Transient rapamycin treatment can increase lifespan and healthspan in middle-aged mice

1Successful interventions that increase healthy longevity in people could have profound benefits for quality of life, productivity, and reduced healthcare costs. The drug rapamycin is a promising candidate for such an intervention, as it has been shown to increase lifespan in numerous species and to delay or reverse multiple age-associated phenotypes in mice including cognitive decline, cardiac dysfunction, immune senescence, and cancer. Recently, a six week treatment with the rapamycin derivative RAD001 was reported to improve immune function in elderly people, as measured by response to influenza vaccine, suggesting that at least some of the effects on aging in mice are conserved in humans. Despite these impressive results, the utility of rapamycin or other mTOR inhibitors to delay aging may be limited by side effects. The high doses of rapamycin and its derivatives used clinically to prevent organ transplant rejection are associated with adverse events, including impaired wound healing, edema, elevated circulating triglycerides, impaired glucose homeostasis, gastrointestinal discomfort, and mouth ulcers. While many of these side effects have not been observed in mice at the lower doses that extend lifespan, chronic treatment with encapsulated rapamycin (eRapa) in the diet at 14 ppm has been reported to cause gonadal degeneration in males, increased risk of cataracts, and impaired response to a glucose tolerance test.

2Based on the premise that transient treatment with rapamycin during middle-age might be more suitable for clinical efforts to promote healthy aging than continuous treatment throughout life, we set out to investigate whether a single three-month treatment regimen can extend lifespan and healthspan in C57BL/6JNia mice starting at 20–21 months of age. We initially used a treatment regimen consisting of intraperitoneal (i.p.) injections of 8 mg/kg rapamycin daily for 90 days. This dose was selected because we have previously found that it increases survival and alleviates disease phenotypes in short-lived mouse models of dilated cardiomyopathy, muscular dystrophy, and the severe mitochondrial disease Leigh Syndrome. Based on efficacy in the Leigh Syndrome mouse model and serum drug levels in wild type mice, we estimate that this treatment regimen is comparable to dietary delivery of eRapa at approximately 378 ppm, or 27-fold higher levels than initially shown to extend lifespan in mice when continuous treatment is initiated at either 9 months or 20 months of age. Because prior studies have noted differences in the magnitude of lifespan extension following continuous rapamycin treatment in male versus female animals, we examined the effect of this regimen in both sexes independently. Serum rapamycin levels did not differ significantly between male and female animals in our study (Figure 1—figure supplement 1).

3During the three-month treatment period, we noted a significant decline in body weight of male mice receiving rapamycin injections relative to vehicle treated controls (Figure 1A), although food intake remained similar during the treatment (Figure 1—figure supplement 2). Decreased body weight persisted for several weeks following cessation of treatment (Figure 1A). This was accompanied by a striking increase in median life expectancy from the end of treatment of 60% (p=0.02, Figure 1B, Table 1) and an increase in overall median lifespan from birth of 16% (p=0.03, Figure 1—figure supplement 3, Table 2). This effect is larger than both the absolute and relative magnitude of lifespan extension resulting from continuous treatment to death with 14 ppm eRapa starting at around the same age in UMHET3 mice. The longest-lived rapamycin-treated male in our cohort survived for 710 days post treatment to approximately 1400 days of age. Based on a survey of the literature, this is likely one of the longest-lived wild type C57BL/6 animals ever reported.

