ArticlePDF AvailableFasting-Refeeding Impacts Immune Cell Dynamics and Mucosal Immune ResponsesAugust 2019Cell 178(5):1072-1087.e14DOI:10.1016/j.cell.2019.07.047Authors: Motoyoshi NagaiMotoyoshi NagaiThis person is not on ResearchGate, or hasn t claimed this research yet. Ryotaro NoguchiRyotaro NoguchiThis person is not on ResearchGate, or hasn t claimed this research yet. Daisuke TakahashiKeio University Takayuki MorikawaTakayuki MorikawaThis person is not on ResearchGate, or hasn t claimed this research yet.Show all 23 authorsHide Download full-text PDFRead full-textDownload full-text PDFRead full-textDownload citation Copy link Link copied Read full-text Download citation Copy link Link copiedCitations (55)References (56)Figures (2)Abstract and FiguresNutritional status potentially influences immune responses; however, how nutritional signals regulate cellular dynamics and functionality remains obscure. Herein, we report that temporary fasting drastically reduces the number of lymphocytes by ∼50% in Peyer s patches (PPs), the inductive site of the gut immune response. Subsequent refeeding seemingly restored the number of lymphocytes, but whose cellular composition was conspicuously altered. A large portion of germinal center and IgA+ B cells were lost via apoptosis during fasting. Meanwhile, naive B cells migrated from PPs to the bone marrow during fasting and then back to PPs during refeeding when stromal cells sensed nutritional signals and upregulated CXCL13 expression to recruit naive B cells. Furthermore, temporal fasting before oral immunization with ovalbumin abolished the induction of antigen-specific IgA, failed to induce oral tolerance, and eventually exacerbated food antigen-induced diarrhea. Thus, nutritional signals are critical in maintaining gut immune homeostasis. Lymphocyte Dynamics in PPs and BM in Response to Fasting and Refeeding (A-D) Numbers of the indicated cell subsets were measured in PPs (A and B) and the right tibia and femur BM (C and D) of mice fasted for 36 h (blue background) and refed with CE2 (red background). 24, 36, and 60 h in (A) and (B), n = 12; the other datasets, n = 9. (E) The mean numbers of the indicated cells from mice fasted for 36 h or refed for 48 h were normalized by the value for mice fed ad libitum (each group, n = 9).…  Trafficking of Naive B Cells between PPs and the BM during Fasting (A) The experimental protocols for adoptive cell transfer of CTV-labeled PP cells. (B and C) Numbers of CTV + total (B, n = 12-13) and indicated cell subsets (C, n = 7-8) in the ad libitum-fed and fasting groups. (D) The experimental protocols for lymphocyte trafficking from PPs in KikGR mice. (E) Numbers of KikGR-Red + indicated cell subsets in the ad libitum-fed (n = 10) and fasting groups (n = 12). Data represent the means ± SEM. Student s t test. *p 0.05; **p 0.01; n.s., not statistically significant.… Figures - uploaded by Koji HaseAuthor contentAll figure content in this area was uploaded by Koji HaseContent may be subject to copyright. Discover the world s research20+ million members135+ million publications700k+ research projectsJoin for freePublic Full-text 1Content uploaded by Koji HaseAuthor contentAll content in this area was uploaded by Koji Hase on Oct 14, 2019 Content may be subject to copyright. ArticleFasting-Refeeding Impacts Immune Cell Dynamicsand Mucosal Immune ResponsesGraphical AbstractHighlightsdFasting drastically reduces lymphocyte levels in Payer’spatchesdNaive B cells migrate to bone marrow during fasting and thenback upon refeedingdNutritional signals are essential to maintain CXCL13expression by stromal cellsdFasting causes GC B cell death and attenuates antigen-specific IgA responseAuthorsMotoyoshi Nagai, Ryotaro Noguchi,Daisuke Takahashi, ..., Keiyo Takubo,Taeko Dohi, Koji HaseCorrespondencehase-kj@pha.keio.ac.jpIn BriefTemporary fasting drastically reduces thelevels of B cells in Peyer’s patches, withgerminal center B cells undergoingapoptosis and naive cells migrating to thebone marrow and only egressing uponrefeeding.Nagai et al., 2019, Cell 178, 1072–1087August 22, 2019 ª2019 Elsevier Inc.https://doi.org/10.1016/j.cell.2019.07.047 ArticleFasting-Refeeding Impacts Immune Cell Dynamicsand Mucosal Immune ResponsesMotoyoshi Nagai,1,2Ryotaro Noguchi,1,2Daisuke Takahashi,1Takayuki Morikawa,3Kouhei Koshida,1Seiga Komiyama,1Narumi Ishihara,1Takahiro Yamada,1Yuki I. Kawamura,2Kisara Muroi,1Kouya Hattori,1Nobuhide Kobayashi,1Yumiko Fujimura,1Masato Hirota,1Ryohtaroh Matsumoto,1Ryo Aoki,4,5Miwa Tamura-Nakano,6Machiko Sugiyama,2,7Tomoya Katakai,8Shintaro Sato,9,10Keiyo Takubo,3Taeko Dohi,1,2and Koji Hase1,10,11,*1Division of Biochemistry, Faculty of Pharmacy and Graduate School of Pharmaceutical Science, Keio University, Tokyo 105-8512, Japan2Department of Gastroenterology, Research Center for Hepatitis and Immunology, Research Institute, National Center for Global Health andMedicine, Chiba 272-8516, Japan3Department of Stem Cell Biology, Research Institute, National Center for Global Health and Medicine, Tokyo 162-8655, Japan4Division of Gastroenterology and Hepatology, Department of Internal Medicine, Keio University School of Medicine, Tokyo 160-8582, Japan5Institute of Health Sciences, Ezaki Glico Co., Ltd., Osaka 555-8502, Japan6Communal Laboratory, Research Institute, National Center for Global Health and Medicine, Tokyo 162-8655, Japan7Laboratory for Immunobiology, Graduate School of Medical Life Science, Yokohama City University, Kanagawa 230-045, Japan8Department of Immunology, Graduate School of Medical and Dental Sciences, Niigata University, Niigata 951-8510, Japan9Mucosal Vaccine Project, BIKEN Innovative Vaccine Research Alliance Laboratories, Research Institute for Microbial Diseases, OsakaUniversity, Osaka 565-0871, Japan10International Research and Development Center for Mucosal Vaccines, the Institute of Medical Science, the University of Tokyo (IMSUT),Tokyo 108-8639, Japan11Lead Contact*Correspondence: hase-kj@pha.keio.ac.jphttps://doi.org/10.1016/j.cell.2019.07.047SUMMARYNutritional status potentially influences immune re-sponses; however, how nutritional signals regulatecellular dynamics and functionality remains obscure.Herein, we report that temporary fasting drasticallyreduces the number of lymphocytes by 50% inPeyer’s patches (PPs), the inductive site of the gutimmune response. Subsequent refeeding seeminglyrestored the number of lymphocytes, but whosecellular composition was conspicuously altered. Alarge portion of germinal center and IgA+B cellswere lost via apoptosis during fasting. Meanwhile,naive B cells migrated from PPs to the bone marrowduring fasting and then back to PPs during refeedingwhen stromal cells sensed nutritional signals andupregulated CXCL13 expression to recruit naiveB cells. Furthermore, temporal fasting before oralimmunization with ovalbumin abolished the induc-tion of antigen-specific IgA, failed to induce oraltolerance, and eventually exacerbated food anti-gen-induced diarrhea. Thus, nutritional signals arecritical in maintaining gut immune homeostasis.INTRODUCTIONInappropriate calorie intake is a global health problem. In devel-oping countries, the nutritional deficiency often compromisesvaccination efficacy and increases the risk of infectious dis-eases (Kaufman et al., 2011; Savy et al., 2009; Scrimshaw andSanGiovanni, 1997). Furthermore, childhood malnutrition is apredisposing factor for environmental enteropathy characterizedby intestinal dysfunction, increased intestinal permeability, andmicrobial dysbiosis (Brown et al., 2015; Humphrey, 2009). Inindustrialized countries, on the other hand, excessive food intakeaccompanied by a lack of exercise has augmented the incidenceof obesity (World Health Organization, 2016), which is a signifi-cant risk factor for cardiovascular disease, metabolic syn-dromes, and cancer (Basen-Engquist and Chang, 2011; Grundy,2004; Poirier et al., 2006). Low-grade inflammation due toobesity is significantly implicated in the development of thesediseases (Visscher and Seidell, 2001). These observationsindicate that nutritional status has a significant impact on theimmune system.The gastrointestinal mucosa is directly exposed to exogenousfood ingredients and thus inevitably faces drastic changes in thenutritional status of the lumen during food uptake and fasting.We previously demonstrated that intestinal tissue is highly sus-ceptible to deprivation of luminal nutrients, as temporal fastingarrested epithelial cell proliferation while refeeding induced hy-perproliferation in the intestinal epithelium (Okada et al., 2013).Given that epithelial cell turnover constitutes a robust first-linebarrier to external antigens, mucosal barrier function may bemore vulnerable during fasting than during feeding. Consideringthat fasting relieves the burden of food-borne antigens andmicroorganisms on the gut mucosa, it is thus reasonable todecelerate epithelial cell turnover temporarily to minimize energyexpenditure under nutrient deprivation.The gut mucosal barrier consists of not only intestinal epithe-lium but also an underlying immune system that establishes thesecond-line barrier. The gut mucosal immune response is charac-terized by the production of dimeric or polymeric immunoglobulin1072 Cell 178, 1072–1087, August 22, 2019 ª2019 Elsevier Inc. ABCDEFigure 1. Characterization of PPs during Fasting and Refeeding(A) Serial sections of PP were stained with H E. PPs were obtained from mice fed ad libitum (left), fasted for 36 h (middle), or refed with CE2 for 48 h (right). Scalebar, 200 mm.(B) Immunostaining of PPs from mice fed ad libitum (left), mice fasted for 36 h (middle), and mice refed with CE2 for 48 h (right). Scale bar, 200 mm.(legend continued on next page)Cell 178, 1072–1087, August 22, 2019 1073 A (IgA) to the mucosal surface (Lycke and Bemark, 2017).Secretory IgA (S-IgA) plays vital roles in host defense againstpathogens, inhibition of microbial metabolite penetration, andregulation of the gut microbial community (Mantis et al., 2011;Uchimura et al., 2018; Wei et al., 2011). To efficiently induceS-IgA response, luminal antigens are actively taken to gut-associ-ated lymphoid tissue, such as Peyer’s patches (PPs), that serveas an inductive site of mucosal immunity. In PPs, germinal center(GC) reactions, namely, class switch recombination to IgA as wellas affinity maturation, occur continuously with the aid of follicularhelper T (Tfh) cells. IgA class-switched B cells subsequentlyegress PPs and then home to the intestinal lamina propria viamesenteric lymph nodes (MLNs), the thoracic duct, and bloodcirculation, during which IgA+B cells terminally differentiate intoIgA-producing plasma cells.Multiple lines of research have uncovered a link between im-mune cell function and metabolic status (Kau et al., 2011; Manand Kallies, 2015). For example, upon T cell receptor (TCR) stim-ulation, effector T (Teff) cells enhance the uptake and utilizationof glucose to promote aerobic glycolysis. Activated Teff cellsalso upregulate glutaminolysis. Such metabolic reprogrammingis essential for Teff cells to meet the energy demand of clonalexpansion and effector functions, such as the production ofinflammatory cytokines (Carr et al., 2010). Furthermore, IgA+plasma cells in the intestine preferentially utilize glycolysis for en-ergy metabolism, whereas naive B cells in PPs usually gain ATPthrough aerobic metabolism in mitochondria (Kunisawa et al.,2015). Stimulation with lipopolysaccharides (LPS) or B cell re-ceptor (BCR) ligation upregulates glucose transporter 1 (Glut1)expression in B cell activating factor (BAFF)-pretreated B cells,which eventually undergo metabolic reprogramming to glycol-ysis (Caro-Maldonado et al., 2014). Because B cell-specificGlut1 depletion leads to decreased B cell number and antibodyproduction, glycolytic rewiring is critical for B cell activation.Upregulation of Glut1 expression in activated B cells is primarilymediated by activation of the phosphatidylinositol 3-kinase(PI3K)-Akt pathway, which in turn enhances the mechanistictarget of rapamycin (mTOR) signaling. Excessive activationof the PI3K-Akt-mTOR pathway increases the frequencies ofTfh cells and GC B cells in PPs, while the opposite is true, asdisruption of mTORC1 or mTORC2 diminishes GC reactionsand production of S-IgA (Zeng et al., 2016). Given