (Extended Data Fig. 3a,g). Similarly, refeeding after these types of DR all led to quick fat mass accumulation (Fig. 1i–m and Extended Data Fig. 3b–e,h–k). Alternate-day fasting as an intermittent fasting regime has been considered to be as beneficial for reducing fat mass as daily DR51. We found that refeeding after alternate-day fasting also led to similar quick fat mass accumulation as refeeding after other types of DR (Fig. 1n and Extended Data Fig. 3g,m–p). These results demonstrate that refeeding after dieting leads to quick fat mass accumulation in mice.

To reveal the mechanism of fat mass increase after SDR, first we analyzed the energy expenditure. As shown in Fig. 2a,b and Extended Data Fig. 3r–y, we found that energy expenditure, physical activity and body temperature were not significantly altered after SDR. These data demonstrate that quick fat mass accumulation after SDR is not due to a decrease in energy expenditure.

a,b, Average energy expenditure (EE) from day 4 to day 5 in mice fed ad libitum (AL) or mice in the SDR(10–25–65%) (a) or SDR(33.3% × 3) group (b), n = 7 and 6 biologically independent mice for AL in a and b, respectively, n = 8 for SDR(10–25–65%), n = 6 for SDR(33.3% × 3). c, Hypothalamic mRNA levels of orexigenic genes Agrp, Npy and anorexigenic genes Pomc, Cart of mice in the SDR(10–25–65%) group (n = 8 biologically independent samples per group). d, Cumulative food intake of mice fed AL (n = 9 biologically independent mice), in the SDR(10–25–65%) group (n = 9) or SDR(33.3% × 3) group (n = 8). e,f, Refeeding with food as much as in D0 after feeding with 10%, 25% and 65% food (e) or 33.3% food (f) daily from day 1 to day 3 markedly increased body fat mass (n = 9 biologically independent mice per group). SDR(10–25–65%)−100% × 23 and SDR(33.3% × 3)−100% × 23, 10%, 25% and 65% food or 33.3% food daily provided from day 1 to day 3 and subsequently 100% food as much as in D0 daily provided for 23 d. g, Cumulative food intake of mice in e,f (n = 9 biologically independent mice per group). h, Fecal TG content in SDR(10–25–65%) group (n = 9 biologically independent samples for D0, D1, D2, D3, D4, D6 and D9, n = 8 for D5, n = 10 for D7 and D8). i, Representative images of Oil-Red-O-stained sections of small intestinal villi at ZT10 (D0L–D9L) or ZT15 (D4D, day 4 at dark phase) in SDR(10–25–65%) group. j, TG content of small intestinal tissues in i (n = 7 biologically independent samples for D0, n = 6 for D1, D6 and D9, n = 8 for D2, D3, D4D and D4). k, Post-gavage serum TG levels in mice fed AL or in the SDR(10–25–65%) group. n = 7, 8, 7, 8 and 8 biologically independent samples for AL at 0, 1, 2, 4, 6 h, respectively, n = 8, 8, 7, 7, 8 for SDR(10–25–65%). l, Fecal BODIPY concentrations of mice fed AL or in the SDR(10–25–65%) group after oral gavage with olive oil containing BODIPY-labeled fatty acid for 10 min to 2 h at D5 (n = 9 biologically independent mice per group). m, Representative fluorescent section images of small intestinal villi from mice fed AL or in the SDR(10–25–65%) group after oral gavage as in l for 2 h at D5. n,o, Small intestine (n) and serum (o) BODIPY concentrations of mice in m (n = 9 biologically independent mice per group). p, Representative fluorescent section images of iWAT and eWAT from mice in m. q, BODIPY concentrations of iWAT and eWAT in p (n = 9 biologically independent mice per group). r,s, Respiratory exchange ratio (RER) of mice fed AL or mice in SDR(10–25–65%) (r) or SDR(33.3% × 3) (s) group. n = 7 and 6 biologically independent mice for AL in r and s, respectively, n = 8 for SDR(10–25–65%), n = 6 for SDR(33.3% × 3). t,u, Food intake of mice fed AL or mice in the SDR(10–25–65%) (t) or SDR(33.3% × 3) (u) group (n = 8 biologically independent mice per group). The data shown in e,f were performed simultaneously with a single control experiment. Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test. a, *P < 0.05; b, **P < 0.01; c, ***P < 0.001; NS, not significant; versus AL or D0 group. Exact P values can be found in Source Data Fig. 2.

Source data

Then we investigated whether food intake was involved in quick fat mass accumulation after SDR. As shown in Fig. 2c, refeeding after SDR significantly increased hypothalamic messenger RNA levels of orexigenic genes Agrp and Npy and decreased hypothalamic mRNA levels of anorexigenic genes Pomc and Cart. Consistently, we found that food intake during the first several days of refeeding after dieting is markedly increased (Extended Data Figs. 1e and 3f,l,q). To determine whether increased food intake contributes to the quick fat mass accumulation after SDR, we analyzed cumulative food intake. Notably, cumulative food intake of mice for all the tested types of dieting was similar to or even lower than that of mice fed ad libitum (Fig. 2d and Extended Data Fig. 4a–d), indicating that cumulative food intake has no significant positive contribution to quick fat mass accumulation after SDR. To investigate the effect of increased food intake during the first several days of refeeding after SDR on fat mass, after feeding with 10–25–65% food or 33.3% × 3 food for 3 d, we fed the mice with 100% food per day as consumed just before SDR (Extended Data Fig. 4e,f). Similarly, refeeding with 100% food after SDR also led to quick fat mass accumulation (Fig. 2e,f and Extended Data Fig. 4g–j), although the cumulative food intake of mice in these two SDR groups was significantly lower than that of the mice fed ad libitum (Fig. 2g). These data demonstrate that the increased food intake during the first several days of refeeding is not the major cause of quick fat mass accumulation after SDR.

To further investigate the mechanism of quick fat mass accumulation after SDR, we analyzed whether increased intestinal lipid absorption was involved. We found that fecal triglyceride (TG) of mice fed with 10–25–65% food for 3 d was significantly decreased (Fig. 2h). Moreover, both Oil-Red-O-stained sections of small intestinal villi and direct quantification of small intestinal TG show that small intestinal TG was markedly increased after SDR (Fig. 2i,j). Gavage of olive oil and measurement of serum TG levels also suggest enhanced intestinal lipid absorption after SDR (Fig. 2k). To clearly monitor the intestinal lipid absorption, gavage of BODIPY-labeled fatty acid combined with olive oil was performed. The fecal BODIPY levels dramatically decreased after SDR (Fig. 2l) and the fluorescence intensity of villus sections from proximal jejunum was significantly increased after SDR (Fig. 2m), which is further confirmed by direct quantification of intestinal and serum BODIPY levels (Fig. 2n,o). These data show that enhanced intestinal lipid absorption contributes to the quick fat mass accumulation after SDR.

To further understand how increased intestinal lipid absorption finally leads to quick fat mass accumulation after SDR, we analyzed the fatty acid uptake and lipid synthesis in WAT. Gavage of BODIPY-labeled fatty acid showed that the fluorescence intensity of iWAT and eWAT sections significantly increased after SDR (Fig. 2p), which is further confirmed by direct quantification of BODIPY levels (Fig. 2q). The results from gavage of BODIPY-labeled fatty acid demonstrate that refeeding after SDR increases fatty acid uptake in WAT. Furthermore, fatty acid and TG synthesis significantly increased in iWAT after SDR when analyzed by high-throughput RNA sequencing (Extended Data Fig. 4k). Moreover, an increase in the respiratory exchange ratio after SDR was observed (Fig. 2r,s and Extended Data Fig. 4l,m), indicating a decreased total lipid oxidation in mice after SDR52. In addition, food intake in both dark and light phase significantly increased after SDR (Fig. 2t,u and Extended Data Fig. 4n,o). These data demonstrate that increased fatty acid uptake and lipid synthesis in WAT and decreased total lipid oxidation contribute to the quick fat mass accumulation after SDR.

To seek potential dietary interventions to prevent fat mass increase after dieting, we measured serum metabolites by non-targeted metabolomics before, during and after SDR. Differential serum metabolites were majorly concentrated on protein digestion and absorption, amino acid metabolism, biosynthesis of amino acids and related pathways (Fig. 3a,b) and the levels of most amino acids were obviously changed (Extended Data Fig. 5a). Further quantification of serum amino acids by liquid chromatography−tandem mass spectrometry (LC–MS/MS) showed that essential amino acids (EAAs) except methionine and some semi-essential and non-essential amino acids were significantly increased during and after SDR (Fig. 3c and Extended Data Fig. 5b). To further explore the potential effect of amino acids on quick fat mass accumulation after SDR, we fed mice with an HP diet, low-protein diet or normal-protein (NP) diet supplemented with EAAs after feeding with 10–25–65% food for 3 d (Fig. 3d). As shown in Fig. 3e and Extended Data Fig. 5c–f, all three diets fed after SDR prevented quick fat mass accumulation and especially the HP diet even partially maintained the fat loss induced by SDR. The effect of feeding with the HP diet after SDR on fat mass was further confirmed by the size and weight of iWAT and eWAT (Fig. 3f,g and Extended Data Fig. 5g–i). H&E staining also revealed that feeding with an HP diet after SDR markedly blocked the increase in adipocyte size (Fig. 3h and Extended Data Fig. 5j–l). Consistently, refeeding with an HP diet after the SDR markedly blocked the hypertrophy of iBAT and the accumulation of enlarged lipid droplets in iBAT cells (Extended Data Fig. 5m–o) and inhibited hepatic steatosis (Extended Data Fig. 5p,q).

a,b, KEGG pathway enrichment analysis based on the significantly altered metabolites in serum of mice in SDR(10–25–65%) group (n = 4 biologically independent mice per group). c, Serum essential amino acid levels of mice in SDR(10–25–65%) group (n = 4 biologically independent mice for D0 and D4, n = 3 for D2). d, Experimental design of mice refeeding with different diets after feeding with 10%, 25% and 65% NP diet provided from day 1 to day 3. LP, low-protein diet. e, An HP diet maintained fat mass loss induced by SDR with 10%, 25% and 65% food intake from day 1 to day 3. n = 18 biologically independent mice for AL-NP and SDR-NP, n = 9 for other groups. f, Representative images of isolated iWAT from mice as treated with NP or HP diet after SDR in e. g, The weight of iWAT from mice in f (n = 10 biologically independent mice for D0, n = 6 for other groups). h, Representative images of H&E-stained sections of iWAT in f. i,j, HP diet maintained fat mass loss induced by SDR with 33.3% food intake daily for 3 d (i) or DR with 66.7% food intake daily for 12 d (j). n = 9 and 8 biologically independent mice for AL-NP in i and j, respectively, n = 9 for other groups. k, HP diet maintained fat mass loss in female mice induced by SDR with 10%, 25% and 65% food intake from day 1 to day 3 (n = 10 biologically independent mice per group). Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test. a, *P < 0.05; b, **P < 0.01; c, ***P < 0.001; except indicated, SDR-HP group versus AL-NP or SDR-NP group, DR-HP group versus AL-NP or DR-NP group. Exact P values can be found in Source Data Fig. 3.

Source data

To further confirm the reverse effect of HP diet on fat gain after dieting, we fed mice with HP diet after feeding with 33.3% food daily for 3 d or 66.7% food daily for 12 d. As expected, feeding with HP diet after these two types of dieting prevented quick fat mass accumulation after dieting and even partially maintained the fat-loss effect of dieting (Fig. 3i,j and Extended Data Fig. 6a–h). In addition, feeding with HP diet after feeding with 10%–25%–65% food for 3 d in female mice also had a similar effect (Fig. 3k and Extended Data Fig. 6i–l).

These data show that fat mass increase after dieting can be prevented by dietary intervention, especially in the form of an HP diet.

To investigate the mechanism of how HP diets prevent quick fat mass accumulation after SDR, we analyzed the energy expenditure. As shown in Fig. 4a, energy expenditure in mice fed HP diet after SDR was significantly lower than that of mice fed NP diet after SDR. These results show that the inhibitory effect of the HP diet on quick fat mass accumulation after SDR was not due to the increase in energy expenditure.

a, Energy expenditure (EE) of mice fed ad libitum (AL) (n = 4 biologically independent mice) or fed with NP (n = 6) or HP (n = 6) diet after SDR with 10%, 25% and 65% food from day 1 to day 3. b, Cumulative food intake of mice as treated in a (AL-NP, n = 18 biologically independent mice; SDR-NP, n = 18; SDR-HP, n = 9). c, Experimental design of pair-feeding studies. Mice were pair-fed to the amount of daily caloric intake consumed by those in SDR-HP group. HF, high-fat diet; PF, pair-feeding. d, Daily caloric intake of mice in different regimens in c. n = 11 biologically independent mice for AL-NP, SDR-NP and SDR-NP(PF), n = 10 for SDR-HP and SDR-HF(PF). e, Pair-feeding with NP diet or high-fat diet after SDR markedly increased body fat mass. n = 11 mice for AL-NP, SDR-NP and SDR-NP(PF), n = 10 for SDR-HP and SDR-HF(PF). f, Fecal TG of mice as treated in a (n = 10, 7, 10, 9, 9, 9 and 8 biologically independent samples for SDR-NP on D0, D4, D5, D6, D7, D8 and D9, respectively, n = 10, 7, 9, 8, 9, 10 and 10 for SDR-HP). g,h, LCFA composition of cecal feces from mice as treated in a (n = 10 biologically independent mice per group). i,j, Total intestinal (i) and serum (j) bile acid concentrations of mice as treated in a. n = 8 and 10 biologically independent samples for AL-NP(D0) in i and j, respectively, n = 4 and 5 for SDR-NP(D4) or SDR-HP(D4) in i and j, respectively. k–n, Intestinal and serum primary (k,l) and secondary (m,n) bile acid composition from mice in (i,j). n = 8 and 10 biologically independent samples for AL-NP(D0) in k,m and l,n, respectively, n = 4 and 5 for SDR-NP(D4) or SDR-HP(D4) in k,m and l,n, respectively. o, Fecal BODIPY concentrations of mice in AL-NP (n = 7 biologically independent mice), SDR-NP (n = 9) and SDR-HP (n = 9) groups after oral gavage with olive oil containing BODIPY-labeled fatty acid for 10 min to 2 h at D5. p, Representative fluorescent section images of small intestinal villi from mice in o after oral gavage for 2 h. q,r, Small intestine (q) and serum (r) BODIPY concentrations of mice in p. n = 7 and 8 biologically independent mice for AL-NP(D5) in q and r, respectively, n = 8 and 7 for SDR-NP(D5), n = 9 and 8 for SDR-HP(D5). s, Representative fluorescent section images of iWAT and eWAT from mice in o after oral gavage for 2 h. t, BODIPY concentrations of iWAT and eWAT in s (n = 9 biologically independent mice per group). u,v, Respiratory exchange ratio (RER) (u) and average RER (v) of mice as treated in a. n = 4 biologically independent mice for AL-NP, n = 6 for SDR-NP and SDR-HP. Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test (a,b,e–h,o,q,r,t,v) and one-tailed Student's t-test (i–n). a, *P < 0.05; b, **P < 0.01; c, ***P < 0.001; NS, not significant; except indicated, SDR-HP group versus AL-NP or SDR-NP group, SDR-NP(PF) or SDR-HF(PF) group versus SDR-HP group. Exact P values can be found in Source Data Fig. 4.

