Lactobacillus Alleviates Colitis Caused by Chemotherapy Via Biolm Formation

Background: Severe colitis is a common side effect of chemotherapy in cancer patients. A widely consumed probiotic, Lactobacillus, has been reported to alleviate colitis. However, the gastric acid has a powerful bactericidal effect, which restricts to clinical trial success. In this study, we attempted to enhance the viability of probiotics in a gastric acid environment and improve the colitis induced by dextran sulfate sodium (DSS) and chemotherapeutic docetaxel. Methods: We puried Lactobacillus from diverse brands of yogurt and estimated their growth at pH 6.8 and pH 2.0. In the further investigation, the bacterial biolm formation was used to dene the mechanism by which administration of LGG via oral gavage alleviates the colitis and intestine permeability of the mice induced by DSS and docetaxel. The potential benet of probiotics on the treatment of breast cancer metastasis has been assessed as well. Results: Lactobacillus growth was unexpectedly faster in the pH 2.0 than in the neutral pH medium during the rst hour. However, its growth was greatly reduced after the rst hour in the pH 2.0 medium but was maintained at neutral pH. As expected, Escherichia coli (E. coli) could not grow well in an acidic medium. Lactobacillus rhamnosus (LGG) administered in the fasting state via oral gavage signicantly improved the preventive effect in the colitis caused by DSS and docetaxel. Further mechanistic investigations suggested that LGG reduced the permeability of the intestine and decreased the expression of proinammatory cytokines, TNFα, IL-1β, and IL-6, in colitis by biolm formation. We examined the chemotherapeutic effect of docetaxel in a breast cancer model and found that increasing the docetaxel dose may reduce tumor growth and metastasis in the lung but did not benet survival due to severe colitis. However, the LGG supplement signicantly improved the survival of tumor-bearing mice following a high dose of docetaxel treatment. Conclusions: Our ndings provide new in acidic conditions in the presence of glucose was previously reported 11 . One study reported that Lactobacillus rhamnosus (LGG), analyzed in a simulated environment with a pH of 2.0 in the presence of 19.4 mM glucose, showed enhanced survival after a 90 min exposure 12 . However, the dynamics of Lactobacilli in the gastric acid environment and the mechanism by which dietary Lactobacillus protects the intestine from inammation caused by chemicals were still not clear.

Results: Lactobacillus growth was unexpectedly faster in the pH 2.0 than in the neutral pH medium during the rst hour. However, its growth was greatly reduced after the rst hour in the pH 2.0 medium but was maintained at neutral pH. As expected, Escherichia coli (E. coli) could not grow well in an acidic medium. Lactobacillus rhamnosus (LGG) administered in the fasting state via oral gavage signi cantly improved the preventive effect in the colitis caused by DSS and docetaxel. Further mechanistic investigations suggested that LGG reduced the permeability of the intestine and decreased the expression of proin ammatory cytokines, TNFα, IL-1β, and IL-6, in colitis by bio lm formation. We examined the chemotherapeutic effect of docetaxel in a breast cancer model and found that increasing the docetaxel dose may reduce tumor growth and metastasis in the lung but did not bene t survival due to severe colitis. However, the LGG supplement signi cantly improved the survival of tumor-bearing mice following a high dose of docetaxel treatment.
Conclusions: Our ndings provide new insights into the potential mechanism of probiotic protection of the intestine and provide a novel therapeutic strategy to augment the chemotherapeutic treatment of tumors.