4This rapamycin regimen also resulted in a decline in body weight of female mice together with an increased trend of food intake in vehicle-treated animals with time (Figure 2A, Figure 2—figure supplement 2); however, we failed to detect a similar increase in post-treatment survival and lifespan in females (p=0.261, Figure 2B, Figure 2—figure supplement 1, Tables 1–2). Cox proportional hazards regression with robust standard errors and adjusting for cohort indicated significant evidence of a treatment difference between male and female mice (p=0.023). We speculated that this lack of lifespan extension results from an increase in aggressive hematopoietic cancers in the rapamycin treated females. Histopathological analysis showed that while 6/12 control females had round cell tumors (lymphoma and histiocytic sarcoma of hematopoietic origins), comparable to previous studies in C57BL/6 mice, all the rapamycin-treated females examined (16 out of 16) had round cell tumors (Figure 3A,B and Figure 3—figure supplement 1, p=0.002). Additionally, an uncommon variant of lymphoma with plasmacytoid morphology affected 4 out of 16 mice examined in the rapamycin group and no vehicle treated mice (Figure 3A and Figure 3—figure supplement 1A,B). Round cell tumors affected more organs in rapamycin treated females, with multiple (≥2) organs affected in all of the rapamycin treated females compared to only 3 vehicle treated females (Figure 3C, p=0.01). Together, these data indicate a more aggressive phenotype of hematopoietic tumors in the high dose rapamycin treated females. In contrast, the incidence of non-hematopoietic neoplasms was dramatically decreased in the rapamycin group (Figure 3D). Nine non-hematopoietic neoplasms (5 pituitary adenomas, and 2 pulmonary adenomas, and 2 thyroid adenomas) were detected in 7 out of the 12 examined vehicle treated mice; whereas, only 1 out of 16 rapamycin treated females had non-hematopoietic neoplasia (1 pituitary adenoma) (Figure 3D, p=0.004).

5Both control and rapamycin treated male cohorts had a high incidence of systemic round cell neoplasia which affected ≥2 organs, and had an intravascular component and a morphology most consistent with histiocytic sarcoma, although immunohistochemistry was not performed (Figure 3—figure supplement 2). Several male mice also had eosinophilic (hyaline) protein droplets in the renal tubules consistent with lysozyme accumulation associated with histiocytic sarcoma. Of the examined vehicle treated males, all (14 out of 14) had hematopoietic neoplasia affecting at least one organ. Additionally, 7 other types of neoplasia were detected on histopathology from the vehicle treated male group, including 1 gastric neoplasia, 1 thyroid adenoma, and 5 pulmonary adenomas. One male mouse had a subcutaneous facial mass characterized by sheets of round cells, most consistent with an extramedullary plasma cell tumor or mast cell tumor. Rapamycin treated males also had a high incidence of systemic round cell neoplasia affecting at least one organ (9 out of 13). In the rapamycin treated male group, 12 additional non-round cell tumors were detected on histopathology, including 1 hepatocellular carcinoma, 2 intestinal adenomas, 1 splenic hemangiosarcoma, 5 pulmonary adenomas, 1 pulmonary carcinoma, 1 thyroid adenoma, and 1 Zymbal’s gland adenoma (Figure 3—figure supplement 2). In the male group, no cases of systemic lymphoma with plasmacytoid morphology were seen. Although the number of mice examined is relatively small, based on these observations we speculate that this regimen of rapamycin treatment induced the detrimental side effect of aggressive hematopoietic cancers specifically in female mice, which occurred earlier in life and were sufficient to prevent lifespan extension in these animals despite other beneficial effects, such as reduced non-hematopoietic cancers. When data from both sexes are pooled together, daily injection of 8 mg/kg rapamycin for three months resulted in a non-significant (p=0.16) increase in life expectancy of 23% (Table 3). Between randomization and the end of treatment, 5 male mice (3 vehicle, 2 rapamycin) died of non-age related causes and were excluded from the analysis. Their inclusion in the analysis of survival does not significantly affect the results as described above (Figure 1—figure supplement 4).

6In order to assess whether transient treatment with a lower dose and different delivery of rapamycin might reduce side effects in female mice and increase lifespan in both sexes, we utilized dietary eRapa at 126 ppm. Mice were fed a diet containing eRapa or the encapsulation control eudragit diet for 90 days starting at 20–21 months of age and then returned to a standard chow diet. In contrast with the effects of daily injection of 8 mg/kg rapamycin (Figures 1A and 2A), body weight and food intake were largely unaffected by this dietary rapamycin regimen (Figure 4 A,C,E, Figure 4—figure supplement 1), except for a transient, small increase in body weight in female mice fed eRapa, as determined by Student’s t test (Figure 4E). Median post-treatment life expectancy was significantly increased by 42% when female and male animals were considered together (p=0.002, Figure 4B, Table 3) and overall lifespan was increased by 13% (p=0.003, Figure 4—figure supplement 2A, Table 4). Similar results were observed stratifying on sex; three months of 126 ppm eRapa administration significantly increased post-treatment survival in both males and females independently (Figure 4D,F, Figure 4—figure supplement 2B,C and Tables 1–2). Cox proportional hazards regression with robust standard error estimates did not find evidence that sex modifies the treatment effect (p=0.904). Between randomization and the end of treatment, 1 male mouse died of non-age related causes and was excluded from the analysis. Its inclusion in the analysis does not significantly affect the results as described above (Figure 4—figure supplement 3A,B).