that mTORserves as a nutrient sensor, the whole-body nutritional statusin response to energy intake and starvation may considerablyaffect the immune response. Indeed, fasting or fasting-mimicdiet exerts a protective effect on bacterial sepsis and colitis byalleviating expression of proinflammatory cytokines (Okadaet al., 2017; Rangan et al., 2019), whereas glucose supplemen-tation protects against influenza viral infection (Wang et al.,2016). However, the impact of nutritional signals on mucosal bar-riers, lymphocyte dynamics, and effector functions in the contextof fasting and feeding remains unknown. Furthermore, most an-imal studies on calorie restriction and intermittent fasting haveemployed older mice, which may not adequately reflect biolog-ical responses during childhood and children susceptible tonutritional deficiency.In the present study, we analyzed the influence of fasting-refeeding on cellular dynamics and functionality of the gut im-mune system mainly in juvenile mice. We observed that PP lym-phocytes, particularly naive B cells, exhibit dynamic movement tothe bone marrow (BM) during temporal fasting and then swiftlymigrate back to PP in response to foodintake. A similar oscillationof naive B cells based on the circadian rhythm was observed to alesser extent in ad libitum-fed mice. Stromal CXCL13 expressionfor recruiting and retaining B cell subsets was regulated by meta-bolic status depending on aerobic glycolysis. The frequency ofmemory-like B cell subsets markedly decreased after fasting-refeeding treatment, leading to attenuation of antigen-specificIgA production and oral tolerance to a food antigen, which even-tually increased susceptibility to antigen-induced diarrhea.RESULTSFasting Has a Profound Effect on the Morphology of PPsTo explore the impact of nutrient deficiency on the gut immunesystem, we maintained juvenile mice (around 6 weeks old) underfasting conditions for 36 h. The numbers of total cells, B cells, andT cells in PPs decreased by half after fasting compared with thoseof ad libitum-fed mice (Figure S1A). Such a drastic change wasevident in PPs throughout the small intestine (Figure S1B). Immu-nofluorescent analysis demonstrated that the size of PPs wasmarkedly reduced during fasting and was restored in responseto refeeding. Despite this macroscopic alteration, the underlyingmicrostructure composed of B cell follicles and T cell regions wasnot disturbed during fasting and refeeding (Figures 1A and 1B).To explore the cause of decrease in the number of lympho-cytes in PPs of fasted mice, we analyzed apoptotic cells andobserved that fasting induced apoptosis in considerable numberof B cell follicles, which were rescued by refeeding (Figure 1C).Apoptotic cells were mainly detected in the GC region as alight-blue zone by hematoxylin counterstaining (Figure 1D).Furthermore, transmission electron microscopy (TEM) showeda higher frequency of apoptotic cells characterized by nuclearchromatin condensation and fragmentation (Figure 1E, upperright) as well as phagocytes engulfing apoptotic bodies in theGC region during fasting (Figure 1E, lower right). Thus, a largeportion of GC B cells, which include proliferating B cells andIgA class-switched B cells, were eliminated in PPs by cell deathfollowed by phagocytosis in fasted mice.Naive B Cells Circulate between PPs and BM inResponse to Nutritional StatusTotal cell number steeply declined with nutrient deprivation from24 to 36 h and then gradually recovered by refeeding for an(C and D) PPs were stained with TUNEL (C) or for cleaved caspase-3, counterstained with hematoxylin (D). PPs were obtained from mice fed ad libitum (left),fasted for 36 h (middle), or refed with CE2 for 48 h (right). Scale bar, 200 mm (upper panel), 50 mm (lower panel). In quantification of TUNEL assay, PPs wereobtained from the ad libitum (n = 7), fasting (n = 6), or refeeding group (n = 8).(E) Sections of PPs were observed by TEM. Scale bar, 10 mm (upper left and lower), 2.0 mm (upper right).Data represent the means ±SEM. ANOVA followed by Tukey’s test (C). *p 0.05; **p 0.01.1074 Cell 178, 1072–1087, August 22, 2019 AC (h)E (h)BD (h)(h)Fasting RefeedingTotal cellB220+ cellnaive B cellimmature B cellCD4+ T cellCD8+ T cellGranulocyteMacrophageGMPMEPCMPCLPMPP(CD34Flt3)ST-HSC(CD34Flt3)LT-HSC(CD34Flt3)MPP2(SLAM)MPP3(SLAM)MPP4(SLAM)ST-HSC(SLAM)LT-HSC(SLAM) 01234Total cells (x107)**04824 72012345B220+ cells (x107)**01.50.00.51.02.04824 72IgM+ IgD+ Naive B cells (x106)****002464824 72IgM- IgD-Immature B cells (x106)****002468104824 72**Total cells (x107)** ****02448720.00.51.01.52.02.596B cells (x107)** ****0.00.51.01.52.0024487296*T cells (x106)*****01234024487296CD95+ GL7+GC B cells (x106)********0.00.51.01.52.02.5024487296**IgM+ IgD+ Naive B cells (x107)****0.00.51.01.5024487296IgA+ B cells (x106)**** **0.00.51.01.52.02.5024487296*Ki67+ B cells (x107)0.00.51.01.5****04824 7204812162024Naive B cells (x106)****0.50.01.51.02.00 4 8 12162024Naive B cells /PP # (x106)0.81.0n.s.1.21.41.6IgMZT0 ZT8 ZT16IgDF(ZT)(ZT)GB220CD3εBM PPB220+CD3ε- gated0.69 0.04542.756.51.56 0.06141.057.40.71 0.07244.055.20.94 6.0622.370.71.81 14.026.058.21.18 8.8824.165.8Figure 2. Lymphocyte Dynamics in PPs and BM in Response to Fasting and Refeeding(A–D) Numbers of the indicated cell subsets were measured in PPs (A and B) and the right tibia and femur BM (C and D) of mice fasted for 36 h (blue background)and refed with CE2 (red background). 24, 36, and 60 h in (A) and (B), n = 12; the other datasets, n = 9.(E) The mean numbers of the indicated cells from mice fasted for 36 h or refed for 48 h were normalized by the value for mice fed ad libitum (each group, n = 9).(legend continued on next page)Cell 178, 1072–1087, August 22, 2019 1075 additional 72 h reflecting the histological observations (Figures1A, 1B, and 2A). A similar tendency was observed for total B cells(Figure 2A); however, cell behavior was different among theB cell subpopulations. IgM+IgD+naive B cells were restored topre-fasting (healthy) levels within 72 h after refeeding, whereasCD95+GL7+GC and IgA+B cells failed to recover (Figure 2B).Consequently, PPs were predominantly filled with naive B cellsafter the mice experienced fasting and refeeding. Of note, totalCD4+and CD8+T cells also decreased after 36 h of fastingand slowly increased with refeeding (Figure S1C), but becauseB cell subsets are deemed highly susceptible to food deprivationand intake, we focused on analyzing B cell dynamics.GC B cells were eliminated by apoptosis and would be newlyinduced in response to food intake, which explains the delayedrecovery during the refeeding period. On the other hand, therapid recovery of naive B cells during the refeeding stage raisesthe possibility that this cell subset may migrate to effector sites(i.e., intestinal lamina propria) and/or extraintestinal lymphoid tis-sue during fasting and then gain re-entry to PPs during refeeding.To examine this possibility, we carefully analyzed cell dynamicsin multiple tissues during fasting and refeeding. Fasting did notaffect cell numbers of cecal patches (CPs), the MLNs, or thesmall intestine lamina propria (SILP) (Figure S1A). The numberof splenic B cells, which mainly consisted of naive B cells, signif-icantly decreased during fasting and then recovered after 48 h ofrefeeding (Figure S1D). In sharp contrast, total BM cell numberincreased during fasting and decreased after refeeding (Fig-ure 2C). In particular, the naive B cell number expanded morethan 4-fold during fasting and then steeply declined to basallevels by refeeding (Figures 2D and 2E). These dynamics ofBM naive B cells were complementary to that of PP and splenicB cells (Figures 2B and S1D). The accumulation of BM naiveB cells did not result from enhanced B cell generation becauseimmature B (B220+IgMIgD) cells, as well as Ki67+B cells,decreased during fasting. Hematopoietic stem/progenitor cells(HS/PCs) also decreased significantly during fasting and rapidlyrecovered in response to refeeding (Figures 2D, 2E, and S2A;Table S1). Moreover, the immunofluorescent analysis revealedthat naive B cells were localized not only in the vascular regionbut also the BM cavity during fasting (Figures S2B and S2C).These data suggest that naive B cells most likely translocate tothe BM during nutrient deprivation.To rigorously confirm the bias of lymphocyte trafficking underfasting conditions, we adoptively transferred fluorescent dye-labeled PP cells from ad libitum-fed donors to either fasting orad libitum-fedrecipients (Figure3A). After 18 h, the transferred lym-phocytes preferentially migrated to the BM only when the recipientmice were fasted (Figure 3B). These BM-migrating cells weremainly naive B cells (Figure 3C). We also assessed lymphocytetraffickingwith knock-in mice carrying Kikume-Green Red (KikGR)(Tomura et al., 2014). We selectively irradiated PPs with a 430-nmlaser to induce photoconversion in PP lymphocytes before fasting(Figure 3D) and confirmed that PP naive B cells preferentiallymigrated to the BM, but not the spleen and MLN, during fasting(Figure 3E). Based on these observations, we considered thatnaive B cells might shuttle between the BM and PPs during thefasting and refeeding stages, respectively. Similar events wereobserved even under germ-free conditions (Figure S3), indicatingthat the gut microbiota is unlikely to contribute to the regulationof lymphocyte dynamics in response to fasting and refeeding.Notably, such B cell dynamics in response to food intake anddeprivation was well conserved at all life stages (Figure S4).Circadian Oscillation in Lymphocyte Trafficking WasObserved in Ad Libitum-Fed MiceThe feeding behavior of mice is distinct between the daytime andnighttime. Because mice are nocturnal, they do not feed duringthe day. We observed slight body weight change of 25.0 ±0.10 g and 23.6 ±0.10 g at zeitgeber time 0 (ZT0) and ZT12,respectively, in 6-week-old BALB/c male mice; body weight wasthen restored by nighttime feeding. Therefore, we assumed thatPP lymphocytes may exhibit circadian oscillation in response tonutritional status. Indeed, our time course analysis demonstratedthat the number of PP naive B cells slightly decreased until ZT12.Conversely, in the BM, naive B cells increased with a peak valuearound ZT12 and then gradually declined at ZT16–24 (Figures2F and 2G). Given that feeding behavior is minimal at ZT0–12and active at ZT12–24 under physiological conditions, lympho-cyte dynamics during daytime and nighttime were similarto thoseseen in the fasting-refeeding model, albeit to a lesser extent (Fig-ures 2A–2E). We considered that food intake plays a central role inthe circadian oscillations of naive B cells, and thus fasting disturbsthe oscillation by retaining this lymphocyte population in the BM.Nutritional Status Affects Chemokine Expression in PPand BMWe further explored the molecular basis controlling lymphocytedynamics in response to nutritional status. Chemokine-chemo-kine receptor interactions are critical in the regulation of immunecell trafficking. Among them, the CXCL13-CXCR5 axis is essen-tial for the migration and retention of B cell subsets in lymphoidtissues including PPs (Ansel et al., 2000). We observed a signif-icant decrease in Cxcl13 expression in PPs during fasting, whichwas recovered by refeeding (Figure 4A). Interestingly, the oppo-site expression pattern was observed in the BM. Although theCCL20-CCR6 axis is also indispensable for the maturation ofB cell follicles in PPs (Varona et al., 2001), Ccl20 expression re-mained stable during fasting (Figure 4B). These observations(F) Representative flow cytometry dot-plots of B220/CD3εgated (upper panels) and IgM/IgD gated on B220+CD3εB cells (lower panels) in the BM of mice fed adlibitum.