Background
Colitis occurs in patients presenting with complex symptoms of acute abdominal pain and direct or rebound tenderness, possibly associated with neutropenia, fever, and/or diarrhea 1 . The dextran sulfate sodium (DSS)-induced colitis mouse model 2 develops characteristics similar to those of ulcerative colitis patients. Two taxanes, paclitaxel and docetaxel [3][4][5] , are widely used to treat advanced breast cancer and other solid tumors in the clinic. However, patients receiving taxane-based chemotherapy may present with colitis symptoms. Taxane-induced colitis can cause serious consequences and restrict the use of chemotherapy. Treatment of taxane-induced colitis with intravenous uids and antibiotics shortens the duration of symptoms, but there is still no effective way to prevent taxane-induced colitis.
The mucous layer overlaying the colonic epithelium consists of mucus glycoprotein (mucins or MUC) and trefoil factors (TFF) secreted by goblet cells, building up the rst defense barrier between the luminal contents and mucosal cells in the gut 6,7 . Because of this barrier, pathogens and antigens cannot access the underlying epithelium and are thereby blocked from invading the body 8 . In this context, in ammation is measured by increased myeloperoxidase activity, presence of proin ammatory cytokines, such as IL-1β, IL-6, TNFα, IFNγ, and histological scores, indicating the severity of mucosal epithelial damage, glandular alterations, and the severity of lamina propria cellular in ltration 9,10 . It is critical to develop an effective treatment to reduce the colitis caused by paclitaxel-based chemotherapy.
There has been a growing interest in Lactobacillus as a probiotic that reduces host gut in ammation. Yogurt, the product of milk fermentation by Lactobacillus, has a bene cial impact on human health. Lactobacillus is a genus of rod-shaped, gram-positive, and anaerobic or microaerophilic bacteria that are "friendly" and normal residents in the oro-gastrointestinal, urinary, and genital tracts. It is mainly present in fermented milk like yogurt, cheeses, fruit juices, wine, and sausages. Lactobacillus is considered a probiotic to prevent and reduce diarrhea, help weight loss, improve lactose digestion in lactose-intolerant individuals, and increase levels of short-chain fatty acids, such as butyrate, which promote gut health. Gastric acid in the stomach activates an enzyme that breaks down proteins as well as kills microbes.
Germs and bacteria, good or bad, are destroyed within 15 minutes in the highly acidic environment of pH 1.5 to 3.5 in the stomach 11 . Therefore, it was important to determine if Lactobacillus could survive in the acidic stomach environment. Increased survivability of probiotic lactobacilli in acidic conditions in the presence of glucose was previously reported 11 . One study reported that Lactobacillus rhamnosus (LGG), analyzed in a simulated environment with a pH of 2.0 in the presence of 19.4 mM glucose, showed enhanced survival after a 90 min exposure 12 . However, the dynamics of Lactobacilli in the gastric acid environment and the mechanism by which dietary Lactobacillus protects the intestine from in ammation caused by chemicals were still not clear.
Here we demonstrated the effect of Lactobacillus on reducing of in ammation in the DSS-and docetaxelinduced colitis mouse models and elucidated the protective mechanisms on the colonic barrier function.

Materials And Methods
Bacteria Various Lactobacillus species were isolated from commercial yogurts, including Suki, Oikos, and Yoplait yogurt purchased from the supermarket (Table 1). Lactobacillus rhamnosus was purchased from ATCC (LGG, ATCC# 53103, Manassas, VA). All Lactobacilli were grown in the De Man, Rogosa, and Sharpe (MRS) broth (Hardy Diagnostics, Santa Maria, CA) at 37°C and incubated in a BactronEZ SHEL LAB anaerobic chamber (Sheldon Manufacturing, Inc. Cornelius, OR). MRS agar plates were made with 55g Lactobacilli MRS broth of dehydrated culture media and 15 g agar (Fisher Scienti c) in one liter of water. The lysogeny broth (LB) medium and agar plates for Escherichia coli (E. coli) growth were purchased from Criterion and prepared according to the manufacturer's instructions.
Mice BALB/c mice (6-8 weeks of age) were purchased from Jackson Laboratory. Mice were housed and fed in a speci c pathogen-free animal house. All animal experiments were approved by the Institutional Animal Care and Use Committee of Nanjing Medical University.