Source data

Then we investigated whether food intake was involved in blocking quick fat mass accumulation after dieting by HP diet. As shown in Fig. 4b and Extended Data Fig. 6m–o, cumulative food intake of mice in the HP group significantly decreased during the refeeding period compared to that of mice in the NP groups no matter with or without dieting; however, the daily food intake after dieting in the HP group was similar to that of the NP group without dieting (Extended Data Fig. 6p–s). As shown in Fig. 2e,f and Extended Data Fig. 4e–j, refeeding with 100% food after SDR also led to a quick fat mass accumulation, so therefore the decrease of food intake during the first several days of refeeding induced by the HP diet is unlikely to be the major reason preventing quick fat mass accumulation after SDR. To further confirm whether calorie intake and/or diet composition was involved, we performed a pair-feeding experiment with NP or high-fat diet containing the same daily calorie consumed by the mice refeeding with an HP diet after SDR (Fig. 4c,d and Extended Data Fig. 7a,b). As shown in Fig. 4e and Extended Data Fig. 7c–f, pair-feeding with NP or high-fat diet after SDR induced obvious fat mass increase, which was even more severe than that in mice refeeding ad libitum with NP diet during the late refeeding period (Fig. 4e). These data show that the difference in diet composition rather than that in caloric intake has the major contribution for blocking fat mass increase after dieting by HP diet.

To further investigate how the HP diet blocks fat mass increase after SDR, we analyzed whether intestinal lipid absorption was involved. We found that both fecal TG and cecal fecal total long-chain fatty acids (LCFAs) of mice fed the HP diet after SDR significantly increased compared to the mice fed NP diet after SDR (Fig. 4f,g), suggesting that the HP diet leads to a decrease in intestinal lipid absorption. Profiling of LCFAs in cecal feces showed that refeeding after SDR reduced the levels of 15 LCFAs in all 22 detected LCFAs and the reduction of six LCFAs including the most abundant C16:0 and C18:0 induced by refeeding after SDR were blockable by an HP diet (Fig. 4h). Bile acids are known to facilitate digestion and absorption of dietary lipids53,54. Consistently, the increase of both intestinal and serum total bile acids in mice refeeding with NP diet after SDR was observed compared to those in mice fed ad libitum and the increase was repressed in mice refeeding with an HP diet after SDR (Fig. 4i,j). Both primary and secondary bile acids had a similar alteration as total bile acids in mice refeeding with NP and HP diet after SDR (Fig. 4k–n). Furthermore, gavage of BODIPY-labeled fatty acid combined with olive oil showed that fecal BODIPY levels significantly increased in the HP group (Fig. 4o). Moreover, the fluorescence intensity of villus sections from proximal jejunum in mice gavaged with BODIPY-labeled fatty acid significantly decreased in the HP group compared to that of the NP group with SDR (Fig. 4p), which is further confirmed by direct quantification of the intestinal and serum BODIPY levels (Fig. 4q,r). These data show that the decrease in intestinal lipid absorption induced by the HP diet contributes to the prevention of fat mass increase after SDR.

To further understand how decreased intestinal lipid absorption finally blocks the fat mass accumulation after SDR, we analyzed the fatty acid uptake and lipid synthesis in WAT. Gavage of BODIPY-labeled fatty acid showed that the fluorescence intensity of iWAT and eWAT sections was significantly decreased in the HP diet group compared to the mice fed NP diet after SDR (Fig. 4s), which is further confirmed by direct quantification of BODIPY levels (Fig. 4t). The results from gavage of BODIPY-labeled fatty acid demonstrate that the HP diet after SDR induces a decrease in lipid uptake in WAT. Furthermore, fatty acid and TG synthesis was significantly decreased in the iWAT of mice in the HP diet group when analyzed by high-throughput RNA sequencing (Extended Data Fig. 7g). In addition, the respiratory exchange ratio of mice fed with an HP diet after SDR was dramatically decreased (Fig. 4u,v), indicating that the HP diet increased total lipid oxidation in mice. These data demonstrate that decreased fatty acid uptake and lipid synthesis in WAT and increased total lipid oxidation induced by HP diet contribute to the prevention of fat mass increase after SDR.

Diet is a key determinant of the gut microbiota configuration42 and microbiota can affect intestinal lipid absorption and lipid metabolism in adipose tissues49,55,56. To investigate whether microbiota are involved in the quick fat mass accumulation and increased intestinal lipid absorption induced by refeeding after SDR, we collected cecal feces of mice refed after SDR with 10%, 25% and 65% food from day 1 to day 3 and analyzed the microbiota composition by 16S rRNA gene sequencing. To assess how refeeding after SDR affects the gut microbial community structure, we investigated the α and β diversity associated with each group and compared the microbial diversity within and between communities. Principal-coordinate analysis revealed distinct clustering between microbiota from mice fed ad libitum, NP diet after SDR and HP diet after SDR (Fig. 5a). Moreover, bacterial α diversity analyzed by Shannon index was reduced after refeeding with NP diet after SDR on day 4 and was returned to normal level for the mice fed an HP diet after SDR on day 4 (Fig. 5b). Next, we calculated the percentage of bacterial taxa in each group. We observed a dramatic increment up to approximately 50% in the class Bacilli, the family Lactobacillaceae and the genus Lactobacillus during refeeding with NP diet after SDR on day 4 compared to feeding ad libitum (Fig. 5c–e); however, the relative abundance of Bacilli, Lactobacillaceae and Lactobacillus was clearly lower in mice fed an HP diet than those of mice fed a NP diet after SDR on day 4 (Fig. 5c–e). Analysis with the linear discriminant analysis (LDA) effect size (LEfSe) method revealed that the mice refeeding with NP diet after SDR was characterized by the genus Lactobacillus, which also belongs to the characterized phylum Firmicutes, class Bacilli, order Lactobacillales and family Lactobacillaceae (Fig. 5f). Songbird analysis also showed that the highest ranked genus based on the log ratios between microbiota in the SDR-NP group on day 4 and microbiota in mice fed ad libitum was Lactobacillus (Fig. 5g). These data show that refeeding after SDR significantly changes the composition of gut microbiota and induces a dramatical enrichment of Lactobacillus.

a, Principal-coordinate analysis (PCoA) based on the weighted UniFrac of microbiota composition of cecal feces in mice fed ad libitum (AL) with NP diet (n = 14 biologically independent mice) or fed with NP diet (n = 7) or HP diet (n = 7) after SDR with 10%, 25% and 65% food from day 1 to day 3 at the indicated time points. Each dot represents an individual mouse. PCoA1 and PCoA2 represent the percentage of variance explained by each coordinate. P = 0.001. b, Comparison of α-diversity of microbiota in a using Shannon index (n = 14 mice for AL-NP, n = 7 for SDR-NP or SDR-HP). c,d, Class-level (c) and family-level (d) proportional abundance of microbiota in a (n = 14 mice for AL-NP, n = 7 for SDR-NP or SDR-HP). e, Genus-level proportional abundance of Lactobacillus in microbiota from a (n = 7 biologically independent mice per group). f, Cladogram generated from LEfSe and LDA score showing the most differentially significant abundant taxa enriched in microbiota from mice in a. p, phylum; c, class; o, order; f, family; g, genus. g, Songbird analysis of genera from mice in AL-NP(D0) and SDR-NP(D4) groups. The top and bottom five ranked features are highlighted with the indicated color. h, Phylogenetic tree of the isolated L. murinus Lam-1 with their relatives based on 16S rRNA gene sequences. The bar indicates sequence divergence. i, The abundance of Lam-1 determined by qPCR of 16S rDNA in cecal feces of mice as treated in a (n = 8 biologically independent mice per group). j,k, The 16S rDNA copies of Lam-1 (j) and total bacterial load (k) quantified by qPCR in cecal feces in i (n = 8 biologically independent mice per group). l,m, Effect of the indicated antibiotics on the growth of Lam-1 (n = 3 independent experiments per group and in triplicate for each experiment). PURO, puromycin; NEO, neomycin; MTZ, metronidazole; KAN, kanamycin; VAN, vancomycin; CHL, chloramphenicol; TET, tetracycline; AMP, ampicillin; PEN, penicillin. n, Experimental design of mice treated with or without penicillin given in drinking water. Arrows show the time points of feces collection, BODIPY-labeled fatty acid gavage and sampling. o,p, The abundance of Lam-1 (o) and total bacterial load (p) in feces of mice from regimens in n (n = 7 and 8 biologically independent samples for SDR on D0 and D4 respectively, n = 9 for SDR-PEN). q, Representative fluorescent section images of small intestinal villi of mice from regimens in n with or without penicillin treatment after oral gavage with olive oil containing BODIPY-labeled fatty acid for 2 h at D5. r, BODIPY concentrations of small intestine in q (n = 8 biologically independent mice for SDR, n = 9 for SDR-PEN). s, Representative fluorescent section images of iWAT and eWAT from mice in q. t, BODIPY concentrations of iWAT and eWAT in s (n = 8 mice for SDR, n = 9 for SDR-PEN). u, Penicillin treatment ameliorated fat mass accumulation induced by SDR with 10%, 25% and 65% food from day 1 to day 3 (n = 10 biologically independent mice for SDR, n = 12 for SDR-PEN). v, Daily food intake of mice in u (n = 10 mice for SDR, n = 12 for SDR-PEN). Data are presented as mean ± s.d. Statistical significance was determined by permutation multivariate analysis of variance (a), Kruskal–Wallis and Dunn's tests (b) and two-tailed Student's t-test (e,i–k,m,o,p,r,t–v). a, *P < 0.05; b, **P < 0.01; ***P < 0.001; NS, not significant. Exact P values can be found in Source Data Fig. 5.

Source data

To investigate whether the predominant Lactobacillus bacteria with an abundance up to about 50% during refeeding after SDR contribute to the increase in intestinal lipid absorption and fatty acid uptake in WAT, we isolated and sequenced eight colonies from the cecal feces of mice in the SDR(10–25–65%) group on day 4. These colonies had the identical 16S rRNA gene sequence and phylogenetic analysis on the basis of 16S rRNA gene sequences show that Lactobacillus murinus (NR_112689) is the closest strain type to the colonies (Fig. 5h). One of the colonies was randomly picked and termed Lam-1 for further investigation. Quantification of Lam-1 by qPCR further confirmed that refeeding with a NP diet after SDR dramatically increased the abundance of Lam-1 and this increase was sufficiently suppressed when refed with an HP diet (Fig. 5i,j). In addition, the total bacterial load of cecal feces was similar in all the mice with or without SDR (Fig. 5k).

To investigate the role of gut microbiota in the quick fat mass accumulation induced by refeeding after SDR, we treated the mice with antibiotic mixture during the refeeding period after SDR (Extended Data Fig. 8a). The antibiotic-induced microbiome depletion was confirmed by gel electrophoresis of fecal DNA and quantification with qPCR (Extended Data Fig. 8b,c). Expectedly, gavage of BODIPY-labeled fatty acid showed that intestinal fatty acid absorption was significantly reduced in mice treated with an antibiotic mixture after SDR (Extended Data Fig. 8d–f). Meanwhile, fatty acid uptake in WAT was significantly decreased in mice treated with an antibiotic mixture after SDR (Extended Data Fig. 8g,h). Furthermore, treatment with an antibiotic mixture significantly ameliorated the fat mass accumulation induced by refeeding after SDR with 10%, 25% and 65% food from day 1 to day 3 (Extended Data Fig. 8i) and also repressed the increase of food intake and fat mass percentage and the decrease of lean mass percentage (Extended Data Fig. 8j–n). It has been reported that antibiotic-induced microbiome depletion alters metabolic homeostasis, including the increase of GLP-1 and decrease of luminal bile acids, suggesting the reduction of intestinal absorption of lipids57. To exclude the effect of antibiotic-induced microbiome depletion, we detected antibiotic susceptibility of Lam-1 and found that penicillin specifically inhibited the growth of Lam-1 at the concentration as low as 10 ng ml−1 (Fig. 5l,m). We next treated mice with penicillin 1 week before SDR (Fig. 5n). Penicillin treatment dramatically inhibited the growth of Lam-1 without affecting the total bacterial load (Fig. 5o,p) and significantly reduced intestinal fatty acid absorption and fatty acid uptake in iWAT of mice during refeeding after SDR (Fig. 5q–t). Moreover, penicillin treatment significantly ameliorated the fat mass accumulation induced by refeeding after SDR and also repressed the increase of fat mass percentage and the decrease of lean mass percentage without altering food intake (Fig. 5u,v and Extended Data Fig. 8o–t). These data suggest that refeeding after SDR increases fat mass accumulation, intestinal lipid absorption and fatty acid uptake in WAT via gut microbiota.