(G) The circadian fluctuation in the number of naive B cells in PPs (upper) and the BM (lower). The mean number of PP cells was normalized by the number of PPs(each time point, n = 8).Data represent the means ±SEM. ANOVA followed by Dunnett’s test for comparison with mice fed ad libitum (A–D) or ZT0 (G). *p 0.05; **p 0.01. CMP,common myeloid progenitor; GMP, granulocyte/macrophage progenitor; MEP, megakaryocyte/erythrocyte progenitor; MPP, multipotent progenitor; ST/LT-HSC, short term/long term-hematopoietic stem cell.See also Figures S1,S2,S3, and S4 and Table S1.1076 Cell 178, 1072–1087, August 22, 2019 imply that differential CXCL13 expression in PPs and BM at leastpartly account for the localization of naive B cells during fastingand refeeding.The sphingosine 1-phosphate (S1P)-S1P receptor type 1(S1P1) axis promotes egress of lymphocytes from peripherallymphoid tissue into circulatory fluids under physiological condi-tions (Matloubian et al., 2004). However, treatment with FTY720failed to prevent the fasting-dependent naive B cell traffickingfrom PPs to the BM (Figure S5), indicating that the S1P-S1P1axis is likely not involved in regulating naive B cell dynamics inresponse to nutritional status.mTOR Signaling Partially Contributes to LymphocyteDynamicsTo gain mechanistic insight into lymphocyte dynamics, we as-sessed the activation status of Akt-mTOR signaling, a sensorABCDEFigure 3. Trafficking of Naive B Cells between PPs and the BM during Fasting(A) The experimental protocols for adoptive cell transfer of CTV-labeled PP cells.(B and C) Numbers of CTV+total (B, n = 12–13) and indicated cell subsets (C, n = 7–8) in the ad libitum-fed and fasting groups.(D) The experimental protocols for lymphocyte trafficking from PPs in KikGR mice.(E) Numbers of KikGR-Red+indicated cell subsets in the ad libitum-fed (n = 10) and fasting groups (n = 12).Data represent the means ±SEM. Student’s t test. *p 0.05; **p 0.01; n.s., not statistically significant.Cell 178, 1072–1087, August 22, 2019 1077 ACDEBFigure 4. Fasting Downregulates CXCL13 Expression in PPs Independent of mTOR Signaling(A) Cxcl12 and Cxcl13 mRNA expression in PPs (upper) and the BM (lower). PPs and the BM were obtained from mice fed ad libitum, fasted for 36 h, or refed withCE2 for 48 h (PP, each group, n = 24; BM, ad libitum, fasting, n = 14; refeeding, n = 8).(B) Ccl20 mRNA expression in PPs. PPs were obtained from mice fed ad libitum or fasted for 36 h (each group, n = 9).(C) Quantification of phosphorylated p70 S6 kinase (left) and Akt (right) in PPs (n = 18) and the BM (n = 6) using whole tissue lysates from ad libitum, fasting,refeeding, and rapamycin-treated groups.(legend continued on next page)1078 Cell 178, 1072–1087, August 22, 2019 and integrator of external nutritional stimuli for regulating cellularmetabolism and physiology, during fasting and refeeding. In thesame experimental setting, we also treated the ad libitum-fedgroup with rapamycin, which mainly inhibits mTORC1 as wellas mTORC2 to a lesser extent (Sarbassov et al., 2006), to deter-mine mTOR signaling dependency. As anticipated, fasting sup-pressed phosphorylation of Akt-mTOR signaling molecules,such as p70 S6 kinase, Akt, IRS-1, and PTEN, in both PPs andBM (Figures 4C and S6). Phosphorylation of these moleculesrecovered to normal or even higher (e.g., p70 S6 kinase in PPs)levels after 24-h refeeding compared with that of the controlad libitum-fed group (Figure 4C).Rapamycin treatment also decreased phosphorylation levelsof most Akt-mTOR signaling molecules in PPs and the BM,except for Akt in PPs (Figures 4C and S6). Notably, rapamycinsignificantly reduced the number of all B cell subsets in PPs (Fig-ure 4D). Consistent with previous studies (Limon and Fruman,2012; Zeng et al., 2016), rapamycin markedly decreased GCB cell number by less than 20% of vehicle control (Figure 4D).In the BM, rapamycin treatment decreased the levels of imma-ture B cells, but not naive B cells (Figure 4D). Meanwhile, rapa-mycin treatment did not alter expression levels of chemokines,including Cxcl12 and Cxcl13, in PPs and the BM (Figure 4E).Thus, mTOR signaling may be critical for B cell survival but notfor chemokine expression.CXCL13 Production by Stromal Cells Requires Warburg-like Aerobic GlycolysisTo further dissect the role of nutrient signals in the regulation ofchemokine expression, we focused on cellular metabolism. Injuvenile mice, fasting drastically changed systemic nutritionalstatus with hypoglycemia accompanied by elevated plasmab-hydroxybutyrate (BHB) levels (Figures 5A and 5B). We theo-rized that such nutritional changes may affect chemokineexpression in lymphoid tissue. To test this, we took advantageof the lymph node-derived stromal cell line BLS12, which abun-dantly produces CXCL13 upon stimulation with tumor necrosisfactor-a(TNF-a) and anti-lymphotoxin-breceptor (LTbR) antag-onistic antibodies (Katakai et al., 2008)(Figure 5C). We initiallyinvestigated the intracellular metabolic status of BLS12 cellsby detecting Glut1 and MitoSOX expression, which representfunctional markers for glycolysis and mitochondrial respiration,respectively (Kunisada et al., 2017). We found that exposure toTNF-aand anti-LTbR antibodies strongly skewed cellular meta-bolism toward glycolysis (Figure 5D). Correspondingly, glucosedeprivation in activated BLS12 cells prominently decreasedCxcl13 expressions (Figure 5E). Further, inhibition of glycolysisby 2-deoxy-D-glucose (2DG) markedly downregulated Cxcl13in activated BLS12 cells (Figure 5F) as well as in PP of micefed ad libitum in association with decreases in B cell subsets(Figures 5G and 5H). On the other hand, neither BHB nor rapa-mycin influenced chemokine expression in BLS12 cells (Figures5I and 5J). These results indicate that metabolic reprogramminginto Warburg-like aerobic glycolysis is a prerequisite for induc-tion of CXCL13 in stromal cells.Repeated Fasting Attenuates Antigen-Specific IgAResponse and Oral Tolerance Leading to Exacerbationof Food Antigen-Induced DiarrheaElimination of GC and IgA+B cells from PPs during fastingincreases the possibility of fasting-refeeding compromising im-mune responses against orally delivered antigens. To investigatethis possibility, mice were orally immunized with ovalbumin(OVA) and cholera toxin (CT) as an adjuvant once a week, fourtimes. In the fasting group, immunization was performed after48 h of refeeding that followed 36-h fasting (Figure 6A); thisregimen did not affect final body weight (Figure 6B). Oral immu-nization with OVA/CT gradually increased fecal OVA-specific IgAuntil day 32 in the ad libitum-fed control group; however, theOVA-specific IgA response was markedly attenuated in the fast-ing group (Figure 6C). Plasma OVA-specific IgA, IgM, and IgGwere also significantly decreased in the fasting group (Figure 6D).Thus, fasted mice failed to gain the booster effect of repeatedimmunization. Considering that a subset of GC B cells differenti-ates into memory B cells, the elimination of GC B cells from PPsby fasting may have resulted in this abnormality. Similarly, fastingattenuated the generation of antigen-specific IgA in response tooral infection with a recombinant Salmonella strain expressing atenuous toxoid fragment C (rSalmonella-ToxC) (Hase et al.,2009; VanCott et al., 1996)(Figure S7).Because antigen-specific IgA and IgG are considered a pro-tective factor against allergic symptoms (Aghamohammadiet al., 2009; Strait et al., 2006; Yamaki et al., 2014), we exploredthe effect of repeated fasting on an OVA-induced diarrhea model(Figure 7A). We found that fasting promoted the development ofdiarrhea without significant increase of OVA-specific plasma IgElevels (Figures 7B and 7C), and repeated fasting diminished in-duction of plasma OVA-specific IgG and fecal OVA-specificIgA on day 14 and 20, respectively (Figures 7D and 7E). TotalIgA in feces also significantly decreased in fasting group onday 14 (Figure 7E). These results suggest that impaired produc-tion of antigen-specific IgG and IgA, as well as total IgA, mayhave led to the exacerbation of diarrhea in the repeated fast-ing group.Furthermore, mice that were fasted before oral administrationof OVA failed to induce oral tolerance to OVA, as evidenced by anincrease in auricular swelling due to delayed-type hypersensitiv-ity after subcutaneous challenge (Figures 7F and 7G). In line withthis, suppression of systemic IgG response to the antigen injec-tion by oral administration of OVA was not observed in the fastingmice, indicative of impaired systemic unresponsiveness (Fig-ure 7H). Altogether, our findings indicate that exacerbation ofdiarrhea most likely resulted from disturbance in antigen-specificmucosal immune responses as well as tolerogenic responses,(D and E) Numbers of indicated cell subsets (D) and Cxcl12 and Cxcl13 mRNA expression (E) in PP and the BM of mice treated with vehicle (PBS) or rapamycin(each group, n = 8).Data represent the means ±SEM. ANOVA followed by Dunnett’s test (A) or Tukey’s test (C). Mann-Whitney U test (B and E). Student’s t test (D). *p 0.05;**p 0.01; n.s., not statistically significant.See also Figures S5 and S6.Cell 178, 1072–1087, August 22, 2019 1079 underscoring the essential role of nutrient signals in the mainte-nance of gut immune homeostasis by securing PP cellularity.DISCUSSIONOur findings demonstrated the dynamic behavior of lympho-cytes during fasting and refeeding. Fasting decreased thenumber of PP lymphocytes while refeeding selectively restorednaive—but not GC and IgA class-switched—B cells. GC B cellsunderwent massive apoptosis due to downregulation ofmTORC1 signaling during fasting, whereas PP-derived naiveB cells migrated into the BM in a CXCL13-dependentmanner—at least in part. CXCL13 expression by peripheralstromal cells was mostly dependent on glycolysis. Moreover,repeated fasting attenuated antigen-specific IgA responseand oral tolerance, which eventually exacerbated antigen-induced diarrhea.Under physiological conditions, a variety of food antigensaccompanied by opportunistic pathogens are continuouslydelivered to the intestinal mucosa. The intestinal mucosa in adulthumans possesses a total surface area of 200 m2. PPs conductimmunosurveillance on the mucosal surface to eliminatepotentially hostile agents. A preconceived notion is that the gutmucosal immune system constitutively induces immune re-sponses as evidenced by active GC reactions (Cesta, 2006).However, our findings revealed that immunological activity inPPs is nearly shut down during fasting where approximatelyhalf of lymphocytes egress PPs or undergo apoptotic cell death.B cells comprise a major population of PPs, where the B cell/Tcell ratio is 5-fold higher than in peripheral lymph nodes (Abbaset al., 2014). We found that B cell populations were highlysusceptible to food deprivation; however, the physiologicalsignificance of this phenomenon remains to be elucidated. Giventhat both luminal antigens and nutrient supply are greatlyFigure 5. Lymph Node-Derived Stromal Cells Depend on Glycolysis to Produce CXCL13(A and B) Plasma glucose (A) and plasma b-hydroxybutyrate (BHB) (B) concentrations (each group, n = 8).(C and D) BLS12 cells were stimulated with indicated stimulants for 24 h. CXCL13 transcripts were detected by qPCR (C, n = 6). The mitochondrial-specificproduction of ROS (%MitoSOX+) and surface GLUT1 expression were detected by flow cytometry (D, n = 3). Data are repres entative of two independentexperiments.(E) Cxcl13 expression in BLS12 cells cultured in control or glucose deprivation medium with TNF-aand anti-LTbR antibodies for 24 h. CXCL13 transcripts weredetected by qPCR (n = 6).(F, I, and J) Cxcl13 expression in BLS12 cells treated with 2DG (F), BHB (I), and rapamycin (J), in the presence of TNF-aand anti-LTbR antibodies for 24 h, weredetected by qPCR (n = 6).(G and H) Cxcl13 expression (G) and numbers of indicated cell subsets (H) in PPs of mice treated with vehicle (PBS) or 2DG (each group, n = 8).Data represent the means ±SEM. ANOVA followed by Tukey’s test (A and B) or Dunnett’ s test (C, D, F, and J). Mann-Whitney U test (E, G, and I). Student’s t test(H). *p 0.05; **p 0.01; n.s., not statistically significant.ACDBFigure 6. Repeated Fasting Suppresses Orally Induced Antigen-Specific Immune Responses(A) Diagram illustrating the protocol for oral immunization with OVA and CT. Mice were fasted for 36 h and refed 48 h before immunization in the fasting-re-fed group.(B) Body weight of mice eating ad libitum or fasting-refed.(C and D) OVA-specific IgA titers in feces (C) and OVA-specific IgA, IgM, and IgG titers in plasma (D) on day 32 from mice eating ad libitum or fasting-refed (eachgroup, n = 10).Data represent the means ±SEM. Mann-Whitney Utest (C and D). *p 0.05; **p 0.01; n.s., not statistically significant.See also Figure S7.Cell 178, 1072–1087, August 22, 2019 1081 ABDCEFGH250 μl PBS100 μl CFA20 μl PBS (right ear)Figure 7. Repeated Fasting Exacerbates Food Antigen-Induced Diarrhea(A) The protocol for OVA-induced diarrheamodel. In the fasting group, mice were fasted for 36 h and refed 24 or 48 h before immunization. Related to Figures 7B–7E.