Lactobacillus puri cation from yogurt
Yogurt was diluted with 10 volumes of phosphate-buffered saline (PBS) and centrifuged at 700 × g for 5 min at 4°C. The supernatant was washed once with an equal volume of cold PBS and centrifuged at 7,000 x g for 15 min at 4°C. Pellets were then resuspended in an equal volume of medium (MRS broth and simulated gastric juice) at 37°C and incubated for 6 h with constant stirring.

Agar plate culture
After serial dilution in a maximum-recovery diluent, the bacteria were cultured on MRS agar and LB agar plates in an anaerobic chamber and a regular incubator at 37°C. The growth was monitored after 24~48 h and photographs were taken with the BioDoc-It TM Imaging System (UVP).

Bacterial growth monitoring
Samples were taken at 0, 0.5, 1, 2, 4, and 6 h, to measure the optical density (OD) at 600 nm using a spectrophotometer (BioTek) before and during incubation.

Bio lm formation assay
The overnight culture was diluted 1:100 in the fresh MRS medium containing 0.5% glucose to provide carbon and energy source and grown in a 96-well plate for 24 h at 37°C in the BactronEZ SHEL LAB anaerobic chamber. The quantitative assays were performed in 3 replicate wells for each treatment. After incubation, the culture was removed and washed twice with water to remove unattached cells and media components, signi cantly lowering the background staining. After15 minutes of staining with 150 µl of 0.1% crystal violet (Sigma) diluted to 2.3% with H 2 O, cells were washed twice with H 2 O, and the plates were dried for several hours. The crystal violet stain was solubilized in 150 µl of 30% acetic acid, and optical density was read in a spectrophotometer (BioTek) at 550 nm.

Induction of Colitis by DSS and Docetaxel
Colitis was induced by feeding BALB/c mice 2.5% dextran sulfate sodium salt (DSS, MP Biomedicals, Santa Ana, CA) in drinking water until the end of the experiment. To generate colitis model by a chemotherapeutic agent, injectable solution of docetaxel Sano was purchased from the pharmacy in our hospital and was administered by intraperitoneal injection at 6mg/kg q4d for 28 days [13][14][15] . 5 X 10 10 of LGG were administered to the mice (n=5 mice per group) by oral gavage with/without fasting 2 h daily for 7 days prior to the DSS or docetaxel administration. The PBS control colitis group served as the primary comparison with the LGG intervention group. Bodyweight, stool consistency, and fecal blood were monitored daily. The mice were euthanized on day 28 following 7 days of interventional feeding.
Immediately after euthanasia by carbon dioxide inhalation, the abdominal skin was sprayed with 70% ethanol, and blood was taken by cardiac puncture. Next, the colons were quickly ushed with cold PBS (10 mM, pH 7.4) to remove feces and blood and the distal segments (1.0 cm) were xed in 10% buffered formalin solution for histological examination.

In vivo intestinal permeability assay
Fluorescein-5-isothiocyanate (FITC)-conjugated dextran (MW 4000; Sigma-Aldrich, St. Louis, MO) was administered at a concentration of 60 mg/100 g of body weight by oral gavage to study intestinal permeability in vivo. After 5 h, serum was collected retro-orbitally, and uorescence intensity was determined with a uorescence spectrophotometer (BioTek) at emission and excitation wavelengths of 485 nm and 528 nm, respectively. FITC concentration was measured from standard curves generated by serial dilution of FITC-dextran 4000, as described in the previously study 16 .

Histological analysis
Tissues were xed overnight with buffered 10% formalin solution (SF93-20; Fisher Scienti c, Fair Lawn, NJ) at 4°C and dehydrated by immersing in a graded ethanol series, 70%, 80%, 95%, 100% ethanol for 40 min each. Tissues were embedded in para n and subsequently cut into ultra-thin slices (5 mm) using a microtome, depara nized with xylene (Fisher), and rehydrated by decreasing concentrations of ethanol and PBS. Tissue sections were stained with hematoxylin and eosin (H&E), and slides were scanned with an Aperio ScanScope, as previously described 17 .