To further investigate the effect of gut microbiota on intestinal absorption and fatty acid uptake in WAT, we transferred cecal feces from donor mice fed ad libitum or fed with NP or HP diet after SDR with 10%, 25% and 65% food from day 1 to day 3 to specific-pathogen-free (SPF) recipient mice (Extended Data Fig. 9a). As expected, intestinal lipid absorption and fatty acid uptake in eWAT of recipients transferred with microbiota from the SDR-NP group were significantly increased, which was attenuated in recipients transferred with microbiota from the SDR-HP group (Fig. 6a–e and Extended Data Fig. 9b).

a, Fecal BODIPY concentrations of fecal microbiota transplantation (FMT)-recipient mice from regimens in Extended Data Fig. 9a after oral gavage with olive oil containing BODIPY-labeled fatty acid for 10 min to 2 h (n = 6 biologically independent mice per group). Mice received FMT from donor mice in the AL-NP, SDR-NP or SDR-HP group. AL, ad libitum. b, Representative fluorescent section images of small intestinal villi and BODIPY concentrations of small intestine from mice in a after oral gavage for 2 h (n = 6 mice per group). c, Serum BODIPY concentrations of mice in b (n = 6 mice per group). d, Representative fluorescent section images of iWAT and eWAT from mice in b. e, BODIPY concentrations of iWAT and eWAT in d (n = 6 mice per group). f, Experimental design of germ-free (GF) mice gavaged with Lactobacillus LC2W or Lactobacillus Lam-1. Arrows indicate the time points of gavage and sampling. g, Fecal BODIPY concentrations of mice from regimens in f after oral gavage with olive oil containing BODIPY-labeled fatty acid for 10 min to 2 h at D5 (n = 8 biologically independent mice per group). h, Representative fluorescent section images of small intestinal villi and BODIPY concentrations of small intestine from mice in g after oral gavage for 2 h (n = 8 mice per group). i, Serum BODIPY concentrations of mice in h (n = 8 mice per group). j, Representative fluorescent section images of iWAT and eWAT from mice in h. k, BODIPY concentrations of iWAT and eWAT in j (n = 8 mice per group). l, The weight of iWAT and eWAT in j (n = 8 mice per group). m, Schematic of the gnotobiotic (GB) mouse model illustrating the timeline of gavage with eight bacterial strains and confirmation of colonization in GF mice for subsequent LC2W or Lam-1 gavage as indicated in f. n, Fecal BODIPY concentrations of GB mice from regimens in m after oral gavage with olive oil containing BODIPY-labeled fatty acid for 10 min to 2 h at D5 (n = 8 biologically independent mice per group). o, Representative fluorescent section images of small intestinal villi and BODIPY concentrations of small intestine from mice in n after oral gavage for 2 h (n = 9 mice for GB + LC2W, n = 8 for GB + Lam-1). p, Serum BODIPY concentrations of mice in o (n = 9 mice per group). q, Representative fluorescent section images of iWAT and eWAT from mice in o. r, BODIPY concentrations of iWAT and eWAT in q (n = 9 mice per group). s, The weight of iWAT and eWAT in q (n = 9 mice per group). t, Fecal BODIPY concentrations of SPF mice gavaged with LC2W or Lam-1 after intragastric administration of olive oil containing BODIPY-labeled fatty acid for 10 min to 2 h (n = 8 biologically independent mice per group). u, Representative fluorescent section images of small intestinal villi and BODIPY concentrations of small intestine from mice in t after BODIPY gavage for 2 h (n = 8 mice per group). v, Serum BODIPY concentrations of mice in u (n = 8 mice per group). w, Representative fluorescent section images of iWAT and eWAT from mice in u. x, BODIPY concentrations of iWAT and eWAT in w (n = 8 mice per group). y, Oral gavage of Lam-1 markedly increased body fat mass (n = 9 biologically independent mice for SPF + vehicle and SPF + LC2W, n = 8 for SPF + Lam-1). z, Cumulative food intake of mice in y (n = 9 mice for SPF + vehicle and SPF + LC2W, n = 8 for SPF + Lam-1). Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test. *P < 0.05; **P < 0.01; ***P < 0.001; NS, not significant. Exact P values can be found in Source Data Fig. 6.

Source data

To determine whether L. murinus Lam-1 plays a causal role in the increase in intestinal lipid absorption, fatty acid uptake in WAT and fat mass accumulation, we gavaged germ-free mice with 2 × 109 colony-forming units (c.f.u.) Lam-1 or Lactobacillus paracasei LC2W for three consecutive days followed by SDR with 10%, 25% and 65% food from day 1 to day 3 (Fig. 6f). We found that Lam-1 gavage significantly increased intestinal lipid absorption and fatty acid uptake in WAT of germ-free mice during refeeding after SDR compared to LC2W gavage (Fig. 6g–k). Meanwhile, Lam-1 gavage significantly upregulated the weight of eWAT in germ-free mice during refeeding after SDR (Fig. 6l).

To further confirm the causal role for Lam-1 in upregulation of intestinal absorption, fatty acid uptake and fat mass accumulation, we used a gnotobiotic mouse model. Gnotobiotic mice were obtained from germ-free mice gavaged with eight bacterial strains representing members of the major bacterial phyla in murine gut and the colonization of bacterial strains in gnotobiotic mice was confirmed by 16S rDNA sequencing (Fig. 6m and Extended Data Fig. 9c). Then we gavaged the gnotobiotic mice with 2 × 109 c.f.u. Lam-1 or LC2W for three consecutive days followed by SDR with 10%, 25% and 65% food from day 1 to day 3. We found that Lam-1 gavage significantly increased intestinal lipid absorption and fatty acid uptake in iWAT of gnotobiotic mice during refeeding after SDR compared to LC2W gavage (Fig. 6n–r). Meanwhile, Lam-1 gavage significantly upregulated the weight of iWAT and eWAT in gnotobiotic mice during refeeding after SDR (Fig. 6s).

We next investigated the effects of Lam-1 administration on intestinal lipid absorption, fatty acid uptake in WAT and fat mass accumulation of SPF mice. Mice were gavaged orally with vehicle or 1010 c.f.u. LC2W or Lam-1 twice with an interval of 24 h. Following the last gavage, we gavaged the mice with BODIPY-labeled fatty acid. We found that Lam-1 gavage significantly increased intestinal lipid absorption and fatty acid uptake in WAT of SPF mice compared to vehicle or LC2W gavage (Fig. 6t–x and Extended Data Fig. 9d–i). Meanwhile, daily gavage of Lam-1 significantly induced fat mass accumulation in SPF mice without altering food intake (Fig. 6y,z and Extended Data Fig. 9j–m).

Collectively, the above studies indicate that Lam-1 enriched by refeeding after SDR can enhance intestinal lipid absorption, increase fatty acid uptake in WAT and induce fat mass accumulation.

Given the effectiveness of depletion of Lam-1 by penicillin and Lam-1 administration, we sought to gain further insight to how gut microbiota orchestrates intestinal lipid absorption and fatty acid uptake in WAT during refeeding after SDR. To this end, we compared the cecal fecal non-targeted metabolomics profiles between the mice fed ad libitum and the mice fed NP and HP diets after SDR with 10%, 25% and 65% food from day 1 to day 3. As expected, refeeding with NP diet after SDR induced significant alterations in the cecal fecal metabolome compared to that in mice feeding ad libitum and these alterations were partially blocked by refeeding with HP diet after SDR (Fig. 7a). Among the metabolites in the blockable-increase group, DL-3-phenyllactic acid (PLA), 4-hydroxyphenyllactic acid (HPLA), 2-hydroxyisocaproic acid (HICA), 2-hydroxy-3-methylbutyric acid (HMBA) and indole-lactic acid (ILA) are the top five significantly increased metabolites induced by refeeding with NP diet after SDR (Fig. 7a). Furthermore, the alterations of these five metabolites between different groups were further confirmed by targeted quantitative analysis using LC–MS/MS with their standards (Fig. 7b–f). Spearman's rank correlation analysis showed that Lactobacillus was most strongly associated with cecal fecal levels of PLA, HPLA, ILA, HICA and HMBA (Fig. 7g,h). Moreover, gavage with Lactobacillus Lam-1 also significantly elevated these five metabolites in cecal feces (Fig. 7i). To further confirm whether Lam-1 produces these five metabolites, we detected these five metabolites in bacterial lysates and culture supernatant. We found that Lam-1 produced much higher amounts of PLA, HPLA, ILA, HICA and HMBA than Escherichia coli and much higher amounts of PLA, HPLA and ILA than Lactobacillus LC2W (Fig. 7j,k). These data demonstrate that refeeding after SDR induces HP-diet partially blockable alterations of cecal fecal metabolites, such as the increase of PLA, HPLA, ILA, HICA and HMBA, which can also be produced by Lam-1 and upregulated by gavage with Lam-1.

a, Heat map of intestinal metabolites from mice fed ad libitum (AL) (n = 6 biologically independent samples) or fed with NP diet (n = 3 samples) or HP diet (n = 3 samples) after SDR with 10%, 25% and 65% food from day 1 to day 3. b–f, Concentrations of cecal fecal PLA (b), HPLA (c), ILA (d), HICA (e) and HMBA (f) from mice in a (n = 4 biologically independent samples per group). g,h, Heat map of Spearman's rank correlation coefficients (two-sided tests) between the relative abundances of gut microbiota genera and the relative concentrations of PLA, HPLA, ILA, HICA and HMBA in cecal feces from mice in SDR-NP(D4) and AL-NP(D0) groups (g) or from mice in SDR-HP(D4) and SDR-NP(D4) groups (h). i, Concentrations of cecal fecal metabolites after oral gavage with vehicle or Lam-1 (n = 9 biologically independent samples per group). j,k, Production of PLA, HPLA, ILA, HICA and HMBA by E. coli, Lactobacillus LC2W or Lactobacillus Lam-1. E. coli was cultured in LB medium and LC2W or Lam-1 was cultured in MRS medium. Metabolites were measured in bacterial lysates (j) or culture supernatant (k) (n = 5 biologically independent samples per group). l, Fecal BODIPY concentrations of mice administrated intragastrically with water or solution containing five compounds (compds), including PLA, HPLA, ILA, HICA and HMBA after oral gavage with olive oil containing BODIPY-labeled fatty acid for 10 min to 2 h (n = 10 biologically independent mice per group). m, Representative fluorescent section images of small intestinal villi from mice in l after BODIPY gavage for 2 h. n, BODIPY concentrations of small intestine in m (n = 10 mice per group). o, Representative fluorescent section images of iWAT and eWAT from mice in m. p, BODIPY concentrations of iWAT and eWAT in (o) (n = 10 mice per group). q, Oral gavage of solution containing the five compounds markedly increased body fat mass (n = 8 biologically independent mice for water, n = 9 for compounds). r, Cumulative food intake of mice in q (n = 8 mice for water, n = 9 for compounds). s, Schematic diagram of refeeding after DR-induced obesity via microbiota and metabolite-mediated increased intestinal lipid absorption and fatty acid uptake in WAT. Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test. * or #P < 0.05; ** or ##P < 0.01; ***P < 0.001; NS, not significant. #The SDR-NP(D4) or SDR-NP(D6) group was compared to the SDR-NP(D0) group. *SDR-HP group was compared to SDR-NP group. Exact P values can be found in Source Data Fig. 7.

Source data

To investigate whether the alterations of metabolites contribute to the enhanced intestinal lipid absorption, increased fatty acid uptake in WAT and fat mass accumulation, we planned to treat mice with the top altered metabolites. Among the top five significantly increased metabolites in blockable-increase group, PLA, HPLA and ILA can be derived from ingested dietary phenylalanine, tyrosine and tryptophan by intestinal microbiota through the reductive pathway using the same enzymes58. Thus, we first investigated the effects of these three metabolites on metabolism and found that neither PLA, HPLA, ILA alone nor their mixtures had significant effect on food intake, intestinal lipid absorption and fatty acid uptake of WAT in mice (Extended Data Fig. 10a–h). In addition, gavage with HICA or HMBA also among the top five significantly increased metabolites had no significant effect (Extended Data Fig. 10i–p); however, when mice were gavaged with the mixtures of PLA, HPLA, ILA, HICA and HMBA followed with BODIPY-labeled fatty acid gavage, the fecal BODIPY levels of mice were significantly decreased without altering food intake (Fig. 7l and Extended Data Fig. 10q). Moreover, the fluorescence intensity of villus sections from proximal jejunum was significantly increased in the mice gavaged with the five metabolites compared to that of mice gavaged with water (Fig. 7m), which is further confirmed by direct quantification of the intestinal BODIPY levels (Fig. 7n). In addition, fatty acid uptake in WAT was also measured and the fluorescence intensity of iWAT and eWAT sections was significantly increased in the mice gavaged with the five metabolites (Fig. 7o), which is further confirmed by direct quantification of BODIPY levels (Fig. 7p). To exclude the potential effects of calories in the five metabolites on intestinal lipid absorption and fatty acid uptake in WAT, we treated mice with glutamic acid and glucose unchanged during refeeding after SDR using the same dose as the total five metabolites. We found that neither glutamic acid nor glucose had significant effect on intestinal lipid absorption, fatty acid uptake of WAT and food intake in mice (Extended Data Fig. 10r–v). Moreover, we found that daily gavage of the five metabolites led to significant fat mass accumulation in mice without altering food intake (Fig. 7q,r and Extended Data Fig. 10w–z). These data show that the five metabolites produced by gut microbiota, such as Lactobacillus Lam-1, can upregulate intestinal lipid absorption, enhance fatty acid uptake in WAT and induce fat mass accumulation.

Here we show that refeeding after dieting promotes the enrichment of intestinal Lactobacillus and its metabolites, which can be blocked by HP diet or antibiotics, thereby enhances intestinal lipid absorption and fatty acid uptake in WAT and leads to increased fat mass gain (Fig. 7s).

In this study, we systematically investigated the effect of refeeding after ten types of dieting on fat mass and the data demonstrate that refeeding after dieting leads to quick fat mass accumulation, which may subsequently contribute to the development of obesity. Due to the prevalence of active and passive dieting, not only in people with overweight or obesity but also in people with normal weight15,16,59,60, it is imperative to comprehensively understand the detrimental effect of feeding ad libitum after dieting and to disseminate this information to the public. Consistent with our observation, it has been reported that weight loss can be achieved through DR in several human studies, but the individuals will usually regain their weight in the near future61,62,63,64,65. In addition, high-fat-diet-induced obesity of animals is widely used to study the mechanisms of obesity and to find strategies to combat obesity66. Our findings provide additional noteworthy and convenient animal models to study obesity, as well as provide an alternative strategy targeting refeeding after dieting to prevent obesity.

Mechanistically, we show that increased intestinal lipid absorption, rather than energy expenditure or increased food intake after SDR, contributes a lot to the fat mass increase after SDR. Targeting intestinal lipid absorption has been successfully applied in preclinical animal studies to treat diet-induced obesity67, which is also likely to be a promising strategy to combat obesity induced by refeeding after SDR. Increased fatty acid uptake and lipid storage in WAT has been reported in the process of weight regain68,69. Similarly, we clearly show that increased fatty acid uptake and lipid synthesis in WAT are involved.

We demonstrate that feeding with HP diet after dieting significantly prevents fat mass accumulation and even partially maintains the fat-loss effect induced by dieting, providing a potential practical way to prevent obesity after dieting. Consistently, it has been reported that HP diets have beneficial effects on weight control and metabolic syndrome70,71, suggesting that HP diet is applicable to prevent obesity after dieting or maintain the fat-loss effect induced by dieting.