(B) Fecal clinical scores. Mice were fed ad libitum or were fasting-refed (each group, n = 14).(C and D) Total and OVA-specific IgE (C), and IgG (D) in plasma (each group, n = 8).(E) Total and OVA-specific IgA titers in feces (each group, n = 14).(F–H) The protocol to assess oral tolerance (F). Mice were fed ad libitum or fasted for 48 h prior to oral administration with 25 mg OVA (tolerance) or PBS (control).Delayed-type hypersensitivity (G), and OVA-specific IgG titers in plasma (H) (each group, n = 8).Data represent the means ±SEM. Mann-Whitney U test (B, D, and E: OVA-specific IgG and IgA). Student’s t test (C–E: total IgE, IgA, and IgG, and OVA-specificIgE). ANOVA followed by Tukey’s test (G) or Dunnett’s test (H). *p 0.05; **p 0.01; n.s., not statistically significant.1082 Cell 178, 1072–1087, August 22, 2019 reduced during the fasting period, diminution of the lymphocytepool may minimize energy expenditure. Despite hypoplasticmorphology under fasting conditions, the fundamental micro-structures of PPs remained intact and showed a rapid restora-tion in response to refeeding. Such plasticity based on dynamicB cell movement characterizes PPs as mucosa-associatedlymphoid tissue. From another point of view, it is possible thatthe hypoplastic status is the default of PPs devoid of nutritionalsignals. Daily food intake most likely confers immunological ac-tivity to PPs by recruiting lymphocytes, mainly naive B cells.The recirculation of naive B cells between PPs and the BM alsooccurred as circadian oscillation under physiological conditions.In mouse lymph nodes, noradrenalin-dependent b2-adrenergicstimuli at night upregulate CCR7 and CXCR4 on lymphocytesto suppress cell egress from lymph nodes (Suzuki et al., 2016).Accordingly, lymphocytes accumulate in the lymph node at nightand circulate in the blood during the day due to decreasednoradrenaline levels. It remains an open question whether thismechanism also contributes to the circadian oscillation of PPnaive B cells. However, considering that fasting abrogated recir-culation of naive B cells between PP and the BM, food intakeshould serve as a primary element in the regulation of PP naiveB cell dynamics. Indeed, the naive B cell population rapidlyexpanded in response to refeeding; this rapid recovery cannotbe explained by enhanced B cell generation, given that thereare multiple steps of differentiation from HSCs into naive B cells,via pro-B, pre-B, immature B, and transitional B cell stages(Allman and Pillai, 2008; Shapiro-Shelef and Calame, 2005).Because naive B cells accumulate in the BM during fasting, weconsidered that the BM may function as a reservoir of naiveB cells to rapidly release the cells to mucosa-associatedlymphoid tissue in response to food intake. In support of thisview, histological analysis of the BM of fasted mice detectedan accumulation of naive B cells in the BM cavity, especially inthe vicinity of blood vessels. Such a perivascular niche is alsoknown as a site for HSC differentiation and proliferation(Oh and No¨r, 2015), suggesting that this region can establish amicroenvironment rich in cell survival factors (e.g., BAFF andgrowth factors) for naive B cells (Schweighoffer and Tybulewicz,2018; Zhang et al., 2004).The Akt-mTOR signaling pathway is known to convergevarious external signals, including growth factors, insulin,glucose, and amino acids (Laplante and Sabatini, 2012; Saxtonand Sabatini, 2017). Induction of GC B cells largely dependson mTOR signaling (Ersching et al., 2017). Over-activation ofPI3K-Akt-mTOR and inhibition of mTOR exert positive and nega-tive effects on GC B in PPs, respectively (Zeng et al., 2016).Furthermore, B cell-specific deletion of Rictor, the core subunitof mTORC2, affects the survival and proliferation of B cells(Lee et al., 2013). Meanwhile, conditional deletion of a coremTORC1 protein in activated B cells arrests GC B cell differenti-ation, leading to a decrease in antigen-specific memory B cellsand plasma cells (Raybuck et al., 2018). In agreement with thesestudies, we also observed that rapamycin treatment prominentlyreduced GC B cells in PPs. Because fasting mitigated mTORactivity in PPs, we consider that the massive cell death of GCB cells during fasting is attributed to the downregulation ofmTORC1 signaling. Furthermore, GC B cells did not recoverrapidly by refeeding and PPs were mainly replenished by naiveB cells. Accordingly, the production of antigen-specific IgA infeces after repeated oral immunization with OVA and CT signifi-cantly decreased when mice were fasted before immunization.These results, together with previous observations, indicatethat repeated fasting may eliminate antigen-specific memoryB cells due to downregulation of mTOR signaling, which eventu-ally attenuates the mucosal immune response.The importance of CXCL13 for the development and mainte-nance of lymph nodes, including PPs, has been well docu-mented (Ansel and Cyster, 2001; Ansel et al., 2000; Okadaet al., 2002). This chemokine is mainly expressed by marginalreticular cells (MRCs) in lymphoid tissue (Katakai et al., 2008).Recent single-cell transcriptome analysis demonstrated thatMRCs and follicular dendritic cells (FDCs) individually expressCXCL13; however, the fact that MRCs outnumber FDCs impli-cates MRCs as a significant source of CXCL13 (Rodda et al.,2018). BLS12 cells, which we used in this study to dissect themolecular basis of CXCL13 production, share many characteris-tics with MRCs. For instance, BLS12 cells express severaladherent molecules, including VCAM-1, ICAM-1, MAdCAM-1,and RANKL. These molecules are also highly upregulated inthe MRC-like network of mucosa-associated lymphoid tissues,such as PPs, nasopharynx-associated lymphoid tissues, andCPs (Katakai et al., 2008). Furthermore, activation of proteinkinase C and nuclear factor kB (NF-kB) pathways upon stimula-tion with TNF-aand anti-LTbR agonist antibodies mediatesCXCL13 expression in BLS12 cells (Katakai et al., 2008; Sutoet al., 2009). We found that activated BLS12 cells undergo meta-bolic reprogramming to aerobic glycolysis, which is critical forthe production of CXCL13. This metabolic shift from oxidativephosphorylation to aerobic glycolysis is well characterized inactivated M1 macrophages and monocytes as well as Th1 andTh17 cells (Kelly and O’Neill, 2015; Michalek et al., 2011; O’Neilland Hardie, 2013). Unexpectedly, activation of mTOR signalingwas found dispensable for metabolic reprogramming andCXCL13 production in MRCs, because rapamycin treatmentfailed to suppress both in in vitro and in vivo settings. Althoughrapamycin administration did not influence Akt phosphorylationin PPs, fasting markedly decreased phosphorylation levels. Aktsenses nutrients and upregulates not only the mTOR pathwaybut also other signaling pathways. For instance, Akt signalingactivates phosphofructokinase 2/fructose-2,6-bisphosphatase2 (PFKFB2), a key regulator of glycolysis (Novellasdemuntet al., 2013; Sreedhar et al., 2017). Notably, inhibition of glycol-ysis by 2DG or glucose deprivation significantly downregulatedCxcl13 expression. Therefore, we speculate that CXCL13expression by MRCs in PPs may be mediated by Akt signalingin an mTORC1-independent manner, although further investiga-tion is required to test this hypothesis.We also found that repeated fasting exacerbates food anti-gen-induced diarrhea in association with defects in antigen-specific IgA response and oral tolerance. In the gastrointestinaltract, dimeric IgA produced in the lamina propria is transportedto the mucosal surface by the action of polymeric Ig receptorsin intestinal epithelial cells (Corthe´sy, 2013; Pabst, 2012). Secre-tory IgA may prevent translocation of luminal antigens in thebody, and antigen-specific monomeric IgA in the blood preventsCell 178, 1072–1087, August 22, 2019 1083 anaphylaxis by competitively inhibiting the association of aller-gens with antigen-specific IgE on mast cells (Yamaki et al.,2014). Selective IgA deficiency (IGAD) is the most common pri-mary antibody deficiency, where IGAD patients frequentlydevelop allergic disorders, including asthma, atopic dermatitis,and food allergies (Aghamohammadi et al., 2009; Schafferet al., 1991). Moreover, allergen-specific IgG protects againstanaphylaxis and food allergy by neutralizing allergens andcross-linking to an inhibitory IgG receptor, FcgRIIB (Straitet al., 2006; Wagenaar et al., 2018). We consider that the exac-erbation of diarrhea in fasting mice may be caused by the atten-uated induction of antigen-specific IgA and IgG, together withinsufficient oral tolerance. The mechanism underlying the abnor-mality in oral tolerance remains unknown. Early works haveshown that PPs are essential for inducing oral tolerance toOVA (Fujihashi et al., 2001), although MLNs can also induceoral tolerance in the absence of PPs (Spahn et al., 2002). Severalmechanisms have been proposed to account for oral tolerance,which includes clonal deletion or anergy of allergen-specificT cells and the induction of regulatory T (Treg) cells (Pabst andMowat, 2012). Another study supports the central role ofeffector/memory-type Treg cells in the establishment of oraltolerance (Siewert et al., 2008). CD103+dendritic cells drivethe differentiation of Treg cells by secreting all-trans retinoicacid. Furthermore, dietary antigens are required to induce pe-ripheral induction of Treg cells in the small intestine (Kim et al.,2016). Fasting may thus affect the frequency of antigen-specificTreg and/or CD103+dendritic cells. Collectively, our findingsdemonstrate that nutritional stimuli are fundamental for main-taining gut immune homeostasis by facilitating antigen-specificIgA and oral tolerance.Multiple studies have defined the beneficial effects of fasting orcalorie restriction regarding metabolic diseases in overnutrition/obesity models. In these studies, time-restricted feeding, calorierestriction, fasting mimicking diets, and short/long-term fastingoptimize nutritional balance to prevent or ameliorate multiple dis-orders, such as metabolic disorders, cardiovascular disease, andautoimmune disorders (Brandhorst et al., 2015; Hatori et al., 2012;Okada et al., 2017; Wei et al., 2017). In sharp contrast, a time-restricted feeding regimen in juvenile mice exacerbated metabolicdisorders (Hu et al., 2019). This is analogous to our finding thatrepeated fasting in juvenile mice promoted food antigen-induceddiarrhea. These observations suggest that fasting and time-restricted feeding during growth may cause detrimental effectsdepending on the feeding regimens and the age of test animals.In conclusion, we found that food intake secures the integrityand function of the gut mucosal immune system through nutri-tional signaling. Nutritional deprivation impairs mucosal immunity,leading to immune barrier dysfunction and excessive allergicresponse. Our study uncovered a novel link between nutritionalsignals and immune cell dynamics and functionality. Furthermore,these findings may promote research and treatment courses forenhancing vaccine efficacy via dietary intervention.STAR+METHODSDetailed methods are provided in the online version of this paperand include the following:dKEY RESOURCES TABLEdLEAD CONTACT AND MATERIALS AVAILABILITYdEXPERIMENTAL MODEL AND SUBJECT DETAILSBMiceBCell culturedMETHOD DETAILSBPreparation of lymphocytes and BM cellsBFlow cytometryBAdoptive transfer of PP cellsBPhotoconversion of PPs from KikGR miceBOral immunization with OVABDetection of antibody responses by ELISABSalmonella infectionBInduction of oral toleranceBImmunofluorescenceBHistological analysisBTransmission electron microscopyBReverse transcription and quantitative PCRBBio-plex detection of phosphorylated proteinsBPlasma parametersdQUANTIFICATION AND STATISTICAL ANALYSISSUPPLEMENTAL INFORMATIONSupplemental Information can be found online at https://doi.org/10.1016/j.cell.2019.07.047.ACKNOWLEDGMENTSWe thank Yuuki Obata, Yutaka Nakamura, Naomi Hoshina, Hiroaki Shiratori,Yuma Kabumoto, Hiyori Tanabe, Seiji Minegishi, and Teruki Hagiwara fortechnical support as well as Michio Tomura, Heiichiro Udono and Yun-GiKim for their valuable discussion and technical consultation. This work wassupported by AMED-Crest (16gm1010004h0101 and 17gm1010004h0102;18gm1010004h0103 to K. Hase), the Japan Society for the Promotion of Sci-ence (17KT0055, 16H01369, 18H04680, 25293114 , and 26116709 to K. Hase),Keio Gijuku Academic Development Funds (to K. Hase), the SECOM Scienceand Technology Foundation (to K. Hase), the Takeda Science Foundation (toK. Hase.), the Science Research Promotion Fund, the Promotion and MutualAid Corporation for Private Schools of Japan (to K. Hase), Daiichi SankyoFoundation of Life Science (to K. Hase), Terumo Foundation for Life Scienceand Arts (to K. Hase), Nagase Science Technology Foundation (to K. Hase),The Tokyo Biochemical Research Foundation (to K. Has e), the National Centerfor Global Health and Medicine (26-110 and 30-1006 to Y.I.K), Yoshida Schol-arship Foundation (to M.N.), and Keio University Doctorate Student Grant-in-Aid Program (to M.N.).AUTHOR CONTRIBUTIONSM.N. and R.N. performed most of the experiments and data analysis. M.N.wrote the manuscript. D.T., K.K., S.K., N.I., T.Y., Y.I.K., M.H., R.M., and M.S.helped with animal experiments. K. Hattori and R.A. performed GF mouse ex-periments. D.T. and K.M. performed KikGR mouse experiments. N.K. and Y.F.performed infection experiments. M.T.