Enzyme-linked immunosorbent assay (ELISA)
The cytokine TNFa, IL-1b, and IL-6 levels in mouse colon mucus were quanti ed using ELISA kits (eBioscience) according to the manufacturer's instructions. Brie y, excess binding sites were blocked with 200 µl of 1x ELISA/ELISPoT Diluent (eBioscience) for 1 h at 22°C. The microtiter plates were coated with the anti-mouse TNFa, IL-1b, or IL-6 antibody at 1:200 overnight at 4°C. After washing three times with PBS containing 0.05% Tween 20, the plates were incubated with the detection antibody in blocking buffer for 1 h at 22°C. The plates were washed three times and avidin conjugated with horseradish peroxidase (HRP), and substrate were added. Subsequently, absorbance at 405 nm using a microtiter plate reader (BioTek Synergy HT) was determined.

Labeling of bacteria with PKH26
Bacteria were labeled with PKH26 Fluorescent Cell Linker Kits (Sigma) in accordance with the manufacturer's instructions. After a wash with PBS, bacterial pellets were suspended in 250-500 µl of diluent C with 2-4 µl of PKH26 and subsequently incubated for 30 min at room temperature. After centrifugation for 5 minutes at 13,000x g, labeled LGG nanovectors were resuspended for further experiments.
Total RNA was isolated from murine tissues using an RNeasy mini kit (Qiagen) according to the manufacturer's instructions. In brief, 100 mg of the tissue was homogenized using a tissue grinder and was disrupted in QIAzol Lysis Reagent. Tissue. The homogenate was mixed with 140 ml of chloroform, centrifuged, the upper aqueous phase was mixed with 1.5 volumes of ethanol, and was loaded into the RNeasy spin column. The ow-through was discarded, and the column was washed with RWT and RPE, respectively. The ow-through was discarded after centrifugation, and the column was washed with RWT and RPE. Total RNA was eluted with RNase-free water, and the quality and quantity of the isolated RNA were analyzed with Nanodrop ND-1000 spectrophotometer (NanoDrop Technologies).

mRNA expression by Quantitative Real-Time PCR
For analysis of mRNA expression, 1 µg of total RNA was reverse transcribed by SuperScript III reverse transcriptase (Invitrogen) and quantitated using primers (Sangon, Shanghai) with SsoAdvancedTM Universal SYBR Green Supermix (BioRad); GAPDH was used for normalization. The primer sequences are listed in Table 2. qPCR was performed using the Applied Biosystems 7500 System, with each reaction performed in triplicate. Analysis and fold-changes were determined using the comparative threshold cycle (Ct) method. The change in mRNA expression was calculated as the fold-change. Gene expression was normalized to the control expression by calculating the ∆ Ct = (Ct of control − Ct of the gene). Setting the expression value of GAPDH to 1.0, the relative expression values were calculated as 2 ΔCt . The change in miRNA or mRNA expression was calculated as fold-change.
Tight junction qPCR array For analysis of tight junction genes mRNA expression, 2 µg of total RNA was reverse transcribed by SuperScript III reverse transcriptase (Invitrogen), and quantitation was performed using mouse Tight Junctions RT 2 Pro ler TM PCR Array (Qiagen, PAMM-143) with SsoAdvancedTM Universal SYBR Green Supermix (BioRad); GAPDH was used for normalization. The data analysis was processed online at https://dataanalysis2.qiagen.com/pcr.

Western blot analysis
Samples were incubated in the SDS loading buffer at 95°C for 5 min and separated by 8% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), followed by electroblotting to polyvinylidene di uoride membranes (Bio-Rad). After blocking nonspeci c binding sites for 1 h in 5% nonfat dried milk in PBST (0.05% Tween 20 in PBS), the membrane was incubated for 1 h at room temperature or overnight at 4°C with the primary antibody. After three washes in PBST, the membrane was incubated with HRP-conjugated goat anti-mouse antibody for 45 min at room temperature, washed three times in PBST, and the signal was detected using enhanced chemiluminescence (ECL kit from Amersham Biosciences).