Our studies further show that refeeding after SDR leads to a dramatic enrichment of L. murinus up to about 50% and its metabolites, which is repressible by HP diet and antibiotics. Notably, it has been reported that the abundance of intestinal L. murinus increased to 6.5% and about 22% after calorie restriction for 3 d and 2 weeks, respectively in mice72 and about 12–18% after calorie restriction for 3–8 weeks in rats33, suggesting that Lactobacillus may contribute to the increased utilization of nutrients under the condition of calorie restriction. Small intestine microbiota has been reported to regulate host absorptive adaptive responses to dietary lipids35. Consistently, we demonstrate that oral gavage of Lactobacillus Lam-1 or its metabolites significantly increased intestinal lipid absorption, enhanced fatty acid uptake in WAT and induced fat mass accumulation. Some Lactobacillus species function as probiotics and have been successfully used as alternative treatments for diarrhea, which occurs due to increased secretion, decreased absorption or both73,74,75, suggesting that Lactobacillus Lam-1 may also be considered as a probiotic to treat diarrhea. Moreover, targeting to inhibit intestinal Lactobacillus by a specific diet or antibiotics is likely to be beneficial to decrease intestinal lipid absorption and combat obesity. In addition, as a limitation of this study, our findings should be tested in murine models of obesity to further reinforce the implications in the context of dieting and obesity.

In summary, dieting still should be encouraged to reduce fat and to control body weight, but the detrimental effects of dieting termination to induce weight regain and obesity should be seriously considered. Taking specific post-dieting diets, such as HP diet, is likely an applicable strategy that can alleviate the detrimental effects of terminating dieting. In addition, our study suggests that inhibition of intestinal Lactobacillus may be an alternative approach to alleviate intestinal lipid absorption and reduce fat mass.

All animals were maintained and used in accordance with the guidelines of and under approval by, the Institutional Animal Care and Use Committee of Shanghai Institute of Nutrition and Health, Chinese Academy of Sciences (ethical committee approval nos. 2014-AN-5, SIBS-2017-ZQW-1, SINH-2021-ZQW-1 and SINH-2022-ZQW-1). Eight-week-old C57BL/6J SPF mice were purchased from Slac and acclimated to the animal facility for 3–5 d. Mice were housed in an SPF facility with a 12-h light–12-h dark cycle, an ambient temperature of 20–24 °C and humidity of 40–60%. Mice were individually housed for 5 d and subsequently housed individually for SDR, DR and refeeding experiments. During the 5-d individually housed period before SDR and DR, the indicated food used in SDR and DR was provided. Average daily food intake of mice in each group in the last 3 d before SDR and DR was measured as the standard daily food intake for the following experiments. SDR and DR was performed by providing the indicated amount of restricted food at ZT12. Normal chow diet was obtained from Shanghai Pu Lu Teng Biotechnology (P1103F). Composition of diets, except normal chow, used in this study is provided in Supplementary Table 1. Briefly, an NP diet containing 20% casein, 0.3% cystine and 49.7% corn starch was prepared according to AIN-93G Growing Rodent Diet (D10012G, Research Diets) except without tert-butylhydroquinone and using corn starch to replace fine granulated sucrose. An HP diet was prepared as the above NP diet except containing 60% casein, 0.9% cystine and 9.1% corn starch. An LP diet was prepared as the above NP diet except containing 5% casein, 0.075% cystine and 64.925% corn starch. An NP diet supplemented with EAAs was prepared as the above NP diet except containing 32.5% corn starch and 2.6% valine, 2.2% isoleucine, 3.32% leucine, 1% methionine, 1.8% phenylalanine, 0.44% tryptophan, 1.76% threonine, 2.96% lysine and 1.12% histidine. The percentage of each supplemented essential amino acid was calculated according to its additional content in the HP diet compared to the NP diet. A high-fat diet was prepared as the above NP diet except containing 26.668% casein, 0.396% cystine, 12.139% corn starch, 17.601% maltodextrin and 33.396% soybean oil. Unless otherwise stated, all animal experiments were performed with male mice.

Body weight and body composition of mice were measured at ZT10. Fat and lean mass of non-anesthetized live mice were detected using an EchoMRI-100H Body Composition Analyzer (EchoMRI) with EchoMRI 2016 software.

Body temperature of mice was measured at ZT3 using a RET-3 rectal probe (Physitemp) attached to a BAT-12 microprobe thermometer (Physitemp).

Metabolic rate and physical activity of the indicated mice were measured using a 16-chamber environment-controlled Comprehensive Lab Animal Monitoring System with Oxymax software (v.4.10) (Columbus Instruments). Mice were acclimatized to the cages for at least 24 h before the beginning of recording. RER refers to the volume of CO2 produced per volume of O2 consumed (both in l h−1). EE (kcal h−1) is calculated as 3.815 × VO2 + 1.232 × VCO2 (both in l h−1)76.

Mice were anesthetized at the indicated time points and tissues of interest were excised, weighed, photographed and then fixed in 4% paraformaldehyde or snap-frozen in liquid nitrogen immediately and subsequently stored at −80 °C. Except where indicated, blood was collected from the heart of mice using 1-ml syringes and centrifuged at 1,000g for 30 min at 4 °C. Serum was collected and frozen at −80 °C. Hypothalamus was obtained at ZT10 from the indicated mice. Feces were collected and stored at −80 °C. Cecal contents were collected and frozen in liquid nitrogen immediately and subsequently stored at −80 °C.

Tissues were fixed in 4% paraformaldehyde for 24–48 h at 4 °C, dehydrated, embedded into paraffin and sectioned (4 µm) for H&E staining. Adipocyte size quantification was performed after H&E staining using ImageJ software (v.1.51).

Total RNA of hypothalamus was extracted with TRIzol (Thermo Fisher, 15596018) and cDNA was synthesized with M-MLV Reverse Transcriptase (Promega, M1705). Quantitative PCR was performed with SYBR Green Master Mix (Applied Biosystems, A25778) using 7900HT Real-Time PCR System (Applied Biosystems). Relative RNA levels were normalized to Actin. The primers used are listed in Supplementary Table 2.

Tissues were fixed in 4% paraformaldehyde for 24–48 h at 4 °C, dehydrated, embedded in OCT compound (Sakura, 4583), sectioned (10 µm), stained with Oil-Red-O solution (Servicebio, G1015) for 10 min and counter-stained with hematoxylin.

Hepatic, small intestinal and fecal TG was extracted and detected as previously described with minor modifications77. Briefly, liver or proximal jejunum samples were collected and shredded into small pieces using surgical scissors. Feces were freeze-dried and ground up using a mortar and pestle. Then, 70–90 mg liver, 40–60 mg proximal jejunum or 100 mg dried and ground feces was homogenized in 1.2 ml of methanol/chloroform (1:2 v/v) containing five porcelain beads with a diameter of 2 mm using TissueLyser II (QIAGEN) at 30 Hz for 10 min. Then samples were centrifuged at 1,000g for 10 min and 700–800 μl of the supernatant was transferred to a new tube. Subsequently, 150 μl of 0.9% saline was added and the sample was vortexed and centrifuged at 1,000g for 10 min. Then 300–400 μl of the organic phase was transferred to a new tube and air-dried in a fume hood overnight at room temperature. Completely dried samples were finally dissolved in 200–300 μl ethanol. The TG content of these samples dissolved in ethanol or serum samples was determined with an enzymatic assay kit from Shensuo Unf Medical Diagnostics.

Mice received olive oil (15 μl g−1 body weight) or BODIPY 500/510 C1, C12 fatty acid (0.5 μg g−1 body weight, Molecular Probes, D3823) in olive oil (10 μl g−1 body weight) by intragastric gavage at ZT12. Mice had access to water but no food after gavage. For olive oil gavage, blood from the tail was collected before gavage and at 1, 2, 4 and 6 h after gavage to measure serum TG. For BODIPY-labeled fatty acid gavage, feces were collected from 10 min to 2 h after gavage, freeze-dried and ground up using a mortar and pestle and stored at −20 °C. The mice gavaged with BODIPY-labeled fatty acid were anesthetized after gavage for 2 h and subsequently blood and the tissues of interest were collected. Excised proximal jejunum, iWAT and eWAT were frozen in OCT compound (Sakura, 4583) and sectioned, then examined under fluorescent microscope. For BODIPY concentrations measurement, proximal jejunum, iWAT and eWAT were homogenized in RIPA buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 1% Nonidet P-40 (Sangon Biotech, A100109), 1% sodium deoxycholate and 0.1% SDS) and centrifuged to measure the fluorescent signal in the supernatant as previously described with minor modifications78. The fluorescent signal of the serum and extracted tissue samples was read using a plate reader (Varioskan Flash, Thermo Scientific) with an excitation wavelength of 492 nm and emission wavelength of 520 nm. The dried and ground feces were extracted with water/chloroform (1:2 v/v) and the fluorescent signal in organic phase was measured.

The indicated tissues were collected at ZT10 to extract total RNA with TRIzol Reagent. Sequencing libraries were generated using the TruSeq RNA Sample Preparation kit (Illumina) and then sequenced on a HiSeq X TEN system (Illumina) to generate 2 × 150-bp paired-end reads at a depth of 30–60 million reads per library. High-quality reads were obtained using Cutadapt (v.1.15) software. Reference genome index was built by Bowtie2 (v.2.2.6) and the high-quality reads were mapped to the reference genome (mm10) using Tophat2 (v.2.0.14). Differentially expressed genes with a threshold of P < 0.01 versus the expression level in D0 group and average reads per million >2 before and after SDR were obtained from three replicates per condition using DESeq (v.1.30.0). Differentially expressed genes assigned to fatty acid and TG biosynthetic process (GO:0006633 and GO:0019432 assigned by Mouse Genome Informatics) were used to generate heat maps by GraphPad Prism. For the heat map, the average log2 value of each gene in D0 group was set to 0 and samples in other groups were normalized to the average expression level in D0 group.

Serum amino acids were treated with sulfosalicylic acid and labeled with aTRAQ reagent using an aTRAQ Reagent kit (AB Sciex). After mixing with internal standards, including norvaline and norleucine, free amino acid proportions were determined using LC–MS/MS with an Agilent 1200 series HPLC system and a 4000 Q-Trap tandem mass spectrometer (Applied Biosystems) equipped with an electrospray ionization source.

Eight-week-old male C57BL/6J mice were housed individually and provided with NP diet before and during SDR with 10%, 25% and 65% food from day 1 to day 3. Daily caloric intake of mice refeeding with an HP diet after SDR was determined every 24 h. Mice in pair-feeding groups were fed at ZT12 with a NP or high-fat diet with the same amount of daily calorie consumed by the mice refeeding with an HP diet after SDR.

Approximately 60 mg cecal feces were mixed with 200 µl H2O, 800 µl methanol and 1 μl formic acid, homogenized, vortexed and sonicated for 30 min on ice. Then the samples were stored at −20 °C for 3 h and subsequently centrifuged at 14,000g for 20 min. The supernatant was filtrated with a 0.22-µm PTFE hydrophilic filter. Subsequently, 150 µl filtrated sample using PC 21:0/21:0 (Avanti Polar Lipids, 850370 P) as an internal standard was mixed with 2 ml 0.9% NaCl and 4 ml chloroform:methanol (1:1 v/v), vortexed and centrifuged at 1,000g for 10 min. Extracted lipids in the lower phase were dried under a nitrogen gas stream, methylated with 4% sulfuric acid in 2 ml methanol, incubated in a hot water bath at 80 °C for 1 h, then cooled to room temperature and combined with 2 ml n-hexane. After vortexing and centrifugation, the supernatant was dried under a nitrogen gas stream. Finally, extracts were resuspended in isooctane and analyzed by gas chromatography MS (Agilent 6890N-5975B) with flame ionization detector using a SP-2560 capillary column (length 100 m, internal diameter 0.25 mm and film size 0.2 μm).

Blood and whole intestines, including their contents with the exception of cecal feces, were collected at ZT10. For serum samples, 100 µl serum were mixed with internal standards and 400 µl methanol:acetonitrile (1:1 v/v). Each intestinal sample was weighed, homogenized with internal standards in 4 ml H2O and then combined with 8 ml methanol:acetonitrile (1:1 v/v), vortexed and sonicated for 30 min on ice twice. Subsequently, the serum or intestinal samples were stored at −20 °C for 1 h and centrifuged at 14,000g for 20 min. The supernatant was dried under vacuum and dissolved in 100 µl methanol:water (1:1 v/v). Samples and authentic bile acid standards with a series of gradient concentrations were analyzed using a LC–MS/MS with UHPLC (Acquity UPLC BEH C18, Waters) coupled to 5500 Q-Trap mass spectrometry system (AB Sciex). MS was operated in the negative ionization mode working in the multiple reaction mode. The peak area and retention time data were extracted and calculated using the MultiQuant software (v.3.0.3). Bile acid concentrations were determined using standard curves for each bile acid.

Bacterial genomic DNA was extracted from cecal contents or feces of mice and 10 ng purified DNA was used for PCR amplification. The V3 and V4 regions of the 16S rRNA genes were amplified by PCR using the following barcoded universal forward primer CCTAYGGGRBGCASCAG and reverse primer GGACTACNNGGGTATCTAAT for cecal feces or forward primer CCTACGGGNGGCWGCAG and reverse primer GACTACHVGGGTATCTAATCC for feces. The PCR product was subsequently pooled in equal molar concentrations. Samples were then sequenced using the Novaseq 6000 platform (Illumina) to generate 2 × 250-bp paired-end reads. The high-quality clean reads were obtained by QIIME 2 (v.2019.4) and R packages (v.3.2.0)79 and then were searched against the GreenGenes (v.13.8) reference database. β-diversity analysis was performed to investigate the structural variation of microbial communities across samples based on weighted UniFrac and visualized via PCoA. The significant difference between groups was assessed by permutation multivariate analysis of variance using Adonis test based on weighted UniFrac with 999 permutations (R vegan package). α-diversity was calculated on the basis of the gene profile of each sample according to the Shannon index using Kruskal–Wallis and Dunn's tests. The taxa abundances were assessed at class, family and genus taxonomic levels. To identify group-specific microbial biomarkers, the LEfSe method was used to assess differences in relative abundance of microbial taxa using an LDA score threshold of 5 as described previously80. Songbird multinomial regression analysis was used to identify differentially abundant genera among groups as described previously81.

Fresh cecal feces were collected from mice in the SDR(10–25–65%) group on day 4, then samples were diluted at a 1:10 dilution in sterile anaerobic PBS. A small proportion of the dilutions was spread on MRS agar plates and incubated in an anaerobic chamber (5% H2, 5% CO2 and 90% N2) at 37 °C for 48 h. Individual colonies were randomly picked and cultured in MRS medium for an additional 24 h. Full-length 16S rRNA genes from individual colonies were amplified using the universal 16S primers 27F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492R (5′-CTACGGCTACCTTGTTACGA-3′) and sequenced. The isolated L. murinus (strain Lam-1) was stored in 20% or 6% glycerol at −80 °C until further use. Reference sequences of type strains were obtained by Blast from the GenBank database and multiple sequences were assembled and aligned with ClustalW. Phylogenetic analysis was performed using the neighbor-joining algorithm of MEGA v.7.0.