7In parallel with survival, we also examined healthspan parameters in these cohorts. The effects of rapamycin on age-associated decline in muscle function and motor coordination were assessed by testing forelimb grip strength and Rotarod performance. Assessments were performed in the same animals before onset of treatment, at the end of the 90 day treatment, and 90 days after the end of the treatment. While eudragit-fed mice showed a steady decline in both grip strength and Rotarod performance, rapamycin-fed mice scored significantly better than control mice in both assays after treatment (Figure 4G,H), suggesting that the healthspan promoting effect of rapamycin continues after treatment is discontinued. Furthermore, only rapamycin-fed mice showed a progressive increase in Rotarod performance during the 3 days of training and over the whole study, with mice performing better at 6 months compared to baseline (Figure 4H, large bracket).

8In the course of routine animal husbandry, we noted that mice treated with either rapamycin regimen produced consistently smaller fecal pellets than age-matched controls (Figure 5A–D). This difference was present in both dry (Figure 5C,D) and freshly excreted feces (Figure 5—figure supplement 1A), indicating that water content is not a major factor affecting feces size. In addition, feces size was affected independently of the method of rapamycin delivery (Figure 5C,D), and the effect was persistent after cessation of treatment (Figure 5—figure supplement 1B,C). We hypothesized that changes in the microbiome may underlie this phenotype and therefore analyzed the fecal microbiome for each of the cohorts used in this study by deep-sequencing of bacterial 16S rRNA. Distance based permutation multivariate analysis of variance (MANOVA) indicated that rapamycin treatment induced a significant change in the composition of fecal microbiome (p=0.018 for injection cohorts. p=0.015 for feeding cohorts.p=0.005 for pooled cohorts), even after accounting for delivery method and batch effects (Figure 6—figure supplements 1–3).

9Among the most notable changes in fecal bacterial DNA content seen in the global microbiome analysis was a significant increase in prevalence of segmented filamentous bacteria (SFB, Candidatus Arthromitus sp.) in the rapamycin treated animals (Figure 6A, Figure 6—figure supplement 2A). SFB are intestinal Gram-positive bacteria with a segmented and filamentous morphology, and are not normally present at high levels in aged mice. The SFB genome lacks a majority of virulence factors and SFB are not invasive; however, their tight adhesion to the intestinal epithelial cell induces differentiation of host immune cells. The increase in SFB following rapamycin treatment was confirmed by real-time PCR of DNA from fecal samples obtained from both mice receiving injections or encapsulated rapamycin (Figure 6B, Figure 6—figure supplement 2B), as well as by semi-quantitative histological scoring of the small intestine in an independent cohort of mice obtained from the Harrison Lab at the Jackson Laboratory and injected with 8 mg/kg/day of for 3 months at the University of Washington (Figure 6C,D, Figure 6—figure supplement 2C). Since increased SFB DNA was observed both in mice injected with rapamycin and mice fed dietary eRapa, this effect is independent of mode of drug delivery. To the best of our knowledge, this represents the first pharmacological intervention to increase SFB in any animal. It will be of interest to determine whether these and other effects of rapamycin on the microbiome are shared across species and play any causal role in the beneficial or detrimental effects of this drug.

10Taken together, our data demonstrate that a single three-month regimen of rapamycin is sufficient to robustly increase life expectancy in middle-aged mice, comparable to the effects previously reported for life-long treatment, while also improving measures of healthspan and substantially altering the microbiome. This work extends prior evidence indicating that short-term rapamycin treatment can improve health in mice, including one experiment suggesting that 4 mg/kg rapamycin every other day for 6 weeks enhanced survival until around 30 months of age in a small cohort, and studies reporting improvements in cardiac and immune function following transient treatment with rapamycin. In the animals treated with the 126 ppm eRapa diet in this study, the improvements in lifespan and health were achieved without overt detrimental side effects, although it is possible that some side effects were undetected, and we did not explicitly test for cataracts, gonadal degeneration, and other adverse outcomes. In the case of the 8 mg/kg/day injection regimen, serious side effects were noted in female, but not male mice. Intriguingly, a dramatic shift toward aggressive hematopoietic cancers and away from non-hematopoietic cancers was observed in these female mice. This is consistent with a similar weak trend seen in kidney transplant patients receiving rapamycin to prevent organ rejection, suggesting a possible conservation of mechanism and clinical relevance. Our data indicate a need to carefully consider sex effects when optimizing treatment regimens and mechanism of drug action. They also illustrate the importance of better understanding the effects of mTOR inhibitors on differential cancer risk, particularly as mTOR inhibitors are being tested and used clinically for a variety of purposes including the treatment of some rare forms of cancer. The importance of evaluating potential risks and adverse side effects when developing interventions to promote healthy aging should not be underestimated.