-N. performed electron microscopyanalysis. T.Y. created the graphical abstract. Y.I.K. provided experimental re-sources and discussed data. T.K. provided the cell line. S.S. provided tetanustoxin. T.M. and K.T. analyzed the bone marrow data. T.D. conceived this study.K. Hase supervised the study. T.D. and K. Hase interpreted the data andrevised the manuscript.DECLARATION OF INTERESTSThe authors declare no competing interests.1084 Cell 178, 1072–1087, August 22, 2019 Received: December 5, 2018Revised: April 30, 2019Accepted: July 25, 2019Published: August 22, 2019REFERENCESAbbas, A.K., Lichtman, A.H., and Pillai, S. (2014). Cellular and molecular immu-nology, Eighth Edition (Elsevier), pp. 299–323.Aghamohammadi, A., Cheraghi, T., Gharagozlou, M., Movahedi, M., Rezaei,N., Yeganeh, M., Parvaneh, N., Abolhassani, H., Pourpak, Z., and Moin, M.(2009). IgA deficiency: correlation between clinical and immunological pheno-types. J. Clin. Immunol. 29, 130–136.Allman, D., and Pillai, S. (2008). Peripheral B cell subsets. Curr. Opin. Immunol.20, 149–157.Ansel, K.M., and Cyster, J.G. (2001). Chemokines in lymphopoiesis andlymphoid organ development. Curr. Opin. Immunol. 13, 172–179.Ansel, K.M., Ngo, V.N., Hyman, P.L., Luther, S.A., Fo¨rster, R., Sedgwick, J.D.,Browning, J.L., Lipp, M., and Cyster, J.G. (2000). A chemokine-driven positivefeedback loop organizes lymphoid follicles. Nature 406, 309–314.Basen-Engquist, K., and Chang, M. (2011). Obesity and cancer risk: recent re-view and evidence. Curr. Oncol. Rep. 13, 71–76.Brandhorst, S., Choi, I.Y., Wei, M., Cheng, C.W., Sedrakyan, S., Navarrete, G.,Dubeau, L., Yap, L.P., Park, R., Vinciguerra, M., et al. (2015). A Periodic Dietthat Mimics Fasting Promotes Multi-System Regeneration, Enhanced Cogni-tive Performance, and Healthspan. Cell Metab. 22, 86–99.Brown, E.M., Wlodarska, M., Willing, B.P., Vonaesch, P., Han, J., Reynolds,L.A., Arrieta, M.C., Uhrig, M., Scholz, R., Partida, O., et al. (2015). Diet and spe-cific microbial exposure trigger features of environmental enteropathy in anovel murine model. Nat. Commun. 6, 7806.Caro-Maldonado, A., Wang, R., Nichols, A.G., Kuraoka, M., Milasta, S., Sun,L.D., Gavin, A.L., Abel, E.D., Kelsoe, G., Green, D.R., and Rathm ell, J.C.(2014). Metabolic reprogramming is required for antibody production that issuppressed in anergic but exaggerated in chronically BAFF-exposed B cells.J. Immunol. 192, 3626–3636.Carr, E.L., Kelman, A., Wu, G.S., Gopaul, R., Senkevitch, E., Aghvanyan, A.,Turay, A.M., and Frauwirth, K.A. (2010). Glutamine uptake and metabolismare coordinately regulated by ERK/MAPK during T lymphocyte activation.J. Immunol. 185, 1037–1044.Cesta, M.F. (2006). Normal structure, function, and histology of mucosa-associated lymphoid tissue. Toxicol. Pathol. 34, 599–608.Corthe´sy, B. (2013). Multi-faceted functions of secretory IgA at mucosal sur-faces. Front. Immunol. 4, 185.Ersching, J., Efeyan, A., Mesin, L., Jacobsen, J.T., Pasqual, G., Grabiner, B.C.,Dominguez-Sola, D., Sabatini, D.M., and Victora, G.D. (2017). Germinal CenterSelection and Affinity Maturation Require Dynamic Regulation of mTORC1Kinase. Immunity 46, 1045–1058.Fujihashi, K., Dohi, T., Rennert, P.D., Yamamoto, M., Koga, T., Kiyono, H., andMcGhee, J.R. (2001). Peyer’s patches are required for oral tolerance to pro-teins. Proc. Natl. Acad. Sci. USA 98, 3310–3315.Grundy, S.M. (2004). Obesity, metabolic syndrome, and cardiovascular dis-ease. J. Clin. Endocrinol. Metab. 89, 2595–2600.Hase, K., Kawano, K., Nochi, T., Pontes, G.S., Fukuda, S., Ebisawa, M., Kado-kura, K., Tobe, T., Fujimura, Y., Kawano, S., et al. (2009). Uptake throughglycoprotein 2 of FimH(+) bacteria by M cells initiates mucosal immuneresponse. Nature 462, 226–230.Hatori, M., Vollmers, C., Zarrinpar, A., DiTacchio, L., Bushong, E.A., Gill, S.,Leblanc, M., Chaix, A., Joens, M., Fitzpatrick, J.A.J., et al. (2012). Time-restricted feeding without reducing caloric intake prevents metabolic diseasesin mice fed a high-fat diet. Cell Metab. 15, 848–860.Hu, D., Mao, Y., Xu, G., Liao, W., Ren, J., Yang, H., Yang, J., Sun, L., Chen, H.,Wang, W., et al. (2019). Time-restricted feeding causes irreversible metabolicdisorders and gut microbiota shift in pediatric mice. Pediatr. Res. 85, 518–526.Humphrey, J.H. (2009). Child undernutrition, tropical enteropathy, toilets, andhandwashing. Lancet 374, 1032–1035.Katakai, T., Hara, T., Sugai, M., Gonda, H., and Shimizu, A. (2004). Lymph nodefibroblastic reticular cells construct the stromal reticulum via contact withlymphocytes. J. Exp. Med. 200, 783–795.Katakai, T., Suto, H., Sugai, M., Gonda, H., Togawa, A., Suematsu, S., Ebi-suno, Y., Katagiri, K., Kinashi, T., and Shimizu, A. (2008). Organizer-likereticular stromal cell layer common to adult secondary lymphoid organs.J. Immunol. 181, 6189–6200.Kau, A.L., Ahern, P.P., Griffin, N.W., Goodman, A.L., and Gordon, J.I. (2011).Human nutrition, the gut microbiome and the immune system. Nature 474,327–336.Kaufman, D.R., De Calisto, J., Simmons, N.L., Cruz, A.N., Villablanca, E.J.,Mora, J.R., and Barouch, D.H. (2011). Vitamin A deficiency impairs vaccine-elicited gastrointestinal immunity. J. Immunol. 187, 1877–1883.Kawamoto, T. (2003). Use of a new adhesive film for the preparation of multi-purpose fresh-frozen sections from hard tissues, whole-animals, insects andplants. Arch. Histol. Cytol. 66, 123–143.Kelly, B., and O’Neill, L.A. (2015). Metabolic reprogramming in macrophagesand dendritic cells in innate immunity. Cell Res. 25, 771–784.Kim, K.S., Hong, S.-W., Han, D., Yi, J., Jung, J., Yang, B.-G., Lee, J.Y., Lee, M.,and Surh, C.D. (2016). Dietary antigens limit mucosal immunity by inducingregulatory T cells in the small intestine. Science 351, 858–863.Kunisada, Y., Eikawa, S., Tomonobu, N., Domae, S., Uehara, T., Hori, S., Fur-usawa, Y., Hase, K., Sasaki, A., and Udono, H. (2017). Attenuation ofCD4+CD25+Regulatory T Cells in the Tumor Microenvironment by Metformin,a Type 2 Diabetes Drug. EBioMedicine 25, 154–164.Kunisawa, J., Sugiura, Y., Wake, T., Nagatake, T., Suzuki, H., Nagasawa, R.,Shikata, S., Honda, K., Hashimoto, E., Suzuki, Y., et al. (2015). Mode of Bioen-ergetic Metabolism during B Cell Differentiation in the Intestine Determines theDistinct Requirement for Vitamin B1. Cell Rep. 13, 122–131.Laplante, M., and Sabatini, D.M. (2012). mTOR signaling in growth control anddisease. Cell 149, 274–293.Lee, K., Heffington, L., Jellusova, J., Nam, K.T., Raybuck, A., Cho, S.H.,Thomas, J.W., Rickert, R.C., and Boothby, M. (2013). Requirement for Rictorin homeostasis and function of mature B lymphoid cells. Blood 122,2369–2379.Limon, J.J., and Fruman, D.A. (2012). Akt and mTOR in B Cell Activation andDifferentiation. Front. Immunol. 3, 228.Lycke, N.Y., and Bemark, M. (2017). The regulation of gut mucosal IgA B-cellresponses: recent developments. Mucosal Immunol. 10, 1361–1374.Man, K., and Kallies, A. (2015). Synchronizing transcriptional control of T cellmetabolism and function. Nat. Rev. Immunol. 15, 574–584.Mantis, N.J., Rol, N., and Corthe´sy, B. (2011). Secretory IgA’s complex roles inimmunity and mucosal homeostasis in the gut. Mucosal Immunol. 4, 603–611.Matloubian, M., Lo, C.G., Cinamon, G., Lesneski, M.J., Xu, Y., Brinkmann, V.,Allende, M.L., Proia, R.L., and Cyster, J.G. (2004). Lymphocyte egress fromthymus and peripheral lymphoid organs is dependent on S1P receptor 1. Na-ture 427, 355–360.Michalek, R.D., Gerriets, V.A., Jacobs, S.R., Macintyre, A.N., MacIver, N.J.,Mason, E.F., Sullivan, S.A., Nichols, A.G., and Rathmell, J.C. (2011). Cuttingedge: distinct glycolytic and lipid oxidative metabolic programs are essentialfor effector and regulatory CD4+ T cell subsets. J. Immunol. 186, 3299–3303.Novellasdemunt, L., Tato, I., Navarro-Sabate, A., Ruiz-Meana, M., Me´ndez-Lucas, A., Perales, J.C., Garcia-Dorado, D., Ventura, F., Bartrons, R., andRosa, J.L. (2013). Akt-dependent activation of the heart 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase (PFKFB2) isoenzyme by amino acids.J. Biol. Chem. 288, 10640–10651.O’Neill, L.A.J., and Hardie, D.G. (2013). Metabolism of inflammation limited byAMPK and pseudo-starvation. Nature 493, 346–355.Oh, M., and No¨r, J.E. (2015). The Perivascular Niche and Self-Renewal of StemCells. Front. Physiol. 6, 367.Cell 178, 1072–1087, August 22, 2019 1085 Okada, T., Ngo, V.N., Ekland, E.H., Fo¨rster, R., Lipp, M., Littman, D.R., andCyster, J.G. (2002). Chemokine requirements for B cell entry to lymph nodesand Peyer’s patches. J. Exp. Med. 196, 65–75.Okada, T., Fukuda, S., Hase, K., Nishiumi, S., Izumi, Y., Yoshida, M., Hagiwara,T., Kawashima, R., Yamazaki, M., Oshio, T., et al. (2013). Microbiota-derivedlactate accelerates colon epithelial cell turnover in starvation-refed mice.Nat. Commun. 4, 1654.Okada, T., Otsubo, T., Hagiwara, T., Inazuka, F., Kobayashi, E., Fukuda, S.,Inoue, T., Higuchi, K., Kawamura, Y.I., and Dohi, T. (2017). Intermittent fastingprompted recovery from dextran sulfate sodium-induced colitis in mice.J. Clin. Biochem. Nutr. 61, 100–107.Pabst, O. (2012). New concepts in the generation and functions of IgA. Nat.Rev. Immunol. 12, 821–832.Pabst, O., and Mowat, A.M. (2012). Oral tolerance to food protein. MucosalImmunol. 5, 232–239.Poirier, P., Giles, T.D., Bray, G.A., Hong, Y., Stern, J.S., Pi-Sunyer, F.X.,and Eckel, R.H.; American Heart Association; Obesity Committee of theCouncil on Nutrition, Physical Activity, and Metabolism (2006). Obesityand cardiovascular disease: pathophysiology, evaluation, and effect ofweight loss: an update of the 1997 American Heart Association ScientificStatement on Obesity and Heart Disease from the Obesity Committee ofthe Council on Nutrition, Physical Activity, and Metabolism. Circulation113, 898–918.Rangan, P., Choi, I., Wei, M., Navarrete, G., Guen, E., Brandhorst, S.,Enyati, N., Pasia, G., Maesincee, D., Ocon, V., et al. (2019). Fasting-Mimicking Diet Modulates Microbiota and Promotes Intestinal Regenera-tion to Reduce Inflammatory Bowel Disease Pathology. Cell Rep. 26,2704–2719.Raybuck, A.L., Cho, S.H., Li, J., Rogers, M.C., Lee, K., Williams, C.L., Shlom-chik, M., Thomas, J.W., Chen, J., Williams, J.V., and Boothby, M.R. (2018).B Cell-Intrinsic mTORC1 Promotes Germinal Center-Defining TranscriptionFactor Gene Expression, Somatic Hypermutation, and Memory B Cell Gener-ation in Humoral Immunity. J. Immunol. 