Cecal contents of mice refeeding with NP or HP diet after SDR or feces of mice upon SDR with or without antibiotic treatments were collected at the indicated time points. Bacterial DNA was isolated as previously described with modifications82. Briefly, approximately 40 mg cecal contents or feces were homogenized in 1 ml PBS containing five porcelain beads with a diameter of 2 mm using TissueLyser II (QIAGEN) at 30 Hz for 15 min and then centrifuged at 12,000g for 5 min. The supernatant was discarded and the pellet was suspended in 0.5 ml of DNA extraction buffer (100 mM Tris-HCl, 1.5 M NaCl, 100 mM Na3PO4, 100 mM EDTA and 1% CTAB, pH 8.0), vortexed and supplemented with 10 µl 20 mg ml−1 proteinase K, then incubated at 37 °C for 1 h on a shaker with a speed of 200 r.p.m. Each tube was then supplemented with 60 μl 20% SDS and incubated at 65 °C for 2 h and subsequently at 95 °C for 20 min. Following centrifugation at 12,000g for 10 min, the clear liquid phase was extracted with an equal volume of phenol and chloroform and then precipitated by isopropanol and resuspended in 60 μl of TE buffer (10 mM Tris and 1 mM EDTA, pH 8.0).

Bacterial DNA was evaluated by electrophoresis with 1% agarose gel and/or analyzed by quantitative PCR. Genomic DNA extracted from E. coli and Lactobacillus Lam-1 was used to establish quantitative PCR standard curves for total bacteria and Lam-1, respectively. Total bacterial 16S rDNA was amplified with universal 16S primers (5′-TCCTACGGGAGGCAGCAGT-3′ and 5′-GGACTACCAGGGTATCTAATCCTGTT-3′)83. Lam-1 16S rDNA was amplified with 5′-CAATGGGCGAAAGCCTGATG-3′ and 5′-TTAGCCGTGGCTTTCTGGTT-3′. The total bacterial or Lam-1 16S rDNA copies were calculated according to corresponding standard curves.

Puromycin (Yeasen Biotech, 60210ES25), vancomycin (Meilunbio, MB1260), neomycin (A610366), metronidazole (A600633), kanamycin (A506636), chloramphenicol (A100230), tetracycline (A100422), ampicillin (A100339) and penicillin (A600135) from Sangon Biotech were used to detect their effects on the growth of Lactobacillus Lam-1. 104 c.f.u. Lam-1 was seeded into 250 μl MRS medium containing the indicated concentrations of antibiotics in 96-well plates and then incubated in an anaerobic chamber at 37 °C. After 18 h, the absorbance at 600 nm was monitored.

Antibiotic mixture treatments were performed as described previously with modifications84. After feeding with 10%, 25% and 65% food from day 1 to day 3, mice were fed ad libitum and gavaged daily for 5 days with a high dose of antibiotic mixture including 10 mg vancomycin, 10 mg ampicillin, 10 mg neomycin and 10 mg metronidazole suspended in 0.2 ml water. Then the mice were gavaged daily with a low dose of antibiotic mixture with 2 mg vancomycin, 4 mg ampicillin, 4 mg neomycin and 4 mg metronidazole suspended in 0.2 ml water. For penicillin treatment, mice were given penicillin via daily drinking water at a dose of 10 µg ml−1 1 week before SDR with 10%, 25% and 65% food from day 1 to day 3. The fat and lean mass and food intake were measured to monitor the effect of antibiotic treatments on obesity induced by refeeding after SDR. To investigate the effect of antibiotic mixture treatments on intestinal lipid absorption and fatty acid uptake in WAT, after feeding with 10%, 25% and 65% food from day 1 to day 3, mice were immediately gavaged with a high dose of antibiotic mixture as mentioned above and fed ad libitum. Eighteen hours later, mice were gavaged with a high dose of antibiotic mixture again. Six hours later, mice were gavaged with BODIPY-labeled fatty acid (0.5 μg g−1 body weight) and blood and the tissues of interest were collected at the indicated time points. For penicillin treatment, mice were given penicillin via daily drinking water at a dose of 200 µg ml−1 1 week before SDR with 10%, 25% and 65% food from day 1 to day 3. After refeeding for 24 h, mice were gavaged with BODIPY-labeled fatty acid and the tissues of interest from mice treated with or without penicillin were collected at the indicated time points.

Fresh cecal feces of donor mice fed ad libitum or fed with NP or HP diet after SDR with 10%, 25% and 65% food from day 1 to day 3 were collected at ZT12 on day 4 and homogenized in sterile PBS (250 mg cecal feces per 1 ml PBS) followed by filtration through a 70-µm strainer. Recipient mice were immediately gavaged with 200 µl of the filtrate. After 24 h, the mice received the cecal fecal microbiota again and were subsequently gavaged with BODIPY-labeled fatty acid (0.5 μg g−1 body weight). Feces, blood and the tissues of interest were collected at the indicated time points for fatty acid absorption assay.

Germ-free C57BL/6J mice were bred and maintained in contamination-controlled flexible film isolators at GemPharmatech and fed an irradiated sterile normal chow diet. Fecal samples from germ-free mice were collected routinely and assessed by aerobic and anaerobic culture and Gram staining to verify germ-free status. Eight bacterial strains representing members of the major bacterial phyla in murine gut were isolated from cecal contents and identified by sequencing of full-length 16S rRNA genes. Bacteroides thetaiotaomicron GB8-1, Parabacteroides distasonis GB8-2 and Enterococcus casseliflavus GB8-6 were anaerobically grown in GAM medium. Bifidobacterium pseudolongum GB8-4 was anaerobically grown in GAM medium supplemented with 10% rumen fluid. Clostridium innocuum GB8-7 and Streptococcus alactolyticus GB8-8 were anaerobically grown in TSB-YE medium. E. coli GB8-3 was aerobically grown in TSB-YE medium. Dubosiella newyorkensis GB8-5 was anaerobically grown in PYG medium supplemented with 5% rumen fluid, 5% bovine serum, 0.1% vitamin K1 and 0.5% hemin. Bacteria in late log phase were collected by centrifugation, resuspended in PBS containing 30% glycerol (109 c.f.u. per ml for each bacterial strain) and the eight bacterial strains were mixed with each other in equal volume. Germ-free C57BL/6J mice were gavaged with the mixture (0.2 ml) at the age of 5 weeks and 6.5 weeks, respectively. Bacterial colonization was confirmed by 16S rDNA sequencing of fecal microbiota in mice at the age of 6.5 weeks and 9 weeks.

Lactobacillus paracasei LC2W (formerly referred to as Lactobacillus casei LC2W) was cultured in MRS medium at 37 °C in an anaerobic chamber (5% H2, 5% CO2 and 90% N2)85, confirmed by sequencing of full-length 16S rRNA genes and stored in PBS containing 6% glycerol at −80 °C until further use. Germ-free or gnotobiotic mice at age of 9 weeks were gavaged with 0.2 ml LC2W (2 × 109 c.f.u.) or Lam-1 (2 × 109 c.f.u.) for three consecutive days and then fed with 10%, 25% and 65% food in the next 3 d. After feeding ad libitum for 24 h, mice were gavaged with BODIPY-labeled fatty acid (0.5 μg g−1 body weight) and feces, blood and the tissues of interest were collected at the indicated time points. The presence of LC2W or Lam-1 in feces was verified by 16S rDNA sequencing. SPF mice were first gavaged with 0.2 ml PBS containing 6% glycerol (vehicle) or 0.2 ml LC2W (1010 c.f.u.) or Lam-1 (1010 c.f.u.) in PBS containing 6% glycerol prewarmed at 37 °C in a water bath for 5–10 min. After 24 h, mice were gavaged with the same amount of vehicle, LC2W or Lam-1 again and subsequently BODIPY-labeled fatty acid (0.5 μg g−1 body weight) was gavaged. Feces, cecal contents, blood and the tissues of interest were collected at the indicated time points. To investigate the effect of Lam-1 on fat mass accumulation, SPF mice were gavaged with the vehicle, LC2W or Lam-1 as mentioned above every day and fat and lean mass and food intake were measured daily.

Blood and cecal feces were collected at ZT10. For serum samples, 100 µl serum were mixed with 400 µl methanol:acetonitrile (1:1 v/v). For cecal fecal samples, approximately 80 mg cecal feces were mixed with 200 µl H2O and 800 µl methanol:acetonitrile (1:1 v/v), homogenized, vortexed and sonicated for 30 min on ice. Then the serum or cecal fecal samples were stored at −20 °C for 1 h and subsequently centrifuged at 14,000g for 20 min. The supernatant containing metabolites was dried under vacuum, dissolved in 100 µl acetonitrile:water (1:1 v/v) and analyzed using LC–MS/MS with an UHPLC (1290 Infinity LC, Agilent Technologies) coupled to a quadrupole time-of-flight mass spectrometer (AB Sciex TripleTOF 6600). The MS was operated in both negative and positive ionization modes with the capillary voltage set to 5.5 kV. The raw MS data were extracted and peak-identified and the annotation of isotopes and adducts was performed using the R-package CAMERA86. Compounds were identified by comparing of accurate m/z value and spectra with a database established with available authentic standards.

Metabolites in serum with variable importance in the projection values >1 obtained from the orthogonal partial least-squares discriminant analysis (OPLS-DA) model and P values <0.05 from two-tailed Student's t-test were considered as differential. The KEGG library was applied to the enrichment analysis of metabolic pathway based on the differential metabolites in serum. EAAs, semi-EAAs and non-EAAs in serum were used to generate a heat map. Metabolites in cecal feces that significantly increased in both the SDR-NP(D4) and SDR-HP(D4) groups compared to the AL-NP(D0) group (excluding metabolites that significantly decreased in the SDR-HP(D4) group compared to SDR-NP(D4) group), were considered as unblockable increases. Metabolites significantly decreased in both the SDR-NP(D4) and SDR-HP(D4) groups compared to AL-NP(D0) group (excluding metabolites that significantly increased in the SDR-HP(D4) group compared to SDR-NP(D4) group), were considered as unblockable decreases. Metabolites significantly decreased in the SDR-NP(D4) group compared to AL-NP(D0) group and significantly increased in SDR-HP(D4) group compared to SDR-NP(D4) (or with no significant change compared to AL-NP(D0) group) were considered as blockable decreases. Metabolites significantly increased in the SDR-NP(D4) group compared to AL-NP(D0) group and significantly decreased in the SDR-HP(D4) group compared to SDR-NP(D4) (or with no significant change compared to AL-NP(D0) group) were considered as blockable increases. These classified metabolites were used to generate a heat map. Spearman's rank correlation coefficients (two-sided tests) between the relative concentrations of metabolites in cecal feces and the relative abundances of gut microbiota genera (average relative abundance >0.1% and LDA score >2 between two groups) were performed in GraphPad Prism to generate heat maps.

E. coli DH5α cultured in LB medium, Lactobacillus LC2W and Lam-1 cultured in MRS medium in the early stationary growth phase with a concentration of approximately 109 c.f.u. per ml were collected. After centrifugation of 2 ml bacterial cultures, the resultant bacterial pellets were mixed with 400 µl H2O, 800 µl methanol and 1.2 μl formic acid. A volume of 0.4 ml cell-free supernatant or 0.4 ml LB or MRS medium was mixed with 800 µl methanol and 1.2 μl formic acid. Then the samples were homogenized, vortexed and sonicated for 30 min on ice. For cecal fecal samples, about 60 mg cecal feces were mixed with 200 µl H2O, 800 µl methanol and 1 μl formic acid, homogenized, vortexed and sonicated for 30 min on ice. Then all the samples were stored at −20 °C for 3 h and subsequently centrifuged at 14,000g for 20 min. The supernatant was filtrated with a 0.22-µm PTFE hydrophilic filter for LC–MS/MS analysis. Samples and commercial compounds including DL-3-PLA (TCI, P1981 and P1168), HPLA (TCI, H0542), ILA (TCI, I0031), HICA (Sigma-Aldrich, 219819) and HMBA (Sigma-Aldrich, 219835) at different concentrations as a standard curve were analyzed by LC–MS/MS with an Agilent 1200 series HPLC system and a 4000 Q-Trap tandem mass spectrometer (Applied Biosystems) equipped with an electrospray ionization source. The metabolites in each sample were quantified according to the standard curves using Analyst v.1.6.2 software.

Mice were gavaged with water (vehicle), 60 mg ml−1 PLA, 30 mg ml−1 HPLA, 15 mg ml−1 ILA, 40 mg ml−1 HICA, 20 mg ml−1 HMBA, a mixture of three metabolites (60 mg ml−1 PLA, 30 mg ml−1 HPLA and 15 mg ml−1 ILA), a mixture of five metabolites (30 mg ml−1 PLA, 15 mg ml−1 HPLA, 8 mg ml−1 ILA, 40 mg ml−1 HICA and 20 mg ml−1 HMBA), 113 mg ml−1 glutamic acid or 113 mg ml−1 glucose each at 150 µl. Each solution used for gavage was adjusted to pH 7.0 with sodium hydroxide. After 24 h, mice were gavaged with the same amount of water or the indicated metabolites again and subsequently BODIPY-labeled fatty acid (0.5 μg g−1 body weight) were gavaged. Feces, blood and the tissues of interest were collected at the indicated time points. To investigate the effect of metabolites on fat mass accumulation, mice were gavaged daily with 150 µl water or a mixture of five metabolites as mentioned above and the fat and lean mass and food intake were monitored.

Data are expressed as mean ± s.d. Except indicated, statistical significance was analyzed using two-tailed Student's t-test and differences were considered statistically significant at P < 0.05. No statistical methods were used to predetermine sample sizes but our sample sizes are similar to those reported in previous publications50,77,78. Mice were randomly allocated to groups based on weight and body fat mass. The investigators were not always able to blinded due to limitations in available personnel; however, data collection and analysis for metabolic rate and physical activity, high-throughput RNA and 16S rDNA sequencing, quantification of bile acids, targeted quantification of metabolites and non-targeted metabolomics were blinded. No samples or animals were excluded from the analyses unless due to technical discrepancy in the treatment we had to terminate the study for this particular sample or animal. Statistical analysis was performed using Microsoft Excel 2016 and GraphPad Prism v.8.0. The experiments to obtain the data shown in Fig. 1b–e, Fig. 1i–k, Fig. 1l–n, Fig. 2e,f, Extended Data Fig. 3h–q, Extended Data Fig. 3x,y, Extended Data Fig. 4a and Fig. 2d, or Extended Data Fig. 4c,d were performed simultaneously. To clearly show these data, the data obtained from the same experiment were shown in different figures with the same data of the control group fed ad libitum.