11This study extends other recent work aimed at developing mid-life interventions to promote healthy aging. Of particular note are two studies reporting improved healthspan from short-term treatments in older mice with a JAK pathway inhibitor and increased lifespan after transient treatment with the NAD+ precursor nicotinamide riboside. From a translational perspective, a healthy aging intervention that can be applied for a relatively short period of time during mid- or late-life is likely to have advantages in cost, practicability and quality of life of people, and we look forward to further developments in this area.

12All lifespan and healthspan experiments were performed on 19–20 month old C57BL6/JNia obtained from the National Institute on Aging Aged Rodent Colony. A separate cohort of C57BL6/J ranging from 17 weeks to 100 weeks of age was obtained from the Harrison lab at the Jackson Laboratory for histological analysis of SFB. Animals were housed in individually ventilated cages (Allentown, Allentown, NJ) containing corncob bedding (Andersons, Maumee, OH) and nestlets. Mice were fed irradiated Picolab Rodent Diet 20 #5053 (Lab Diet, St. Louis, MO). Animals were maintained in a specific pathogen free facility within a Helicobacter spp.-free room. Mice were housed in groups (5 per cage at a maximum) and aggressive male mice were isolated to prevent fighting. Mice were acclimatized at least two weeks before onset of experiments. Mice were inspected daily, and medicated for non-life threatening conditions as directed by the veterinary staff. Date of death was recorded when mice were found dead or unlikely to survive longer than 48 hr at the time of inspection. Mice were euthanized according to the following criteria (modified from the Intervention Testing Program protocol) when they showed one of these symptoms: (1) inability to eat or drink, (2) severe lethargy, as indicated by a lack of response such as a reluctance to move when gently prodded, (3) severe respiratory difficulty while at rest, indicated by a regular pattern of deep abdominal excursions or gasping, or showing any combination of the following features: (a) severe balance and gait disturbance, (b) an ulcerated or bleeding tumor, visible to the naked eye and breaking through the skin of the animal, or rapid weight gain associated with visible or palpable masses, or (c) Body Condition Score equal to 1 or loss of 20% of body weight in the course of seven days. Survival data is shown from the end of treatment (23–24 months of age). Deaths before and during the three month treatment period by group were as follows: 2 vehicle-injected males, 3 rapamycin-injected males, 1 eudragit fed female, 1 eRapa fed male. Sentinel mice (Crl:CD1[ICR]; Charles River, Wilmington, MA) were tested quarterly and were negative for endo/ectoparasites, mouse norovirus, mouse hepatitis virus, mouse parvovirus, and rotavirus. Sentinel mice were tested annually for Mycoplasma pulmonis, pneumonia virus of mice, reovirus 3, Sendai virus, and Theiler murine encephalomyelitis virus. All care of experimental animals was in accordance with the University of Washington institutional guidelines and experiments were performed as approved by the Institutional Animal Care and Use Committee.

13For the injection experiments, rapamycin (LC Laboratories) was dissolved in DMSO to 100 mg/ml, then further diluted in 5% PEG-400/5% Tween-80 to a final concentration of 1.2 mg/ml, sterile filtered, and stored at −80°C for long-term storage. 37 rapamycin treated mice (17 males, 20 females) were i.p. injected daily for 3 months with 66 µl/10g body weight for a final dosage of 8.0 mg/kg starting at 20–21 months old. 38 Control mice (18 males, 20 females) were i.p. injected with vehicle solution (5% PEG-400/5% Tween-20) for 3 months. For the feeding model, encapsulated rapamycin was obtained from Rapamycin Holdings, Inc. Irradiated PicoLab Diet 20 5053 pellets were ground and mixed with encapsulated rapamycin at 126 ppm. 300 ml of 1% agar melted in sterile water and 200 ml of sterile distilled water were added per kilogram of powdered chow, in order to make pellets. Pellets were stored at −20°C until use. Control food contained the same concentration of agar and encapsulation material (eudragit) without rapamycin at the concentration that matched the rapamycin chow. 37 mice (19 on eudragit, 18 on rapamycin) received assigned diet treatments at 20–21 months of age, lasting for 90 days.