200, 2627–2639.Rodda, L.B., Lu, E., Bennett, M.L., Sokol, C.L., Wang, X., Luther, S.A., Barres,B.A., Luster, A.D., Ye, C.J., and Cyster, J.G. (2018). Single-Cell RNASequencing of Lymph Node Stromal Cells Reveals Niche-Associated Hetero-geneity. Immunity 48, 1014–1028.Sarbassov, D.D., Ali, S.M., Sengupta, S., Sheen, J.H., Hsu, P.P., Bagley, A.F.,Markhard, A.L., and Sabatini, D.M. (2006). Prolonged rapamycin treatmentinhibits mTORC2 assembly and Akt/PKB. Mol. Cell 22, 159–168.Savy, M., Edmond, K., Fine, P.E.M., Hall, A., Hennig, B.J., Moore, S.E., Mulhol-land, K., Schaible, U., and Prentice, A.M. (2009). Landscape analysis of inter-actions between nutrition and vaccine responses in children. J. Nutr. 139,2154S–2218S.Saxton, R.A., and Sabatini, D.M. (2017). mTOR Signaling in Growth, Meta-bolism, and Disease. Cell 168, 960–976.Schaffer, F.M., Monteiro, R.C., Volanakis, J.E., and Cooper, M.D. (1991). IgAdeficiency. Immunodefic. Rev. 3, 15–44.Schmidt, T.H., Bannard, O., Gray, E.E., and Cyster, J.G. (2013). CXCR4promotes B cell egress from Peyer’s patches. J. Exp. Med. 210,1099–1107.Schweighoffer, E., and Tybulewicz, V.L. (2018). Signalling for B cell survival.Curr. Opin. Cell Biol. 51, 8–14.Scrimshaw, N.S., and SanGiovanni, J.P. (1997). Synergism of nutrition, infec-tion, and immunity: an overview. Am. J. Clin. Nutr. 66, 464S–477S.Shapiro-Shelef, M., and Calame, K. (2005). Regulation of plasma-cell develop-ment. Nat. Rev. Immunol. 5, 230–242.Siewert, C., Lauer, U., Cording, S., Bopp, T., Schmitt, E., Hamann, A., andHuehn, J. (2008). Experience-driven development: effector/memory-like al-phaE+Foxp3+ regulatory T cells originate from both naive T cells and naturallyoccurring naive-like regulatory T cells. J. Immunol. 180, 146–155.Spahn, T.W., Weiner, H.L., Rennert, P.D., Lu¨gering, N., Fontana, A.,Domschke, W., and Kucharzik, T. (2002). Mesenteric lymph nodes are criticalfor the induction of high-dose oral tolerance in the absence of Peyer’s patches .Eur. J. Immunol. 32, 1109–1113.Sreedhar, A., Petruska, P., Miriyala, S., Panchatcharam, M., and Zhao, Y.(2017). UCP2 overexpression enhanced glycolysis via activation of PFKFB2during skin cell transformation. Oncotarget 8, 95504–95515.Strait, R.T., Morris, S.C., and Finkelman, F.D. (2006). IgG-blocking antibodiesinhibit IgE-mediated anaphylaxis in vivo through both antigen interception andFc gamma RIIb cross-linking. J. Clin. Invest. 116, 833–841.Suto, H., Katakai, T., Sugai, M., Kinashi, T., and Shimizu, A. (2009). CXCL13production by an established lymph node stromal cell line via lymphotoxin-beta receptor engagement involves the cooperation of multiple signaling path-ways. Int. Immunol. 21, 467–476.Suzuki, K., Hayano, Y., Nakai, A., Furuta, F., and Noda, M. (2016). Adrenergiccontrol of the adaptive immune response by diurnal lymphocyte recirculationthrough lymph nodes. J. Exp. Med. 213, 2567–2574.Tomura, M., Hata, A., Matsuoka, S., Shand, F.H.W., Nakanishi, Y., Ikebuchi,R., Ueha, S., Tsutsui, H., Inaba, K., Matsushima, K., et al. (2014). Trackingand quantification of dendritic cell migration and antigen trafficking betweenthe skin and lymph nodes. Sci. Rep. 4, 6030.Uchimura, Y., Fuhrer, T., Li, H., Lawson, M.A., Zimmermann, M., Yilmaz, B.,Zindel, J., Ronchi, F., Sorribas, M., Hapfelmeier, S., et al. (2018). AntibodiesSet Boundaries Limiting Microbial Metabolite Penetration and the ResultantMammalian Host Response. Immunity 49, 545–559.e5.VanCott, J.L., Staats, H.F., Pascual, D.W., Roberts, M., Chatfield, S.N., Yama-moto, M., Coste, M., Carter, P.B., Kiyono, H., and McGhee, J.R. (1996). Regu-lation of mucosal and systemic antibody responses by T helper cell subsets,macrophages, and derived cytokines following oral immunization with liverecombinant Salmonella. J. Immunol. 156, 1504–1514.Varanasi, S.K., Donohoe, D., Jaggi, U., and Rouse, B.T. (2017). ManipulatingGlucose Metabolism during Different Stages of Viral Pathogenesis Can Haveeither Detrimental or Beneficial Effects. J. Immunol. 199, 1748–1761.Varona, R., Villares, R., Carramolino, L., Goya, I., Zaballos, A., Gutie´rrez, J.,Torres, M., Martı´nez-A, C., and Ma´rquez, G. (2001). CCR6-deficient micehave impaired leukocyte homeostasis and altered contact hypersensitivityand delayed-type hypersensitivity responses. J. Clin. Invest. 107, R37–R45.Visscher, T.L., and Seidell, J.C. (2001). The public health impact of obesity.Annu. Rev. Public Health 22, 355–375.Wagenaar, L., Bol-Schoenmakers, M., Giustarini, G., Vonk, M.M., van Esch,B.C.A.M., Knippels, L.M.J., Garssen, J., Smit, J.J., and Pieters, R.H.H.(2018). Dietary Supplementation with Nondigestible Oligosaccharides Re-duces Allergic Symptoms and Supports Low Dose Oral Immunotherapy in aPeanut Allergy Mouse Model. Mol. Nutr. Food Res. 62, e1800369.Wang, A., Huen, S.C., Luan, H.H., Yu, S., Zhang, C., Gallezot, J.D.,Booth, C.J., and Medzhitov, R. (2016). Opposing Effects of Fasting Meta-bolism on Tissue Tolerance in Bacterial and Viral Inflammation. Cell 166,1512–1525.Wei, M., Shinkura, R., Doi, Y., Maruya, M., Fagarasan, S., and Honjo, T. (2011).Mice carrying a knock-in mutation of Aicda resulting in a defect in somatichypermutation have impaired gut homeostasis and compromised mucosaldefense. Nat. Immunol. 12, 264–270.Wei, M., Brandhorst, S., Shelehchi, M., Mirzaei, H., Cheng, C.W., Budniak, J.,Groshen, S., Mack, W.J., Guen, E., Di Biase, S., et al. (2017). Fasting-mimicking diet and markers/risk factors for aging, diabetes, cancer, and car-diovascular disease. Sci. Transl. Med. 9, eaai8700.World Health Organization (2016). Report of the Commission on Ending Child-hood Obesity (World Health Organization), pp. 1–68.Yamaki, K., Nakashima, T., Miyatake, K., Ishibashi, Y., Ito, A., Kuranishi, A.,Taguchi, A., Morioka, A., Yamamoto, M., and Yoshino, S. (2014). IgA attenu-ates anaphylaxis and subsequent immune responses in mice: possible appli-cation of IgA to vaccines. Immunol. Res. 58, 106–117.Yamamoto, M., Fujihashi, K., Kawabata, K., McGhee, J.R., and Kiyono, H.(1998). A mucosal intranet: intestinal epithelial cells down-regulate intraepithe-lial, but not peripheral, T lymphocytes. J. Immunol. 160, 2188–2196.1086 Cell 178, 1072–1087, August 22, 2019 Yamazaki, S., Ema, H., Karlsson, G., Yamaguchi, T., Miyoshi, H., Shioda, S.,Taketo, M.M., Karlsson, S., Iwama, A., and Nakauchi, H. (2011). Nonmyelinat-ing Schwann cells maintain hematopoietic stem cell hibernation in the bonemarrow niche. Cell 147, 1146–1158.Zeng, H., Cohen, S., Guy, C., Shrestha, S., Neale, G., Brown, S.A., Cloer, C.,Kishton, R.J., Gao, X., Youngblood, B., et al. (2016). mTORC1 and mTORC2Kinase Signaling and Glucose Metabolism Drive Follicular Helper T Cell Differ-entiation. Immunity 45, 540–554.Zhang, M., Ko, K.H., Lam, Q.L.K., Lo, C.K.C., Xu, D.J.L., Shen, L.,Zheng,B.,Srivastava,G.,andLu,L.(2004).NovelfunctionofTNFcyto-kines in regulating bone marrow B cell survival. Cell. Mol. Immunol. 1,447–453.Cell 178, 1072–1087, August 22, 2019 1087 STAR+METHODSKEY RESOURCES TABLEREAGENT or RESOURCE SOURCE IDENTIFIERAntibodiesAnti-cleaved caspase-3 (Asp175) (polyclonal) Cell Signaling Technology Cat#9661; RRID: AB_2341188Anti-rat IgG (H+L) TRITC (polyclonal) Southern Biotech Cat#3030-03; RRID: AB_619945Anti-rat IgG (H+L) Alexa Fluor 633 (polyclonal) Thermo Fisher Scientific Cat#A21094; RRID: AB_2535749Anti-mouse TER-119 PerCP-Cy5.5 (TER-119) TONBO biosciences Cat#65-5921; RRID: N/AAnti-mouse Lymphotoxin beta R/TNFRSF3 (polyclonal) R and D systems Cat#AF1008; RRID: AB_354531Anti-mouse Ly6A/E PE-Cy7 (Sca-1) Biolegend Cat#122514; RRID: AB_756199Anti-mouse Ly-6G/Gr1 PerCP-Cy5.5 (RB6-8C5) eBioscience Cat#45-5931; RRID: AB_906247Anti-mouse Ki-67 PE-Cy7 (SolA15) eBioscience Cat#25-5698; RRID: AB_11220070Anti-mouse IgM PE (R6-60.2) BD PharMingen Cat#553409; RRID: AB_394845Anti-mouse IgM PE dozzle 594 (RMM-1) Biolegend Cat#406529; RRID: AB_2566585Anti-mouse IgM HRP (polyclonal) Southern Biotech Cat#1020-05; RRID: AB_619903Anti-mouse IgM Alexa Fluor 488 (polyclonal) Thermo Fisher Scientific Cat#A21042; RRID: AB_2535711Anti-mouse IgG HRP (polyclonal) Southern Biotech Cat#1030-05; RRID: AB_2619742Anti-mouse IgD APC (11-26c.2a) Biolegend Cat#405713; RRID: AB_10645480Anti-mouse IgA HRP (polyclonal) Southern Biotech Cat#1040-05; RRID: AB_2714213Anti-mouse IgA FITC(C10-3) BD PharMingen Cat#559354; RRID: AB_397235Anti-mouse GL7 Pacific Blue (GL7) Biolegend Cat#144614; RRID: AB_2563292Anti-mouse Endomucin (V.7C7.1) abcam Cat#ab106100; RRID: AB_10859306Anti-mouse CD95 FITC (Jo2) BD PharMingen Cat#554257; RRID: AB_395329Anti-mouse CD8aPerCP-Cy5.5 (53-6.7) TONBO biosciences Cat#65-0081; RRID: AB_2621882Anti-mouse CD48 FITC (HM48-1) Biolegend Cat#103404; RRID: AB_313019Anti-mouse CD45R/B220 V450 (RA3-6B2) BD Horizon Cat#560472; RRID: AB_1645276Anti-mouse CD45R/B220 PE (RA3-6B2) eBioscience Cat#12-0452; RRID: AB_465671Anti-mouse CD45R/B220 FITC (RA3-6B2) BD PharMingen Cat#553088; RRID: AB_394618Anti-mouse CD45R/B220 APC-eFluor 780 (RA3-6B2) eBioscience Cat#47-0452; RRID: AB_1518810Anti-mouse CD45R/B220 APC (RA3-6B2) eBioscience Cat#17-0452; RRID: AB_469395Anti-mouse CD45R/B220 (RA3-6B2) eBioscience Cat#14-0452; RRID: AB_467254Anti-mouse CD45R/B220 PerCP-Cy5.5 (RA3-6B2) TONBO biosciences Cat#65-0452; RRID: AB_2621892Anti-mouse CD45 BV510 (30-F11) Biolegend Cat#103137; RRID: AB_2561392Anti-mouse CD44 PE-Cy7 (IM7) eBioscience Cat#25-0441; RRID: AB_469623Anti-mouse CD4 PerCP-Cy5.5 (RM4-5) TONBO biosciences Cat#65-0042; RRID: AB_2621876Anti-mouse CD4 APC-eFluor 780 (GK1.5) eBioscience Cat#47-0041; RRID: AB_11218896Anti-mouse CD3εV500 (500A2) BD Horizon Cat#560771; RRID: AB_1937314Anti-mouse CD3εBV605 (145-2C11) Biolegend Cat#100351; RRID: AB_2565842Anti-mouse CD3εAPC (145-2C11) TONBO biosciences Cat#20-0031; RRID: AB_2621537Anti-mouse CD3ε(145-2C11) BD PharMingen Cat#550275; RRID: AB_393572Anti-mouse CD34 FITC (RAM34) eBioscience Cat#11-0341; RRID: AB_465020Anti-mouse CD31 (MEC 13.3) BD PharMingen Cat#550274; RRID: AB_393571Anti-mouse CD16/CD32 Alexa Fluor 700 (93) eBioscience Cat#56-0161; RRID: AB_493994Anti-mouse CD16/CD32 (2.4G2) TONBO biosciences Cat#70-0161; RRID: AB_2621487Anti-mouse CD150/SLAM PE (TC15-12F12.2) Biolegend Cat#115904; RRID: AB_313683Anti-mouse CD135 APC (A2F10) Biolegend Cat#135310; RRID: AB_2107050Anti-mouse CD127/IL-7RaPE (SB/119) Biolegend Cat#121111; RRID: AB_493510(Continued on next page)e1 Cell 178, 1072–1087.e1–e6, August 22, 2019 ContinuedREAGENT or RESOURCE SOURCE IDENTIFIERAnti-mouse CD11b PerC-/Cy5.5 (M1/70) TONBO biosciences Cat#65-0112; RRID: AB_2621885Anti-mouse CD117/c-kit APC-Cy7 (2B8) Biolegend Cat#105826; RRID: AB_1626278Anti-cleaved caspase-3 (Asp175) (polyclonal) Cell Signaling technology Cat#9661; RRID: AB_2341188Anti-mouse GULT1 Alexa Fluor 647 (EPR3915) abcam Cat#ab195020; RRID: AB_2783877Anti-hamster IgG (H+L) FITC (polyclonal) Southern Biotech Cat#6210-02; RRID:N/ABacterial and Virus StrainsrSalmonella–ToxC (DaroA, DaroD)VanCott et al., 1996 N/AChemicals, Peptides, and Recombinant ProteinsCholera toxin (CT) List Biological Laboratories Cat#100BTetanus toxoid (TT) BIKEN foundation N/AFreund’s Adjuvant, Complete (CFA) Sigma Aldrich Cat#F5881Imject Alum Adjuvant Thermo Fisher Scientific Cat#77161Alubmin, Chicken Egg (Ovalbumin) Grade V (OVA) Sigma Aldrich Cat#A5503; CAS: 9006-59-1Rapamycin LC Laboratories Cat#R-5000; CAS: 53123-88-9Fingolimod (FTY720) Cayman Chemical Company Cat#10006292; CAS: 162359-56-02-Deoxy-D-glucose (2DG) abcam Cat#ab142242; CAS: 154-17-63-Hydroxybutyric acid/b-hydroxybutyrate (BHB) Sigma Aldrich Cat#166898; CAS: 300-85-6Murine Tumor Necrosis Factor-a(TNF-a) Peprotech Cat#315-01A7-Aminoactinomycin D (7-AAD) TONBO Cat#13-6993Fixable Viability Dye eFluor 780 Thermo Fisher Scientific Cat#65-0865LIVE/DEAD Fixable Blue Dead Cell Stain Kit, for UV excitation Thermo Fisher Scientific Cat#L23105SYTOX Blue Dead Cell Stain, for flow cytometry Thermo Fisher Scientific Cat#S34857MitoSOX Red Mitochondrial Superoxide Indicator (MitoSOX-PE) Thermo Fisher Scientific Cat#M36008CellTrace Cell Proliferation Kits CellTrace Violet (CellTrace Violet) Thermo Fisher Scientific Cat#C34557Quetol 812 Nissin EM Cat#340Deoxyribonuclease I (DNase I) Sigma Aldrich Cat#DN25; CAS: 9003-98-9Collagenase Wako Cat#032-22364; CAS: 9001-12-1cOmplete, mini Protease Inhibitor Cocktail Roche Cat#04-693-124-001TRIzol Reagent Thermo Fisher Scientific Cat#15596026EagleTaq Universal Master Mix (ROX) Roche Cat#07-260-288-190Power SYBR Green PCR Master Mix Applied Biosystems Cat#4367659Critical Commercial AssaysBio-Plex Pro Cell Signaling Akt Panel 8-Plex Assay Bio Rad Cat#LQ0-0006JK0K0RRLEGEND MAX Mouse OVA Specific IgE ELISA Kit withPre-coated PlatesBiolegend Cat#439807ELISA MAX Mouse Mouse IgE Biolegend Cat#432403Mouse IgA ELISA Quantitation Set Bethyl Laboratories Cat#E90-103Mouse IgG ELISA Quantitation Set Bethyl Laboratories Cat#E90-131DeadEnd Colorimetric TUNEL System Promega Cat#G7130Target Retrieval Solution Dako Cat#S1699Protein Block Serum-Free Dako Cat#X0909ENVISION+ System-HRP labeled Polymer Anti-Rabbit Dako Cat#K4002VECTOR ImmPACT DAB Peroxidase