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

The raw data of RNA sequencing generated in this study have been deposited in the National Center for Biotechnology Information Sequence Read Archive (https://www.ncbi.nlm.nih.gov/sra) under accession no. PRJNA612577. The accession nos. for the 16S rDNA sequencing data reported in this study are National Center for Biotechnology Information BioProject PRJNA868153 and PRJNA868196. The 16S rRNA gene complete sequences of L. paracasei LC2W and isolated strains including Lam-1, B. thetaiotaomicron GB8-1, P. distasonis GB8-2, E. coli GB8-3, B. pseudolongum GB8-4, D. newyorkensis GB8-5, E. casseliflavus GB8-6, C. innocuum GB8-7 and S. alactolyticus GB8-8 generated during this study have been submitted to GenBank under accession nos. OP179631, MZ955456, OP185294, OP185272, OP185287, OP185295, OP185270, OP185274, OP185269 and OP185264, respectively. The reference genome (mm10) is available at https://www.ncbi.nlm.nih.gov/genome. GreenGenes (v.13.8) reference database is available at https://greengenes.secondgenome.com. All other data that support the findings of this study are available as source data files. Source data are provided with this paper.

Kumanyika, S. & Dietz, W. H. Solving population-wide obesity - progress and future prospects. N. Engl. J. Med. 383, 2197–2200 (2020).

Article Google Scholar

Blüher, M. Obesity: global epidemiology and pathogenesis. Nat. Rev. Endocrinol. 15, 288–298 (2019).

Article Google Scholar

Gao, M. et al. Associations between body-mass index and COVID-19 severity in 6·9 million people in England: a prospective, community-based, cohort study. Lancet Diabetes Endocrinol. 9, 350–359 (2021).

Article CAS Google Scholar

Heymsfield, S. B. & Wadden, T. A. Mechanisms, pathophysiology, and management of obesity. N. Engl. J. Med. 376, 254–266 (2017).

Article CAS Google Scholar

van der Klaauw, A. A. & Farooqi, I. S. The hunger genes: pathways to obesity. Cell 161, 119–132 (2015).

Article Google Scholar

Pillon, N. J., Loos, R. J. F., Marshall, S. M. & Zierath, J. R. Metabolic consequences of obesity and type 2 diabetes: balancing genes and environment for personalized care. Cell 184, 1530–1544 (2021).

Article CAS Google Scholar

Franks, P. W. & McCarthy, M. I. Exposing the exposures responsible for type 2 diabetes and obesity. Science 354, 69–73 (2016).

Article CAS Google Scholar

McHill, A. W. & Wright, K. P. Jr. Role of sleep and circadian disruption on energy expenditure and in metabolic predisposition to human obesity and metabolic disease. Obes. Rev. 18, 15–24 (2017).

Article Google Scholar

Milaneschi, Y., Simmons, W. K., van Rossum, E. F. C. & Penninx, B. W. Depression and obesity: evidence of shared biological mechanisms. Mol. Psychiatry 24, 18–33 (2019).

Article CAS Google Scholar

Chao, A. M., Quigley, K. M. & Wadden, T. A. Dietary interventions for obesity: clinical and mechanistic findings. J. Clin. Invest. 131, e140065 (2021).

Article CAS Google Scholar

Di Francesco, A., Di Germanio, C., Bernier, M. & de Cabo, R. A time to fast. Science 362, 770–775 (2018).

Article Google Scholar

Lettieri-Barbato, D., Giovannetti, E. & Aquilano, K. Effects of dietary restriction on adipose mass and biomarkers of healthy aging in human. Aging 8, 3341–3355 (2016).

Article CAS Google Scholar

Most, J., Tosti, V., Redman, L. M. & Fontana, L. Calorie restriction in humans: an update. Ageing Res. Rev. 39, 36–45 (2017).

Article Google Scholar

Speakman, J. R. & Mitchell, S. E. Caloric restriction. Mol. Asp. Med. 32, 159–221 (2011).

Article CAS Google Scholar

Slof-Op ‘t Landt, M. C. T. et al. Prevalence of dieting and fear of weight gain across ages: a community sample from adolescents to the elderly. Int. J. Public Health 62, 911–919 (2017).

Article Google Scholar

Santos, I., Sniehotta, F. F., Marques, M. M., Carraça, E. V. & Teixeira, P. J. Prevalence of personal weight control attempts in adults: a systematic review and meta-analysis. Obes. Rev. 18, 32–50 (2017).

Article CAS Google Scholar

Templeman, I. et al. A randomized controlled trial to isolate the effects of fasting and energy restriction on weight loss and metabolic health in lean adults. Sci. Transl. Med. 13, eabd8034 (2021).

Article Google Scholar

Kraus, W. E. et al. 2 years of calorie restriction and cardiometabolic risk (CALERIE): exploratory outcomes of a multicentre, phase 2, randomised controlled trial. Lancet Diabetes Endocrinol. 7, 673–683 (2019).

Article Google Scholar

Robertson, L. T. & Mitchell, J. R. Benefits of short-term dietary restriction in mammals. Exp. Gerontol. 48, 1043–1048 (2013).

Article CAS Google Scholar

Nas, A. et al. Impact of breakfast skipping compared with dinner skipping on regulation of energy balance and metabolic risk. Am. J. Clin. Nutr. 105, 1351–1361 (2017).

CAS Google Scholar

Pendergast, F. J., Livingstone, K. M., Worsley, A. & McNaughton, S. A. Correlates of meal skipping in young adults: a systematic review. Int. J. Behav. Nutr. Phys. Act. 13, 125 (2016).

Article Google Scholar

Johnstone, A. M. Fasting - the ultimate diet? Obes. Rev. 8, 211–222 (2007).

Article CAS Google Scholar

Maughan, R. J., Fallah, J. & Coyle, E. F. The effects of fasting on metabolism and performance. Br. J. Sports Med. 44, 490–494 (2010).

Article CAS Google Scholar

Trepanowski, J. F. & Bloomer, R. J. The impact of religious fasting on human health. Nutr. J. 9, 57 (2010).

Article Google Scholar

Dulloo, A. G. & Montani, J. P. Pathways from dieting to weight regain, to obesity and to the metabolic syndrome: an overview. Obes. Rev. 16, 1–6 (2015).

Article Google Scholar

Lowe, M. R. Dieting: proxy or cause of future weight gain? Obes. Rev. 16, 19–24 (2015).

Article Google Scholar

Montani, J. P., Schutz, Y. & Dulloo, A. G. Dieting and weight cycling as risk factors for cardiometabolic diseases: who is really at risk? Obes. Rev. 16, 7–18 (2015).

Article Google Scholar

Zmora, N., Suez, J. & Elinav, E. You are what you eat: diet, health and the gut microbiota. Nat. Rev. Gastroenterol. Hepatol. 16, 35–56 (2019).

Article CAS Google Scholar

Albenberg, L. G. & Wu, G. D. Diet and the intestinal microbiome: associations, functions, and implications for health and disease. Gastroenterology 146, 1564–1572 (2014).

Article CAS Google Scholar

Gentile, C. L. & Weir, T. L. The gut microbiota at the intersection of diet and human health. Science 362, 776–780 (2018).

Article CAS Google Scholar

Rinninella, E. et al. Gut microbiota during dietary restrictions: new insights in non-communicable diseases. Microorganisms 8, 1140 (2020).

Article CAS Google Scholar

von Schwartzenberg, R. J. et al. Caloric restriction disrupts the microbiota and colonization resistance. Nature 595, 272–277 (2021).

Article Google Scholar

Fraumene, C. et al. Caloric restriction promotes rapid expansion and long-lasting increase of Lactobacillus in the rat fecal microbiota. Gut Microbes 9, 104–114 (2018).

Article Google Scholar

Chadaideh, K. S. & Carmody, R. N. Host-microbial interactions in the metabolism of different dietary fats. Cell Metab. 33, 857–872 (2021).

Article CAS Google Scholar

Martinez-Guryn, K. et al. Small intestine microbiota regulate host digestive and absorptive adaptive responses to dietary lipids. Cell Host Microbe 23, 458–469 (2018).

Article CAS Google Scholar

Delzenne, N. M., Neyrinck, A. M., Bäckhed, F. & Cani, P. D. Targeting gut microbiota in obesity: effects of prebiotics and probiotics. Nat. Rev. Endocrinol. 7, 639–646 (2011).

Article CAS Google Scholar

Moreno-Navarrete, J. M. & Fernandez-Real, J. M. The gut microbiota modulates both browning of white adipose tissue and the activity of brown adipose tissue. Rev. Endocr. Metab. Disord. 20, 387–397 (2019).

Article CAS Google Scholar

Parséus, A. et al. Microbiota-induced obesity requires farnesoid X receptor. Gut 66, 429–437 (2017).

Article Google Scholar

Krautkramer, K. A., Fan, J. & Bäckhed, F. Gut microbial metabolites as multi-kingdom intermediates. Nat. Rev. Microbiol. 19, 77–94 (2021).

Article CAS Google Scholar

Gutierrez Lopez, D. E., Lashinger, L. M., Weinstock, G. M. & Bray, M. S. Circadian rhythms and the gut microbiome synchronize the host's metabolic response to diet. Cell Metab. 33, 873–887 (2021).

Article CAS Google Scholar

Cani, P. D. et al. Microbial regulation of organismal energy homeostasis. Nat. Metab. 1, 34–46 (2019).

Article CAS Google Scholar

Kolodziejczyk, A. A., Zheng, D. & Elinav, E. Diet-microbiota interactions and personalized nutrition. Nat. Rev. Microbiol. 17, 742–753 (2019).

Article CAS Google Scholar

Schroeder, B. O. & Bäckhed, F. Signals from the gut microbiota to distant organs in physiology and disease. Nat. Med. 22, 1079–1089 (2016).

Article CAS Google Scholar

Das, N. K. et al. Microbial metabolite signaling is required for systemic iron homeostasis. Cell Metab. 31, 115–130 (2020).

Article CAS Google Scholar

Quan, L. H. et al. Myristoleic acid produced by enterococci reduces obesity through brown adipose tissue activation. Gut 69, 1239–1247 (2020).

Article CAS Google Scholar

Araújo, J. R. et al. Fermentation products of commensal bacteria alter enterocyte lipid metabolism. Cell Host Microbe 27, 358–375 (2020).

Article Google Scholar

Canfora, E. E., Meex, R. C. R., Venema, K. & Blaak, E. E. Gut microbial metabolites in obesity, NAFLD and T2DM. Nat. Rev. Endocrinol. 15, 261–273 (2019).

Article CAS Google Scholar

Choi, H., Rao, M. C. & Chang, E. B. Gut microbiota as a transducer of dietary cues to regulate host circadian rhythms and metabolism. Nat. Rev. Gastroenterol. Hepatol. 18, 679–689 (2021).

Article Google Scholar

Li, G. et al. Intermittent fasting promotes white adipose browning and decreases obesity by shaping the gut microbiota. Cell Metab. 26, 672–685 (2017).

Article CAS Google Scholar

Thaiss, C. A. et al. Persistent microbiome alterations modulate the rate of post-dieting weight regain. Nature 540, 544–551 (2016).

Article CAS Google Scholar

de Cabo, R. & Mattson, M. P. Effects of intermittent fasting on health, aging, and disease. N. Engl. J. Med. 381, 2541–2551 (2019).

Article Google Scholar

Bray, M. S. et al. Quantitative analysis of light-phase restricted feeding reveals metabolic dyssynchrony in mice. Int. J. Obes. 37, 843–852 (2013).

Article CAS Google Scholar

de Aguiar Vallim, T. Q., Tarling, E. J. & Edwards, P. A. Pleiotropic roles of bile acids in metabolism. Cell Metab. 17, 657–669 (2013).

Article Google Scholar

Chiang, J. Y. L. & Ferrell, J. M. Bile acids as metabolic regulators and nutrient sensors. Annu. Rev. Nutr. 39, 175–200 (2019).

Article CAS Google Scholar

Kuang, Z. et al. The intestinal microbiota programs diurnal rhythms in host metabolism through histone deacetylase 3. Science 365, 1428–1434 (2019).

Article CAS Google Scholar

Fabbiano, S. et al. Functional gut microbiota remodeling contributes to the caloric restriction-induced metabolic improvements. Cell Metab. 28, 907–921 (2018).

Article CAS Google Scholar

Zarrinpar, A. et al. Antibiotic-induced microbiome depletion alters metabolic homeostasis by affecting gut signaling and colonic metabolism. Nat. Commun. 9, 2872 (2018).

Article Google Scholar

Dodd, D. et al. A gut bacterial pathway metabolizes aromatic amino acids into nine circulating metabolites. Nature 551, 648–652 (2017).

Article CAS Google Scholar

Custers, K. & Van den Bulck, J. Television viewing, computer game play and book reading during meals are predictors of meal skipping in a cross-sectional sample of 12-, 14- and 16-year-olds. Public Health Nutr. 13, 537–543 (2010).

Article Google Scholar

Liechty, J. M. & Lee, M. J. Longitudinal predictors of dieting and disordered eating among young adults in the U.S. Int. J. Eat. Disord. 46, 790–800 (2013).

Article Google Scholar

Turicchi, J. et al. Associations between the rate, amount, and composition of weight loss as predictors of spontaneous weight regain in adults achieving clinically significant weight loss: a systematic review and meta-regression. Obes. Rev. 20, 935–946 (2019).

Article Google Scholar

Engber, D. Unexpected clues emerge about why diets fail. Nat. Med. 25, 1637–1639 (2019).

Article CAS Google Scholar

Melby, C. L., Paris, H. L., Foright, R. M. & Peth, J. Attenuating the biologic drive for weight regain following weight loss: must what goes down always go back up? Nutrients 9, 468 (2017).

Article Google Scholar

Fothergill, E. et al. Persistent metabolic adaptation 6 years after "The Biggest Loser" competition. Obesity 24, 1612–1619 (2016).

Article Google Scholar

Vink, R. G., Roumans, N. J., Arkenbosch, L. A., Mariman, E. C. & van Baak, M. A. The effect of rate of weight loss on long-term weight regain in adults with overweight and obesity. Obesity 24, 321–327 (2016).

Article CAS Google Scholar

Kleinert, M. et al. Animal models of obesity and diabetes mellitus. Nat. Rev. Endocrinol. 14, 140–162 (2018).

Article Google Scholar

Ko, C. W., Qu, J., Black, D. D. & Tso, P. Regulation of intestinal lipid metabolism: current concepts and relevance to disease. Nat. Rev. Gastroenterol. Hepatol. 17, 169–183 (2020).

Article CAS Google Scholar

MacLean, P. S., Higgins, J. A., Giles, E. D., Sherk, V. D. & Jackman, M. R. The role for adipose tissue in weight regain after weight loss. Obes. Rev. 16, 45–54 (2015).