14Body weight was measured at least weekly from the beginning of treatment until three months after the end of treatment. For male mice receiving injections (Figure 1A), at each time point from the start point, sample size was as follows: vehicle N = 19, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 17, 17, 17, 17, 17, 17, 17, 17, 17, 16, 16, 16, 16, 14, 13, 13, 13, 13, 13, 13; rapamycin N = 20, 19, 19, 19, 19, 19, 19, 19, 19, 19, 18, 18, 18, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 16, 16, 16, 16, 16, 16, 16, 15, 15. For female mice receiving injections (Figure 2A), at each time point from the start point, sample size was as follows: vehicle N = 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 19, 19, 19, 19, 18, 18, 18, 18, 17, 17, 17, 16, 15, 15, 13, 12, 11 vehicle; rapamycin N = 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 19, 18, 18, 18, 18, 17, 17, 17, 16, 16, 16, 16, 15, 14, 13, 13. For mice receiving micro-encapsulated diets (Figure 4), at each time from the start point, sample size was as follows: eudragit males N= 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 19, 18, 18, 18, 18, 18, 18, 18; eudragit females N= 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 18, 18, 18, 18, 18, 17, 17, 17, 17, 16, 16, 15, 15; eRapa males N= 19, 19, 19, 19, 19, 19, 19, 19, 19, 18, 18, 18, 18, 18, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17; eRapa females N= 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 17, 17, 17, 17, 17, 17, 16, 16, 16, 16, 15, 15, 15.

15Food intake was measured weekly or twice per week per cage by subtracting the amount of food remaining on the wire rack from the amount given 4–7 days before. Average food intake per mouse was calculated by dividing this value by the number of mice in the cage, then averaging all the values from cages hosting mice under the same treatment and of the same sex.

16Gross examination was performed following the natural death or euthanasia of animals. Tissues from mice were fixed with 10% neutral buffered formalin, routinely processed and embedded in paraffin, and stained with haematoxylin and eosin (H&E). 16 rapamycin injected female, 15 rapamycin injected male, 12 vehicle administered female, and 14 vehicle administered male counterparts in the lifespan study were analyzed for tumor incidence and extent by two board certified veterinary pathologists (J.M.S. and P.M.T). 7/15 rapamycin injected male mice were autolyzed to varying degrees, which may have complicated histological detection of systemic neoplasia, although disease sufficient to result in death was determined for 3 of these mice and systemic neoplasia was detected on histological examination of 5 of these 7 mice. Autolysis also complicated the histological assessment of 2 female vehicle treated mice, although systemic neoplasia was detected on histological examination of one of these two mice. Major organs (including decalcified cross section of the head, skin, lung, heart, liver, kidney, spleen, pancreas, lymph node, salivary gland, gastrointestinal tract and reproductive tract) were examined histologically.

17For semi-quantitative SFB analysis, 14 female and 16 male mice i.p. injected with rapamycin (8 mg/kg/day, daily for 3 months) and 9 female and 17 male mice injected with vehicle (daily for 3 months) were sacrificed the day after the end of treatment, and H&E sections of the gastrointestinal tract were routinely prepared and examined. The amount of SFB in the intestinal tissue was graded on a 0 to 4 scale, with 0 representing no SFB, 1 representing rare SFB, 2 representing mild colonization by SFB, 3 representing moderate colonization by SFB, and 4 representing severe colonization of the small intestine by SFB.

18Images of representative lesions were acquired using NIS-Elements BR 3.2 64-bit and plated in Adobe Photoshop Elements. Image brightness and contrast was adjusted using Auto Smart Fix and Auto White Balance manipulations applied to the entire image. Original magnification is stated.