Substrate Vector Laboratories Cat#SK4105Lamina Propria Dissociation Kit Miltenyi Biotec Cat#130-097-410Foxp3 / Transcription Factor Staining Buffer Set eBioscience Cat#00-5523-00Rneasy Mini Kit QIAGEN Cat#74106iScript Advanced cDNA Synthesis Kit for RT-qPCR BIO RAD Cat#1725038Spotchem II glucose Arkray Cat#77301(Continued on next page)Cell 178, 1072–1087.e1–e6, August 22, 2019 e2 LEAD CONTACT AND MATERIALS AVAILABILITYFurther information and requests for reagents may be directed to, and will be fulfilled by the Lead Contact, Koji Hase (hase-kj@pha.keio.ac.jp)EXPERIMENTAL MODEL AND SUBJECT DETAILSMiceUnless otherwise stated, four to five-week-old male BALB/c mice were purchased from CLEA Japan Inc. (Tokyo, Japan) and wereacclimated for one week under specific pathogen-free (SPF) conditions at the animal facilities of the National Center for Global Healthand Medicine, Faculty of Pharmacy, Keio University (Tokyo, Japan). Knock-in mice carrying Kikume-Green Red (KikGR) cDNA underthe CAG promoter were obtained from RIKEN RBC (Tokyo, Japan) and were maintained under SPF conditions at the animal facilitiesof Faculty of Pharmacy, Keio University (Tokyo, Japan). Germ-free (GF) BALB/cA mice (CLEA Japan Inc.) were maintained in GF vinylisolators at an animal facility in the Faculty of Medicine, Keio University. SPF and GF mice were fed with CE-2 (CLEA Japan) and werekept under a 12:12 h light-dark cycle. During the fasting period, mice were kept in plastic cages without bedding chips or bait, with astainless mesh floor to avoid coprophagia, and with ad libitum drinking water. Irrespective of the fasting and refeeding period length,refeeding or tissue collection was set to begin at 8:00 a.m. for all experiments. To examine the effect of rapamycin, 2DG and FTY720in vivo, rapamycin (5 mg/kg; LC laboratories, Woburn, MA) was administrated intraperitoneally daily for seven consecutive days(Zeng et al., 2016), 2DG (250 mg/kg; Abcam, Cambridge, UK) was also administrated i.p. three times every 12 h (Varanasi et al.,2017), while FTY720 (1 mg/kg; Cayman Chemical Company, Ann Arbor, MI) was orally administrated three times every 12 h duringfasting. All animal experiments were performed according to the Institutional Guidelines for the Care and Use of Laboratory Animals inResearch with approval by the local ethics committees at the National Center for Global Health and Medicine, and Keio University.ContinuedREAGENT or RESOURCE SOURCE IDENTIFIERKetone-H kit Serotech Cat#A350Experimental Models: Cell LinesBLS12 cell Katakai et al., 2004 N/AExperimental Models: Organisms/StrainsMouse: B6.B6129-Gt(ROSA)26Sor tm1(CAG-kikGR)Kgwa (KikGR mice)RIKEN RBC Cat# RBRC04847,RRID:IMSR_RBRC04847Mouse: BALB/cA mice CLEA Japan N/AOligonucleotidesCxcl13 (Mm00444533_m1) Thermo Fisher Scientific Cat# 4331182Primer: Rpl32 Forward: GGCTTTTCGGTTCTTAGAGGA Exigen N/APrimer: Rpl32 Reverse: TTCCTGGTCCACAATGTCAA-30) Exigen N/APrimer: Cxcl12 Forward: TTTCAGATGCTTGACGTTGG Exigen N/APrimer: Cxcl12 Reverse: GCGCTCTGCATCAGTGAC Exigen N/APrimer: Cxcl13 Forward: CTCCAGGCCACGGTATTCTG Exigen N/APrimer: Cxcl13 Reverse: GGAGCTTGGGGAGTTGAAGA Exigen N/APrimer; Ccl20 Forward: GGCAGAAGCAAGCAACTACG Exigen N/APrimer; Ccl20 Reverse: CTTTGGATCAGCGCACACAG Exigen N/ASoftware and AlgorithmsPrism 8 GraphPad Software RRID:SCR_002798FlowJo Version 10 FlowJo, LCC RRID:SCR_008520DIVA software Version 6.2 BD Biosciences RRID:SCR_001456IMARIS Version 9.2.0 ZEISS RRID:SCR_007370ImageJ Software Version 1.49 NIH RRID:SCR_003070OtherMouse diet: CE-2 CLEA Japan N/ASporchem Arkray Cat#EZSP-4433Fiber coupled Blue LED Light source Prizmatix Cat#Silver-LED-430Dial-thickness gauge Mitsutoyo Cat#MDC-25PXe3 Cell 178, 1072–1087.e1–e6, August 22, 2019 Cell cultureBLS12 cells were cultured in DMEM (Sigma-Aldrich) supplemented with 10% FBS, GlutaMAX (GIBCO; Thermo Fisher Scientific),1 mM sodium pyruvate (Sigma-Aldrich), and antibiotics, as previously described (Katakai et al., 2004). Cells were stimulated with10 ng/mL murine TNF-a(PeproTech, Rocky Hill, NJ) and/or goat anti-mouse LTbR agonist antibodies (R D Systems, Minneapolis,MN) for 24 h. In a separate experiment, BLS12 cells were cultured in glucose-deprived medium containing 10% FBS, GlutaMAX,MEM essential amino acids (Thermo Fisher Scientific), 1 mM sodium pyruvate, inorganic salts (1.8 mM CaCl2, 0.8 mM MgSO4,5.3 mM KCl, 44 mM NaHCO3, 110 mM NaCl, and 0.9 mM NaH2PO4-H2O), and antibiotics in the presence of TNF-aand anti-LTbR antibodies for 24 h. Control medium was supplemented with 1 mg/ml glucose in the glucose-deprived medium. For pharma-cological inhibition of mTORC1 or glycolysis, BSL12 cells were treated with medium containing various concentrations of rapamycin(LC Laboratories) or 2-deoxy-D-glucose (2DG; Abcam, Cambridge, UK) for 24 h. For analyzing the effect of b-hydroxybutyrate(BHB; Sigma-Aldrich), BLS12 cells were treated with medium containing 25 mM HEPES (GIBCO) with or without 4 mM BHB. Forflow cytometric analysis, BLS12 cells were detached using Trypsin-EDTA solution (Nacalai Tesque, Kyoto, Japan) and washedwith PBS. Mitochondrial reactive oxygen species (ROS) production was detected using MitoSOX-PE (Thermo Fisher Scientific)according to the manufacturer’s instructions.METHOD DETAILSPreparation of lymphocytes and BM cellsPeyer’s patches (PPs) were cut from the intestine and washed twice with phosphate-buffered saline (PBS; pH 7.2). PPs were thenminced and stirred in 30 mL RPMI 1640 medium (pH 7.2; Sigma-Aldrich, St. Louis, MO) containing 2% fetal bovine serum (FBS;Sigma-Aldrich), 12.5 mM HEPES, 100 U/mL penicillin, 100 U/mL streptomycin, 0.5 mg/mL collagenase (Wako Pure Chemical Cor-poration, Osaka, Japan), and 0.5 mg/mL DNase I (Sigma-Aldrich) for 30 min at 37C. After filtration through a 70-mm cell strainer, thecells were resuspended in PBS with 2% FBS. MLNs and CPs were mechanically dispersed into a single-cell suspension. Laminapropria cells of the small intestine were prepared using the Lamina Propria Dissociation Kit (Miltenyi Biotec, Bergisch Gladbach, Ger-many) according to manufacturer’s instructions. BM cells from the right femur and tibia, and splenocytes were mechanicallydispersed into a single-cell suspension. Red blood cells were removed by RBC lysis reagent (150 mM NH4Cl, 1 mM EDTA, and0.4 mM NaHCO3) and then cells were suspended in PBS with 2% FBS.Flow cytometryFor surface and intracellular staining, non-specific binding was blocked with Fc receptor antibody (clone: 2.4G2; Tonbo Biosciences,San Diego, CA) prior to staining with fluorochrome-conjugated antibodies. For intracellular staining, lymphocytes were fixed,permeabilized, and stained with monoclonal antibodies using the Foxp3 staining set (eBioscience, San Diego, CA) according to man-ufacturer’s instructions. 7-AAD (Tonbo Biosciences), Fixable Viability Dye eFluor 780 (Thermo Fisher Scientific, Waltham, MA), LIVE/DEAD Fixable Blue Dead Cell Stain Kit (Thermo Fisher Scientific), or SYTOX Blue Dead Cell Stain (Thermo Fisher Scientific) was usedto discriminate dead cells. The stained samples were analyzed using LSR II, LSRFortessa, and Aria III flow cytometers with DIVAsoftware (all from BD Biosciences, Franklin Lakes, NJ) and FlowJo software version 10 (FlowJo LLC, Ashland, OR).Adoptive transfer of PP cellsPP cells from ad libitum-fed mice were prepared as described above. The cells were labeled with CellTrace Violet (ThermoFisher Scientific) according to the manufacturer’s instructions and then suspended in PBS. The fluorescence-labeled cells(1 3107cells/mouse) were intravenously injected into recipient ad libitum-fed or 18 h-fasted recipients. Both groups were maintainedunder the same conditions after cell transfer. Eighteen hours later, transferred cells were detected in PPs, BM, and MLNs by flowcytometry.Photoconversion of PPs from KikGR micePhotoconversion of the PPs of KikGR mice was performed as described previously (Schmidt et al., 2013; Tomura et al., 2014). Briefly,KikGR mouse was anesthetized with isoflurane, shaved with an electric razor, and antiseptically prepared with 10% povidone-iodine.The skin was incised anteriorly at the midline below the costal margin, and then abdominal wall was incised. Each PP was sequen-tially drawn out from abdominal cavity, and the surgical site was covered by a piece of sterile aluminum foil with a 5 mm hole punchedin it to leave only the PP exposed. A Silver LED 430 with a high numerical aperture polymer optical fiber light guide and fiber collimator(Prizmatix) was used as a 430-nm blue light source. Each PP was exposed for 2 minutes and immediately replaced into the peritonealcavity to avoid drying. The abdominal cavity and skin were closed with 4–0 nylon suture (Natsume Seisakusho). After photoconver-sion surgery, the mice fasted for 36 h or fed ad libitum. Photoconverted cells in the BM, MLNs and the spleen were analyzed by aFACSAria III flow cytometer (BD Biosciences).Oral immunization with OVAIn the fasting mouse group, 36-h fasts began at 8:00 p.m. on day 0. At 8:00 a.m. on day 4, all mouse groups were given 200 mL of 7.5%sodium bicarbonate in HBSS to neutralize gastric acid 1 h prior to oral immunization with 250 mL OVA (4 mg/mL; Sigma Aldrich) andCell 178, 1072–1087.e1–e6, August 22, 2019 e4 CT (40 mg/mL; List Biological Laboratories, Campbell, CA) in PBS. This immunization protocol with or without fasting was repeatedon day 7, 14, and 21. Fecal samples were collected on day 0, 18, 25, and 32 to measure OVA-specific IgA levels by enzyme-linkedimmunosorbent assays (ELISA), while plasma samples were collected on day 32 to measure OVA-specific IgM, IgA, and IgG levelsby ELISA.Detection of antibody responses by ELISATo measure fecal IgA, fecal samples were homogenized in PBS (1 mL/100 mg feces) containing 1 3Complete Mini Protease InhibitorCocktail (Roche, Basel, Switzerland) and 0.02% sodium azide (Wako Pure Chemical Corporation), followed by centrifugation tocollect the supernatant as a fecal extract. Microlon ELISA plates (96-well; Greiner Bio-One, Kremsmu¨nster, Austria) were coatedwith 1 mg/mL OVA in PBS at 4C overnight. After washing, the wells were blocked with 4-fold diluted Block Ace (DS Pharma Biomed-ical, Saita, Japan) for 1 h at room temperature. After washing four times with PBS containing 0.05% Tween 20 (PBS-T), serially dilutedplasma and fecal extracts were added in duplicate (100 mL/well). Fecal extracts from day 0 or plasma from non-immunized mice wereincluded as a negative control. Horseradish peroxidase (HRP)-conjugated anti-mouse IgA, IgM, or IgG (Southern Biotech, Birming-ham, AL) were used as antibodies. After incubation at room temperature for 1 h, the plates were extensively washed with PBS-T.Specific antibody binding was visualized by adding 3,30,5,50-tetramethylbenzidine as a substrate (Sigma-Aldrich) and then thereaction was terminated by 1.2 M H2SO4. Endpoint titers were expressed as the reciprocal log2of the last dilution, giving OD450values higher than control samples (Yamamoto et al., 1998). For measurements of plasma total IgG and fecal total IgA, MouseIgG ELISA Quantitation Set (Bethyl Laboratories) and Mouse IgA ELISA Quantitation Set (Bethyl Laboratories) were used. For mea-surements of plasma total and OVA-specific IgE, ELISA MAX Mouse IgE (Biolegend) and LEGEND Mc Mouse OVA Specific IgE ELISAKit with Pre-coated Plates (Biolegend) were used according to the manufacturer’s instructions.Salmonella infectionrSalmonella–ToxC (DaroA, DaroD) and TT were kindly provided by the BIKEN Foundation (Osaka, Japan)(VanCott et al., 1996).Ad libitum or fasting group mice were orally immunized with 5 3107CFU of rSalmonella-ToxC on day 4. In the fasting group,36-h fasts began at 8:00 p.m. on day 0, 17, and 31. TT-specific IgA in feces was measured by ELISA. Flat-bottomed, 96-wellMaxiSorp Nunc-Immuno plates were coated overnight with 500 ng/well of TT. Plates were blocked with 2% BSA in PBS, and opticallydiluted fecal extracts and sera were added into the plate wells. The Mouse IgA ELISA Quantitation Set (Bethyl Laboratories) was usedfor antibody detection. To produce HRP signals were visualized by adding 3,30,5,50-tetramethylbenzidine as a substrate (Sigma-Aldrich) and then the reaction was terminated by 1.2 M H2SO4.Food antigen-induced diarrhea modelTo establish diarrhea, female mice were injected intraperitoneally with 100 