Article Google Scholar

van Baak, M. A. & Mariman, E. C. M. Mechanisms of weight regain after weight loss - the role of adipose tissue. Nat. Rev. Endocrinol. 15, 274–287 (2019).

Article Google Scholar

Brehm, B. J. & D’Alessio, D. A. Benefits of high-protein weight loss diets: enough evidence for practice? Curr. Opin. Endocrinol. Diabetes Obes. 15, 416–421 (2008).

Article CAS Google Scholar

Westerterp-Plantenga, M. S., Lemmens, S. G. & Westerterp, K. R. Dietary protein - its role in satiety, energetics, weight loss and health. Br. J. Nutr. 108, S105–S112 (2012).

Article CAS Google Scholar

Pan, F. et al. Predominant gut Lactobacillus murinus strain mediates anti-inflammaging effects in calorie-restricted mice. Microbiome 6, 54 (2018).

Article Google Scholar

Borthakur, A. et al. The probiotic Lactobacillus acidophilus stimulates chloride/hydroxyl exchange activity in human intestinal epithelial cells. J. Nutr. 138, 1355–1359 (2008).

Article CAS Google Scholar

María Remes Troche, J. et al. Lactobacillus acidophilus LB: a useful pharmabiotic for the treatment of digestive disorders. Ther. Adv. Gastroenterol. 13, 1756284820971201 (2020).

Article Google Scholar

Hempel, S. et al. Probiotics for the prevention and treatment of antibiotic-associated diarrhea: a systematic review and meta-analysis. JAMA 307, 1959–1969 (2012).

Article CAS Google Scholar

Talukdar, S. et al. Neutrophils mediate insulin resistance in mice fed a high-fat diet through secreted elastase. Nat. Med. 18, 1407–1412 (2012).

Article CAS Google Scholar

Liu, Y. et al. Liver Patt1 deficiency protects male mice from age-associated but not high-fat diet-induced hepatic steatosis. J. Lipid Res. 53, 358–367 (2012).

Article CAS Google Scholar

Khalifeh-Soltani, A. et al. Mfge8 promotes obesity by mediating the uptake of dietary fats and serum fatty acids. Nat. Med. 20, 175–183 (2014).

Article CAS Google Scholar

Bolyen, E. et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat. Biotechnol. 37, 852–857 (2019).

Article CAS Google Scholar

Segata, N. et al. Metagenomic biomarker discovery and explanation. Genome Biol. 12, R60 (2011).

Article Google Scholar

Morton, J. T. et al. Establishing microbial composition measurement standards with reference frames. Nat. Commun. 10, 2719 (2019).

Article Google Scholar

Wilson, K. Preparation of genomic DNA from bacteria. Curr. Protoc. Mol. Biol. https://doi.org/10.1002/0471142727.mb0204s56 (2001).

Nadkarni, M. A., Martin, F. E., Jacques, N. A. & Hunter, N. Determination of bacterial load by real-time PCR using a broad-range (universal) probe and primers set. Microbiology 148, 257–266 (2002).

Article CAS Google Scholar

Kuss, S. K. et al. Intestinal microbiota promote enteric virus replication and systemic pathogenesis. Science 334, 249–252 (2011).

Article CAS Google Scholar

Chen, C. et al. Complete genome sequence of the probiotic bacterium Lactobacillus casei LC2W. J. Bacteriol. 193, 3419–3420 (2011).

Article CAS Google Scholar

Kuhl, C., Tautenhahn, R., Böttcher, C., Larson, T. R. & Neumann, S. CAMERA: an integrated strategy for compound spectra extraction and annotation of liquid chromatography/mass spectrometry data sets. Anal. Chem. 84, 283–289 (2012).

Article CAS Google Scholar

Download references

This work was supported by grants from the National Natural Science Foundation of China to Q.Z. (31630037, 91740103 and 91940306) and Y.C. (32200969), the National Key R&D Program of China to Q.Z. (2018YFA0800603), the Strategic Priority Research Program of the Chinese Academy of Sciences to Q.Z. (XDB19030103), the Frontier Science of Chinese Academy of Sciences Key Research Projects to Q.Z. (QYZDJ-SSW-SMC022), the Shanghai Leading Talents Program to Q.Z., the Shanghai Postdoctoral Excellence Program to Y.C. and the China Postdoctoral Science Foundation Grant to Y.C. (2019M661658).

CAS Key Laboratory of Nutrition, Metabolism and Food Safety, Shanghai Institute of Nutrition and Health, University of Chinese Academy of Sciences, Chinese Academy of Sciences, Shanghai, China

Wuling Zhong, Hui Wang, Yale Yang, Yali Zhang, Hejin Lai, Yalan Cheng, Huimin Yu, Ning Feng, Rui Huang, Hao Ying & Qiwei Zhai

Department of Orthopaedics, Shanghai Jiao Tong University School of Medicine Affiliated Sixth People's Hospital, Shanghai, China

Shen Liu

Key Laboratory of Synthetic Biology, CAS Center for Excellence in Molecular Plant Sciences, Chinese Academy of Sciences, Shanghai, China

Sheng Yang

GemPharmatech Co., Nanjing, China

Tongyang Hao & Baoyu Zhang

National Clinical Research Center for Eye Diseases, Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China

Fang Zhang

Zhongshan Hospital, State Key Laboratory of Medical Neurobiology, Institute for Translational Brain Research, Ministry of Education Frontiers Center for Brain Science, Fudan University, Shanghai, China

Feifan Guo

School of Life Science and Technology, ShanghaiTech University, Shanghai, China

Qiwei Zhai

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

W.Z., H.W., Y.Y., Y.Z., H.L., Y.C., H.M.Y., N.F., R.H., S.L., S.Y., T.H., B.Z., H.Y., F.Z., F.G. and Q.Z. participated in data interpretation and revising the paper and approved the final version of the manuscript. W.Z. performed most of the experiments. H.W. performed absolute quantification of amino acids, fatty acids and metabolites using MS. W.Z. and Y.Y. measured body composition of mice. T.H. and B.Z. helped with experiments using germ-free and gnotobiotic mice. W.Z., Y.Z., H.L., Y.C., H.M.Y. and N.F. collected tissues of mice. W.Z. and Q.Z. designed the study, analyzed the data and wrote the manuscript. Q.Z. supervised the project.

Correspondence to Qiwei Zhai.

T.H. and B.Z. are employees of GemPharmatech and T.H. has stakeholder interests. All other authors have no competing interests.

Nature Metabolism thanks the anonymous reviewers for their contribution to the peer review of this work.

Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

a, Experimental scheme of mice upon short-term dietary restriction (SDR) with 10%, 25% and 65% food intake from day 1 to day 3. Arrows show the time points to provide food, to measure body weight, body composition and food intake. Each mouse was housed separately for 5 days before the SDR experiment. Average food intake within 3 days before the SDR experiment was measured as the dietary reference for mice fed with a 10%-25%-65% diet in the following three days. ZT, Zeitgeber time. ZT0 indicates the beginning of light time, ZT12 indicates the beginning of dark time. b, c, Refeeding after feeding with 1-day food in 3 days or 4-day food in 6 days markedly increased body fat percentage (b) and decreased lean mass percentage (c). AL, ad libitum (n = 9 biologically independent mice); SDR(10%-25%-65%), 10%, 25% and 65% food provided from day 1 to day 3 (n = 9); SDR(65%-25%-10%), 65%, 25% and 10% food provided from day 1 to 3 (n = 7); SDR(33.3%×3), 33.3% food daily provided from day 1 to 3 (n = 8); SDR(66.7%×6), 66.7% food daily provided from day 1 to 6 (n = 9). SDR(10%-25%-65%) group was compared to AL group. d, e, Body lean mass (d) and daily food intake (e) of mice in (b). n = 9 biologically independent mice for AL, SDR(10%-25%-65%) and SDR(66.7%×6), n = 7 for SDR(65%-25%-10%), n = 8 for SDR(33.3%×3). SDR(10%-25%-65%) group was compared to AL group. f, Experimental scheme of mice upon SDR with 10%, 25% and 65% food intake from day 1 to day 3. Arrows show the time points to provide food, to collect blood and tissue samples. g, Representative images to show intra-abdominal fat at ZT10 (D0L-D12L) or ZT15 (D4D, day 4 at dark phase) from SDR(10%-25%-65%) group. Yellow arrow heads indicate the inguinal adipose tissues (iWAT). h, i, Representative images of isolated inguinal white adipose tissue (iWAT) (h) and epididymal white adipose tissue (eWAT) (i) from mice in (g). The size for each square is 0.5 cm × 0.5 cm. j, The weight of eWAT in (i) (n = 12 biologically independent mice per group). k, Average adipocyte area of iWAT in (Fig. 1h) (n = 6 biologically independent samples per group). l, Representative images of H&E-stained sections of eWAT from mice in (g). m, Average adipocyte area of eWAT in (l) (n = 6 biologically independent samples per group). Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test. a, P < 0.05; b, **P < 0.01; c, ***P < 0.001; versus AL or D0 group. Exact P values can be found in Source Data Extended Data Fig. 1.

Source data

a, Representative images of isolated interscapular brown adipose tissue (iBAT) at ZT10 (D0L-D12L) or ZT15 (D4D, day 4 at dark phase) from mice in SDR(10%-25%-65%) group. b, The weight of iBAT from mice in (a) (n = 12 biologically independent mice per group). c, Representative images of H&E-stained sections of iBAT from mice in (a). d, Representative images of H&E-stained sections of liver from mice in (a). e, Triglyceride (TG) content of liver from mice in (a) (n = 6 biologically independent samples for D0, D4D, D4, D6, D9 and D12, n = 9 for D1 and D3, n = 11 for D2). Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test. *P < 0.05; **P < 0.01; ***P < 0.001; versus D0 group. Exact P values can be found in Source Data Extended Data Fig. 2.

Source data

a, Experimental design of short-term dietary restriction (SDR) in mice feeding with 2-day food in 3 days or 1-day food in 2 days. SDR(40%-70%-90%), 40%, 70% and 90% food provided from day 1 to day 3; SDR(90%-70%-40%), 90%, 70% and 40% food provided from day 1 to 3; SDR(50%×2), 50% food daily provided from day 1 to day 2. b, c, Refeeding after feeding with 2-day food in 3 days or 1-day food in 2 days markedly increased fat mass percentage (b) and decreased lean mass percentage (c). n = 8 biologically independent mice for AL, n = 9 for other groups. SDR(40%-70%-90%) group was compared to AL group. d-f, Body weight (d), body lean mass (e) and daily food intake (f) of mice in (b). n = 8 biologically independent mice for AL, n = 9 for other groups. SDR(40%-70%-90%) group was compared to AL group. g, Experimental design of DR in mice feeding with 66.7% food daily for 12 or 24 days or alternate-day fasting for 15 cycles. DR(66.7%×12), 66.7% food daily provided from day 1 to day 12; DR(66.7%×24), 66.7% food daily provided from day 1 to day 24; DR(0%-AL)×15, fasted in day 1 and fed ad libitum in day 2 for 15 cycles. h, i, Refeeding after feeding with 66.7% food daily for 12 or 24 days markedly increased fat mass percentage (h) and decreased lean mass percentage (i) (n = 9 biologically independent mice per group). j-l, Body weight (j), body lean mass (k) and daily food intake (l) of mice in (h) (n = 9 biologically independent mice per group). m, n, Refeeding after alternate-day fasting for 15 cycles markedly increased fat mass percentage (m) and decreased lean mass percentage (n) (n = 9 biologically independent mice per group). o-q, Body weight (o), body lean mass (p) and daily food intake (q) of mice in (m) (n = 9 biologically independent mice per group). r-u, Energy expenditure (EE) (r, t) and physical activity (s, u) of mice fed ad libitum (n = 7, 7, 6 and 5 biologically independent mice in (r), (s), (t) and (u) respectively), or mice in SDR(10%-25%-65%) (n = 8 biologically independent mice) and SDR(33.3%×3) group (n = 6 biologically independent mice). v-y, Body temperature of mice fed ad libitum or upon DR for the indicated patterns. n = 9 and 8 biologically independent mice for AL in (v, x, y) and (w) respectively, n = 7 for SDR(65%-25%-10%), n = 8 for SDR(33.3%×3), n = 9 for other groups. SDR(10%-25%-65%) or SDR(40%-70%-90%) group was compared to AL group. The data shown in (h-q) or (x, y) were performed simultaneously with a single control experiment. Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test. a, P < 0.05; b, P < 0.01; c, P < 0.001; NS, not significant; versus AL group. Exact P values can be found in Source Data Extended Data Fig. 3.

Source data

a-d, Cumulative food intake of mice fed ad libitum and upon DR for the indicated patterns. n = 9 and 8 biologically independent mice for AL in (a, c, d) and (b) respectively, n = 7 for SDR(65%-25%-10%), n = 9 for other groups. SDR(65%-25%-10%) or SDR(40%-70%-90%) group was compared to AL group. e, Experimental design of SDR in mice feeding with 1-day food in 3 days following with 100% food as much as in D0. f, Daily food intake of mice in different regimens in (e) (n = 9 biologically independent mice per group). SDR(10%-25%-65%)-100%×23 or SDR(33.3%×3)-100%×23, 10%, 25% and 65% food or 33.3% food daily provided from day 1 to day 3, and subsequently 100% food as much as in D0 provided for 23 days. SDR(10%-25%-65%)-100%×23 group was compared to AL group. g, h, Refeeding with food as much as in D0 after feeding with 1-day food in 3 days markedly increased body fat mass percentage (g) and decreased lean mass percentage (h) (n = 9 biologically independent mice per group). SDR(10%-25%-65%)-100%×23 group was compared to AL group. i, j, Body weight (i) and body lean mass (j) of mice in (f) (n = 9 biologically independent mice per group). SDR(10%-25%-65%)-100%×23 group was compared to AL group. k, Heat map of differentially expressed genes associated with fatty acid and triglyceride biosynthesis in iWAT from SDR(10%-25%-65%) group (n = 3 biologically independent samples per group). The average log2 value of each gene in D0 group was set to 0. Samples in D4 or D6 group were shown in triplicate and normalized to the average expression level in D0 group. l, m, Average respiratory exchange ratio (RER) of mice fed ad libitum, or mice in SDR(10%-25%-65%) (l) or SDR(33.3%×3) (m) group. n = 7 and 6 biologically independent mice for AL in (l) and (m) respectively, n = 8 for SDR(10%-25%-65%), n = 6 for SDR(33.3%×3). n, o, Food intake of mice fed ad libitum, or mice in SDR(10%-25%-65%) (n) or SDR(33.3%×3) (o) group between the indicated zeitgeber times (n = 8 biologically independent mice per group). The data shown in (a, Fig. 2d) or (c, d) were performed simultaneously with a single control experiment. Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test. a, *P < 0.05; b, **P < 0.01; c, ***P < 0.001; versus AL group. Exact P values can be found in Source Data Extended Data Fig. 4.