19Fresh blood was collected upon euthanasia via cervical dislocation 24 hr after the last injection, and sera were isolated using serum separators tube (BD, Franklin Lakes, NJ) and immediately stored at −20 C. Rapamycin was extracted with 100 mM ZnSO4 at room temperature and analyzed using a Waters 2795 LC/QuattroMicro MS (Waters, Milford MA).

20Mice were tested on a Rotamex V rotarod (Columbus Instruments, Columbus OH) with a constant acceleration of 0.1 rpm/second over a period of four days. On the first day, all animals were allowed to acclimate to the rotarod with a single round of testing. Over the following three days (labeled day 1, 2, and 3) all mice were tested three times, with a 30 min minimum resting period in between rounds. All mice were subjected to rotarod testing prior to initiation of treatment, upon cessation of treatment, and 90 days after cessation of treatment. Individuals running the rotarod test were not blinded to the treatment.

21Forelimb grip strength was tested with a Chatillon DFE-050 force gauge (AMETEK, Largo FL). Animals were held by the base of the tail, allowed to grip the bar of the gauge, and slowly pulled away from the testing apparatus with a smooth horizontal movement. Each mouse was tested 5 times per round, and maximum grip strength was recorded. Animals were tested prior to initiation of treatment, upon cessation of treatment, and 90 days after cessation of treatment. At least 72 hr rest was allowed between grip strength and rotarod testing. Individuals running the grip strength test were not blinded to the treatment.

22All fecal samples were collected per cage at 3 months of treatment immediately after excretion from the mice analyzed for lifespan analysis and frozen with liquid nitrogen. Microbial DNA was extracted as previously described. Briefly, fecal samples were collected into 800 µL of lysis buffer (500 mM NaCl, 50 mM tris-HCl, 50 mM EDTA, and 4% SDS), homogenized in a Qiagen Tissuelyser II (Valencia CA), and incubated at 70°C for 20 min. The supernatant was mixed with 1 µL of 2 M ammonium acetate, incubated on ice, and then centrifuged at 16,000 × g for 10 min at room temperature. The supernatant was then mixed with an equal volume of ice-cold isopropanol and incubated for 30 min on ice. The contents of the tube were then centrifuged at 4°C for 15 min to pellet DNA. The pellet was rinsed twice with 70% EtOH and re-suspended in 150 µL of tris-EDTA buffer. DNA was further purified using DNeasy kit (Qiagen, Valencia CA) according to the manufacturer’s protocol. For Metagenomic sequencing, sequencing of the V4 region of the 16S rRNA gene was performed on the Illumina MiSeq platform (San Diego CA), as previously described. The raw metagenome data are publicly available at the European Nucleotide Archive (ENA) database (ERP014805).

23Real-time PCR to measure SFB and total bacterial DNA was performed on a CFX384 Real-Time System with a C1000-Touch thermal cycler (BioRad, Hercules, CA) and a StepOnePlus (Applied Biosystems, Foster City, CA) with a Sybr Green method as previously described. 10 ng of DNA collected from fecal samples was analyzed. Primers for SFB are 5’ TGTGGGTTGTGAATAACAAT 3’ and 5’ GCGAGCTTCCCTCATTACAAGG 3’. Primers for the detection of total bacteria (16s rRNA gene) are 5’ TCCTACGGGAGGCAGCAGT 3’ and 5’ GGACTACCAGGGTATCTAATCCTGTT 3’.

24An unpaired two-sample t test was used to compare two experimental groups unless otherwise mentioned. Post-treatment survival data were analyzed using a one-sided Mann Whitney U test (http://vassarstats.net/utest.html). Fisher’s exact test was used for semi-quantitative analysis of SFB and the incidence of malignancies. Rotarod p-values were calculated by applying the glht function for general linear hypotheses for mixed-effects models from the R multcomp package to each of the outputs from the lmer function in the R lme4 package. For the differential abundance analysis of microbiome population, we used the fitZig function in the R package metagenomeSeq.

25The Figure 6—figure supplement 1 heatmap of hierarchical clustering of 16S rRNA gene prevalence was generated using Euclidian Distance using the R package pheatmap. The shift in the microbial population after controlling for the experimental effect was tested using distance matrix based permutation MANOVA implemented in R package vegan as the adonis function using the weighted Unifrac distance. Violin plots and stacked barplot were created using R package ggplot2. In the above analysis we removed a library that exhibited signs of experimental failure (total library size <10000).