mL OVA (1 mg/mL) and alum (1 mg/mL; Thermo FisherScientific) in PBS on day 0 and 7. Starting on day 14, OVA (50 mg) was administered orally every three days (on day 14, 17, and20). In the fasting group, 36-h fasts began at 8:00 p.m. on day 10 and 17. The severity of allergic reactions to OVA was evaluatedbased on total and OVA-specific IgE in plasma and diarrhea occurrence, which was assessed by visually monitoring the mice forup to 1 h after oral challenge. Fecal and plasma samples were collected on day 14 and 20 to measure total and OVA-specific IgAand IgG, respectively.Induction of oral toleranceMice in the fasting group were fasted for 36 h. After 36-h refeeding (9:00 a.m. on day 3), the mice were gavaged with 25 mg OVA in200 mL PBS. Control mice received PBS only. On day 10, the mice were immunized subcutaneously with 100 mg OVA in 100 mL com-plete Freund’s adjuvant (CFA; Sigma-Aldrich). Delayed-type hypersensitivity (DTH) was measured on day 17 as described previously(Fujihashi et al., 2001). Briefly, 20 mL PBS containing 10 mg OVA was injected into the left ear pinna of the mice, while the right ear pinnareceived PBS as a negative control. After 24 h, ear swelling was measured using a dial thickness gauge (Mitutoyo, Kanagawa, Japan).The DTH response was expressed as the difference in ear thickness between the right and left ears. Plasma samples were collectedon day 24 to measure OVA-specific IgG levels by ELISA.ImmunofluorescenceFor immunostaining, PPs were snap-frozen in liquid nitrogen and embedded in OCT compound (Sakura, Tokyo, Japan). Frozen sec-tions (4-mm thick) were fixed in dry ice-cold acetone for 15 min and then completely dried at room temperature for 1 h. After blockingwith an anti-CD16/CD32 antibody (Tonbo Biosciences) in 10-fold diluted Block Ace (blocking buffer; DS Pharma Biomedical) for30 min, the sections were incubated with primary antibodies (hamster monoclonal anti-mouse CD3ε; BD pharmagen, or rat mono-clonal anti-mouse CD45R/B220; eBiosciences) in blocking buffer overnight at 4C. Bound antibodies were detected with FITC-labeled anti-hamster (Southern Biotech) or TRITC-labeled anti-rat antibodies (Southern Biotech) and counterstained with DAPI.The sections were then examined with a confocal microscope (BX50; Olympus, Tokyo, Japan).Frozen BM sections (10-mm thick) were prepared according to the Kawamoto method (Kawamoto, 2003; Yamazaki et al., 2011)and were fixed in 4% paraformaldehyde. After blocking with Protein Block (Dako, Jena, Germany) for 1 h, the fixed sections werewashed with 0.3% (v/v) Triton X-100 in PBS and incubated with Alexa Fluor 488-conjugated anti-IgM antibodies (Thermo FisherScientific), anti-CD31 and anti-endomucin antibody, for 16 h at 4C. The stained sections were again washed with 0.3% (v/v) Tritone5 Cell 178, 1072–1087.e1–e6, August 22, 2019 X-100 in PBS and further stained with DAPI and secondary antibodies (Alexa Fluor 633-conjugated anti-rat IgG; Thermo Fisher Sci-entific) for 4 h at room temperature. All antibodies were diluted in Protein Block. Immunofluorescence data were obtained andanalyzed with a confocal laser scanning microscope (FV1000; Olympus). B cell number and the distance between naive B cellsand vessels were determined using Imaris (Zeiss, Oberkochen, Germany) and ImageJ Software (NIH).Histological analysisPPs were fixed in 4% paraformaldehyde and embedded in paraffin. Tissue sections (3.5-mm thick) were then stained after deparaffi-nization. For histological examination, the sections were stained with hematoxylin (Dako) and eosin (Wako Pure Chemical Corpora-tion). Antigen retrieval was performed by autoclaving the sections in Target Retrieval Solution (Dako). Then, the sections were treatedwith 3% H2O2(Wako Pure Chemical Corporation) in methanol to inactivate endogenous peroxidase. After blocking with Protein BlockSerum-Free (Dako) for 1 h, the sections were incubated with anti-cleaved Caspase-3 (0.5 mg/mL; Cell Signaling Technology, MA,USA) for 16 h at 4C. After washing with PBS, sections were incubated with ENVISION+ System-HRP labeled Polymer Anti-Rabbit(Dako) for 30 min at room temperature. The ImmPACTTMDAB peroxidase substrate (Vector Laboratories) was used for diaminoben-zidine staining, and hematoxylin was used for counterstaining. For TUNEL staining, the DeadEnd Colorimetric TUNEL System(Promega, Madison, WI) were used according to the manufacturer’s instructions. All sections were examined via confocal micro-scopy (BX50).Transmission electron microscopyPPs were pre-fixed in an aldehyde mixture (2% paraformaldehyde and 2% glutaraldehyde in 30 mM HEPES buffer containing100 mM NaCl and 2 mM CaCl2; pH adjusted to 7.4) for 1 h at room temperature and post-fixed in an aldehyde-OsO4mixture (1%OsO4, 1.25% glutaraldehyde, 1% paraformaldehyde, and 0.32% K3[Fe(CN)6] in 30 mM HEPES buffer; pH 7.4) for 1 h at room tem-perature. The fixed blocks were washed three times with Milli Q water (Milli Q Integral; Merck Millipore, Burlington, MA), dehydratedusing a graded ethanol series, and then embedded in Quetol 812 (Nisshin EM, Tokyo, Japan). The resin blocks were sectioned(70-nm thick) using an ultramicrotome (EM UC7; Leica, Wetzlar, Germany), contrasted with uranyl acetate and lead citrate, and finallyexamined with a transmission electron microscope (JEM-1400; JEOL, Tokyo, Japan).Reverse transcription and quantitative PCRTotal RNA from the BM was extracted using TRIzol Reagent (Thermo Fisher Scientific) while total RNA from PPs and BLS12 cells wasisolated using the RNeasy Mini Kit (QIAGEN, Hilden, Germany) according to manufacturer’s instructions. RNA was reverse-transcribed to obtain cDNA using the iScript Advanced cDNA Synthesis Kit (Bio-Rad Laboratories, Hercules, CA). RT-qPCR wasperformed using the 7900HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA) or the CFX384 Real-Time System(Bio-Rad Laboratories) with SYBR Green (Applied Biosystems) and TaqMan assay (Roche). The oligonucleotide primers for Rpl32,Cxcl12, Cxcl13, and Ccl20 were purchased from Exigen (Tokyo, Japan). TaqMan assay probe for Cxcl13 was obtained from AppliedBiosystems.Bio-plex detection of phosphorylated proteinsTissues were homogenized in 500 mL lysis solution (1X cell lysis factor QG and 2 mM phenylmethylsulfonyl fluoride in cell lysis buffer),vortexed, and placed on ice. The tissue homogenate was transferred to a microcentrifuge tube and frozen at 70C. The sampleswere then thawed, sonicated, and centrifuged at 15,000 3gfor 10 min, after which the supernatant was collected. Protein concen-trations were determined using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific) according to manufacturer’s instructions.Lysates were adjusted to 200 mg/mL to detect the following key phosphorylated proteins of Akt downstream signaling, Akt (Ser473),BAD (Ser136), GSK-3a/b(Ser21/Ser9), IRS-1 (Ser636/Ser639), mTOR (Ser2448), PTEN (Ser280), p70 S6 kinase (Thr389), and S6 ribosomalprotein (Ser235/Ser236), in PPs and the BM, followed by quantitative analysis with the Bio-Plex Phosphoprotein Assay (Bio-Plex ProCell Signaling Akt Panel; 8-plex #lq00006jk0k0rr; Bio-Rad Laboratories) and the Bio-Plex 3D System (Bio-Rad Laboratories)according to manufacturer’s instructions.Plasma parametersPlasma glucose levels were measured using Spotchem EZ SP-4430 (Arkray, Kyoto, Japan) with an automated dry chemistry system(Spotchem II; Arkray). Plasma BHB concentration was measured using the Serotec Ketone-H Kit (Serotec Co., Ltd., Sapporo, Japan)according to the manufacturer’s instructions.QUANTIFICATION AND STATISTICAL ANALYSISFor statistical analyses of two or more groups, we used Student’s t test or ANOVA followed by Tukey’s test. When variances were nothomogeneous, the data were analyzed by the non-parametrical Mann-Whitney U test or the Dunnett’s test. Two-way ANOVA wasapplied to the time-course analysis of TT-specific IgA production. Differences with P-values 0.05 were considered statistically sig-nificant. Statistical analyses were performed using GraphPad Prism 8 software (GraphPad Software, Inc., La Jolla, CA). The exper-iments were not randomized, and the investigators were not blinded to allocation during experiments and outcome assessment.Cell 178, 1072–1087.e1–e6, August 22, 2019 e6 Supplemental FiguresFigure S1. The Behavior of Lymphocytes in Multiple Organs during Fasting and Refeeding, Related to Figure 2(A) Numbers of the indicated cell subsets in PPs, the CPs, the MLNs, and SILP of mice fed ad libitum (PPs, n = 9; CPs, MLNs, and SILP, n = 10) or mice fasted for36 h (PPs, n = 12; CPs, MLNs, and SILP, n = 10).(B) Numbers of PP cells from proximal (Prox), middle (Mid) , or distal (Dis) small intestine (each group, n = 10).(C) Numbers of the indicated cell subsets from PPs at each time point. Mice were fasted for 36 h (blue background) and refed with CE2 for 72 h (red background)(0, 84, and 108 h, n = 9; 24, 36, and 60 h, n = 12).(D) Numbers of the indicated cell subsets in the spleen of mice fed ad libitum, mice fasted for 36 h, and mice refed with CE2 for 48 h (each group, n = 8).Data represent the means ±SEM. Student’s t test (A and B). ANOVA followed by Dunnett’s test (C) or Tukey’s test (D). *p 0.05; **p 0.01; n.s., not significant. Figure S2. Flow Cytometry Gating Strategies and Localization of Naive B Cells in the BM, Related to Figure 2(A) Flow cytometry gating strategies on lymphocytes, granulocytes, macrophages, hematopoietic stem cells, and multipotent progenitors in the BM.(B) Sections of the BM, which were obtained from mice fed ad libitum, fasted for 36 h, or refed with CE2 for 48 h, were stained with Abs against IgM (green) andDAPI (gray) shown on the left and Abs against CD31 (blue), Emcn (blue), and DAPI (gray) shown on the right. The solid arrow indicates IgM+naive B cells inintravascular space. Scale bar: 20 mm.(C) The ratio of IgM+naive B cells in DAPI+. Calculated from Figure S2B (each group, n = 8).Data represent the means ±SEM (C) ANOVA followed by Tukey’s test. *p 0.05, **p 0.01. Figure S3. Minimal Contribution of Gut-Microbiota to the Behavior of Lymphocytes during Fasting, Related to Figure 2Numbers of the indicated cell subsets in PPs and the BM of mice fed ad libitum (n = 9) or fasted for 36 h (n = 8). Six-week-old mice were kept under germ-freeconditions. Data represent the means ±SEM. Student’s t test. *p 0.05; **p 0.01. Figure S4. The Behavior of Lymphocytes during Fasting and Refeeding in Aged Mice, Related to Figure 2(A and B) Numbers of the indicated cell subsets in PPs and the BM of mice fed ad libitum, fasted for 36 h, or refed with CE2 for 48 h. Sixteen-week-old (A) and40–50-week-old (B) mice were kept under SPF conditions (each group, n = 8).Data represent the means ±SEM. ANOVA followed by Tukey’s test. *p 0.05; **p 0.01. Figure S5. Minimal Effects of FTY720 on B Cell Dynamics in PPs and the BM, Related to Figure 4(A) Diagram illustrating the FTY720 treatment protocol. FTY720 or vehicle (PBS) were orally administered into ad libitum-fed or fasting mice.(B and C) Body weight (B) and numbers of the indicated cell subsets in PPs and the BM (C) of mice fed ad libitum or fasted for 36 h. Mice were treated with vehicleor FTY720 (each group, n = 8).Data represent the means ±SEM. ANOVA followed by Tukey’s test (B). Student’s t test (C).*p 0.05; **p 0.01; n.s., not statistically significant. Figure S6. Quantification of Phosphorylated Akt-mTOR Signaling Proteins in PPs and the BM, Related to Figure 4(A and B) Quantification of phosphorylated Akt, BAD, GSK-3a/b, IRS-1, PTEN, S6 ribosomal protein, mTOR, and p70 S6 kinase in PPs (A, n = 18) and the BM(B, n = 6) using whole tissue lysates. The mean of mean fluorescent intensity (MFI) from mice fasted for 36 h, mice refed with CE2 for 24 h and 48 h, and micetreated with rapamycin was normalized by that of mice fed ad libitum. Data are represented as a heatmap. Figure S7. The Effect of Fasting on Salmonella-Induced Mucosal Responses, Related to Figure 6(A) Diagram illustrating the protocol for oral immunization model using S. Typhimurium expressing the fragment C of the tetanus toxoid (Salmonella-ToxC ).Mice were fasted for 36 h on day 0, 17, and 31 in the fasting-refed group.(B) Absorbance of tetanus toxoid (TT)-specific IgA in feces from mice eating ad libitum (n = 8) or fasting-refed (n = 7). Data represent the means ±SEM.Two-way ANOVA.