Source data

a, Heat map of serum essential amino acids (EAA), semi-EAA and non-EAA identified by non-targeted metabolomics in mice feeding with 10%, 25% and 65% food from day 1 to day 3 (n = 4 biologically independent mice per group). b, Targeted quantification of serum semi-EAA and non-EAA content of mice as in (a) (n = 4 biologically independent mice for D0 and D4, n = 3 for D2). c, d, A high-protein diet (HP) maintained the decrease of fat mass percentage (c) and increase of lean mass percentage (d) induced by SDR with 10%, 25% and 65% food from day 1 to day 3. n = 18 biologically independent mice for AL-NP and SDR-NP, n = 9 for other groups. NP, normal-protein diet; LP, low-protein diet. e, f, Body weight (e) and body lean mass (f) of mice in (c). n = 18 biologically independent mice for AL-NP and SDR-NP, n = 9 for other groups. g, h, Representative images of isolated iWAT (g) and eWAT (h) from mice as treated with NP or high-protein diet after SDR in (c). i, The weight of eWAT in (h) (n = 10 biologically independent mice for D0, n = 6 for D4, D6, D9 and D12). j, Average adipocyte area of iWAT in (Fig. 3h) (n = 6 biologically independent samples per group). k, Representative images of H&E-stained sections of eWAT in (h). l, Average adipocyte area of eWAT in (k) (n = 6 biologically independent samples per group). m, Representative images of isolated iBAT from mice in (g). n, The weight of iBAT in (m) (n = 10 biologically independent mice for D0, n = 6 for D4, D6, D9 and D12). o, Representative images of H&E-stained sections of iBAT in (m). p, Representative images of H&E-stained sections of liver from mice in (g). q, Triglyceride (TG) content of liver from mice in (g) (n = 7 biologically independent mice for D0, n = 6 for D4, D6, D9 and D12). Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test. a, *P < 0.05; b, **P < 0.01; c, ***P < 0.001; except indicated, SDR-HP group versus AL-NP or SDR-NP group. Exact P values can be found in Source Data Extended Data Fig. 5.

Source data

a, b, A high-protein diet maintained the decrease of fat mass percentage (a) and the increase of lean mass percentage (b) induced by feeding with 33.3% food daily for 3 days (n = 9 biologically independent mice per group). c, d, Body weight (c) and body lean mass (d) of mice in (a) (n = 9 biologically independent mice per group). e, f, A high-protein diet maintained the decrease of fat mass percentage (e) and the increase of lean mass percentage (f) induced by feeding with 66.7% food daily for 12 days. n = 8 biologically independent mice for AL-NP, n = 9 for DR(66.7%×12)-NP and DR(66.7%×12)-HP. g, h, Body weight (g) and body lean mass (h) of mice in (e). n = 8 biologically independent mice for AL-NP, n = 9 for DR(66.7%×12)-NP and DR(66.7%×12)-HP. i, j, A high-protein diet maintained the decrease of fat mass percentage (i) and the increase of lean mass percentage (j) within a few days from the beginning of refeeding in female mice induced by feeding with 10%, 25% and 65% food from day 1 to day 3 (n = 10 biologically independent mice per group). k, l, Body weight (k) and body lean mass (l) of mice in (i) (n = 10 biologically independent mice per group). m, n, Cumulative food intake of male mice fed ad libitum (AL) or fed with NP or high-protein diet after dieting with 33.3% food daily for 3 days (m) or 66.7% food daily for 12 days (n). n = 9 and 8 biologically independent mice for AL-NP in (m) and (n) respectively, n = 9 for other groups. o, Cumulative food intake of female mice fed ad libitum or fed with NP or high-protein diet after SDR with 10%, 25% and 65% food from day 1 to day 3 (n = 10 biologically independent mice per group). p, Daily food intake of male mice fed ad libitum or fed with NP or high-protein diet after SDR with 10%, 25% and 65% food from day 1 to day 3. n = 18 biologically independent mice for AL-NP and SDR(10%-25%-65%)-NP, n = 9 for SDR(10%-25%-65%)-HP. q-s, Daily food intake of mice in (m-o). n = 9, 8 and10 biologically independent mice for AL-NP in (q), (r) and (s) respectively, n = 10 for SDR(10%-25%-65%)-NP or SDR(10%-25%-65%)-HP in female mice, n = 9 for other groups. Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test. a, P < 0.05; b, P < 0.01; c, P < 0.001; SDR-HP group versus AL-NP or SDR-NP group, DR-HP group versus AL-NP or DR-NP group. Exact P values can be found in Source Data Extended Data Fig. 6.

Source data

a, Composition of normal-protein (NP), high-protein (HP) and high-fat (HF) diets represented in percent calories. b, Food intake of mice fed ad libitum, mice refeeding with NP or HP diet after SDR and mice pair-fed with the isocaloric NP or HF diet after SDR. n = 11 biologically independent mice for AL-NP, SDR-NP and SDR-NP(PF), n = 10 for SDR-HP and SDR-HF(PF). c, d, Pair-feeding with normal-protein or high-fat diet after SDR markedly increased body fat percentage (c) and decreased lean mass percentage (d). n = 11 biologically independent mice for AL-NP, SDR-NP and SDR-NP(PF), n = 10 for SDR-HP and SDR-HF(PF). e, f, Body weight (e) and body lean mass (f) of mice in (b). n = 11 biologically independent mice for AL-NP, SDR-NP and SDR-NP(PF), n = 10 for SDR-HP and SDR-HF(PF). g, Heat map of differentially expressed genes associated with fatty acid and triglyceride biosynthesis in iWAT from mice fed normal-protein diet or high-protein diet after SDR with 10%, 25% and 65% food from day 1 to day 3 (n = 3 biologically independent samples per group). The average log2 value of each gene in D0 (NP) group was set to 0. Samples in D4 (NP/HP) or D6 (NP/HP) groups were shown in triplicate and normalized to the average expression level in D0 group. Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test. a, P < 0.05; b, P < 0.01; c, P < 0.001; SDR-NP(PF) or SDR-HF(PF) group versus SDR-HP group. Exact P values can be found in Source Data Extended Data Fig. 7.

Source data

a, Experimental design of mice gavaged with or without a mixture of antibiotics (ABX). Arrows show the time points of water or ABX gavage, feces collection, BODIPY-labeled fatty acid gavage and sampling. SPF, specific pathogen-free. b, Gel electrophoresis image of fecal DNA in mice from regimens in (a) (n = 5 biologically independent samples per group). c, Relative bacterial abundance determined by qPCR of 16S rDNA of feces in (b) (n = 5 biologically independent samples per group). d, Representative fluorescent section images of small intestinal villi in mice from regimens in (a) after oral gavage with BODIPY-labeled fatty acid for 2 h at D5. e, f, Small intestine (e) and serum (f) BODIPY concentrations of mice in (d) (n = 8 biologically independent mice for SDR, n = 7 for SDR + ABX). g, Representative fluorescent section images of iWAT and eWAT from mice in (d). h, iWAT and eWAT BODIPY concentrations of mice in (g) (n = 8 biologically independent mice for SDR, n = 7 for SDR + ABX). i, Treatment with antibiotic mixture ameliorated fat mass accumulation induced by SDR with 10%, 25% and 65% food from day 1 to day 3 (n = 9 biologically independent mice per group). j-n, Food intake (j), fat mass percentage (k), lean mass percentage (l), body weight (m) and body lean mass (n) of mice in (i) (n = 9 biologically independent mice per group). o, p, Penicillin treatment ameliorated the increase of fat mass percentage (o) and the decrease of lean mass percentage (p) induced by SDR with 10%, 25% and 65% food from day 1 to day 3 (n = 10 biologically independent mice for SDR, n = 12 for SDR-PEN). q, r, Body weight (q) and body lean mass (r) of mice in (o) (n = 10 biologically independent mice for SDR, n = 12 for SDR-PEN). s, t, The abundance of Lam-1 (s) and total bacterial load (t) in feces of mice in (o) (n = 7, 9 and 9 biologically independent samples for SDR on D0, D4 and D6 respectively, n = 9, 9 and 8 for SDR-PEN on D0, D4 and D6 respectively). Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test. a, *P < 0.05; b, **P < 0.01; c, ***P < 0.001. NS, not significant. Exact P values can be found in Source Data Extended Data Fig. 8.

Source data

a, Experimental design of fecal microbiota transplantation (FMT) from the donor mice into recipient mice. Fecal microbiota (FM) from mice in AL-NP, SDR-NP or SDR-HP group were transplanted into SPF recipient mice. AL, ad libitum; NP, normal protein diet; HP, high-protein diet; SDR, short-term dietary restriction with 10%, 25% and 65% food from day 1 to day 3. b, The weight of iWAT and eWAT of FMT recipient mice from regimens in (a) after oral gavage with olive oil containing BODIPY-labeled fatty acid for 2 h (n = 6 biologically independent samples per group). c, Genus-level proportional abundance of microbiota analyzed by fecal 16S rDNA sequencing of gnotobiotic (GB) mice obtained from germ-free mice gavaged with 8 indicated bacterial strains (n = 18 biologically independent mice). d, Fecal BODIPY concentrations of SPF mice administrated intragastrically with vehicle or Lam-1 after oral gavage with olive oil containing BODIPY-labeled fatty acid for 10 min to 2 h (n = 8 biologically independent mice per group). e, Representative fluorescent section images of small intestinal villi from mice in (d) after BODIPY gavage for 2 h. f, g, Small intestine (f) and serum (g) BODIPY concentrations of mice in (e). n = 8 biologically independent mice for SPF + Vehicle, n = 8 and 7 biologically independent mice for SPF + Lam-1 in (f) and (g) respectively. h, Representative fluorescent section images of iWAT and eWAT from mice in (e). i, BODIPY concentrations of iWAT and eWAT in (h) (n = 8 biologically independent mice per group). j, k, Oral gavage of Lam-1 markedly increased body fat mass percentage (j) and decreased lean mass percentage (k) (n = 9 biologically independent mice for SPF + Vehicle and SPF + LC2W, n = 8 for SPF + Lam-1). l, m, Body weight (l) and body lean mass (m) of mice in (j) (n = 9 biologically independent mice for SPF + Vehicle and SPF + LC2W, n = 8 for SPF + Lam-1). Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test. *P < 0.05; **P < 0.01; ***P < 0.001. NS, not significant. Exact P values can be found in Source Data Extended Data Fig. 9.

Source data

a, Food intake of mice after gavage with water, DL-3-phenyllactic acid (PLA), 4-hydroxyphenyllactic acid (HPLA) or indole-lactic acid (ILA) for 24 h (n = 6 biologically independent mice per group). b, Fecal BODIPY concentrations of mice in (a) after oral gavage with olive oil containing BODIPY-labeled fatty acid for 10 min to 2 h (n = 6 biologically independent mice per group). c, BODIPY concentrations of small intestine from mice in (b) after oral gavage for 2 h (n = 6 biologically independent mice per group). d, BODIPY concentrations of iWAT and eWAT from mice in (c) (n = 6 biologically independent mice per group). e, i, m, Food intake of mice after gavage with water or solution containing the 3 compounds (compds) (PLA, HPLA and ILA) (e), 2-hydroxyisocaproic acid (HICA) (i) or 2-hydroxy-3-methylbutyric acid (HMBA) (m) for 24 h (n = 7 biologically independent mice for Water or 3 Compds, n = 8 for HICA or HMBA). f, j, n, Fecal BODIPY concentrations of mice in (e), (i) or (m) after oral gavage with olive oil containing BODIPY-labeled fatty acid for 10 min to 2 h (n = 7 biologically independent mice for Water or 3 Compds, n = 8 for HICA or HMBA). g, k, o, BODIPY concentrations of small intestine from mice in (f), (j) or (n) after oral gavage for 2 h (n = 7 biologically independent mice for Water or 3 Compds, n = 8 for HICA or HMBA). h, l, p, BODIPY concentrations of iWAT and eWAT from mice in (f), (j) or (n) after oral gavage for 2 h (n = 7 biologically independent mice for Water or 3 Compds, n = 8 for HICA or HMBA). q, Food intake of mice after gavage with water or solution containing the 5 compounds (PLA, HPLA, ILA, HICA and HMBA) for 24 h (n = 10 biologically independent mice per group). r, Food intake of mice after gavage with water, glutamic acid or glucose for 24 h (n = 8 biologically independent mice per group). s, Fecal BODIPY concentrations of mice in (r) after oral gavage with olive oil containing BODIPY-labeled fatty acid for 10 min to 2 h (n = 8 biologically independent mice per group). t, Representative fluorescent section images of small intestinal villi and BODIPY concentrations of small intestine from mice in (s) after oral gavage for 2 h (n = 8 biologically independent mice per group). u, Serum BODIPY concentrations of mice in (t) (n = 8 biologically independent mice per group). v, Representative fluorescent section images and BODIPY concentrations of iWAT and eWAT from mice in (t) (n = 8 biologically independent mice per group). w, x, Oral gavage of solution containing five compounds (compds) including PLA, HPLA, ILA, HICA and HMBA markedly increased body fat mass percentage (w) and decreased lean mass percentage (x) (n = 8 biologically independent mice for Water, n = 9 for Compds). y, z, Body weight (y) and body lean mass (z) of mice in (w) (n = 8 biologically independent mice for Water, n = 9 for Compds). Data are presented as mean ± s.d. Statistical significance was determined by two-tailed Student's t-test. *, P < 0.05; **, P < 0.01; NS, not significant. Exact P values can be found in Source Data Extended Data Fig. 10.

Source data

Supplementary Tables 1 and 2.

Statistical Source Data.

Statistical Source Data.

Statistical Source Data.

Statistical Source Data.

Statistical Source Data.

Statistical Source Data.

Statistical Source Data.

Statistical Source Data.

Statistical Source Data.

Statistical Source Data.

Statistical Source Data.

Statistical Source Data.

Statistical Source Data.

Statistical Source Data.

Statistical Source Data.

Unprocessed gels.

Statistical Source Data.

Statistical Source Data.

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and Permissions

Zhong, W., Wang, H., Yang, Y. et al. High-protein diet prevents fat mass increase after dieting by counteracting Lactobacillus-enhanced lipid absorption. Nat Metab 4, 1713–1731 (2022). https://doi.org/10.1038/s42255-022-00687-6

Download citation

Received: 12 October 2021

Accepted: 14 October 2022

Published: 01 December 2022

Issue Date: December 2022

DOI: https://doi.org/10.1038/s42255-022-00687-6

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Nature Metabolism (2022)

Nature (2022)