Quanti cation and statistical analysis
Unless otherwise indicated, all statistical analyses were performed with SPSS 16.0 software. Data are presented as mean ± standard deviation (SD). The signi cance of mean values between the two groups was analyzed using the Student's t-test. The differences between individual groups were analyzed by a one-or two-way ANOVA test. The differences were considered signi cant when the p-value was less than 0.05 or 0.01.

Results
Time-dependent effect of acidic pH on growth of Lactobacillus from yogurt Lactobacillus supplement via the oral route must pass through stomach containing gastric acid that can kill most pathogenic germs 18 . The survival of different species of Lactobacillus puri ed from three brands of yogurts was investigated in gastric acid ( Table 1). The bacteria were cultured in the MRS medium (Supplementary Figure 1A) on MRS agar plates (Supplementary Figure 1B) with pH 2.0, mimicking gastric acid pH 19 and compared to those grown at neutral pH 6.8. The cell viability for the acid tolerance test was obtained at 0, 0.5, 1, 4, and 6 h. Unexpectedly, we found that Lactobacillus from all three sources of yogurts not only survived but even grew faster in the rst hour in MRS with pH 2.0 compared to the medium with pH 6.8 ( Figure 1A). The growth of Lactobacillus from all sources in simulated gastric acid quickly increased in the rst hour, reaching the highest population at the end of the rst hour, especially Lactobacillus from the OIKOS yogurt, and then reduced after 1 h. In contrast, all Lactobacillus species were in the lag phase growing slowly in MRS at pH 6.8 in the rst hour, but the growth rates markedly improved after the rst hour of incubation. This effect of pH on Lactobacillus growth was con rmed on MRS agar plates ( Figure 1B). After 4 h, Lactobacillus in the pH 2.0 medium was irreversibly decreased and died, even after transferring to MRS agar plates with neutral pH ( Figure 1A, 1C).
Next, we assessed the growth of LGG, the probiotic with widespread clinical use 20 , in acidic conditions. Consistent with the Lactobacillus from yogurts, the LGG growth was fast in the rst hour without a lag phase and restrained after 1 h in an acidic medium ( Figure 2D, 2E). In the neutral pH medium, the LGG population was increased after 1 h of lag phase, surpassing the group in the acidic medium at 1.5 h. However, E. coli growth was signi cantly inhibited in the pH 2 medium compared to the group in the pH 6.8 medium at all time points (Figure 2 F, 2G). These data suggested that Lactobacillus is tolerant of the gastric acid condition for at least 1 h; the low pH, rather than repressing, dramatically induces Lactobacillus growth for a short time.
To con rm our ndings in vivo, LGG was labeled with the uorescence dye PKH26 21 and administered to BALB/c mice via oral gavage, and the gut bacteria were puri ed from the feces of the intestine. Fluorescence intensity analysis indicated that the LGG/PKH26 signal slightly increased, reaching the peak at 1 h after gavage in mice kept on a regular chow diet. However, the LGG/PKH26 signal in the intestine dramatically increased after 10 min gavage in mice with 6 h fasting (Supplementary Figure 2). Given 1-2 h time of solid diet in the stomach 22 , the LGG administered with solid chow diet could lead to repression of bacterial growth by gastric acid. However, more LGG could reach the intestine if bacteria were administered in the fasting condition.
LGG protects against DSS-induced colitis LGG effects on colitis were investigated in DSS-induced colitis experimental mouse model established by oral administration of 2.5% DSS. Main clinical characteristics of colitis, including diarrhea, fecal bleeding, bodyweight loss, and colon shortening, were observed. The mice were fed with or without LGG (1x10 10 /d/kg, body weight) for 7 days, and body weight and the length and histology of the intestine were estimated 14 days after DSS treatment ( Figure 2A). As shown in Figure 2B, LGG signi cantly recovered bodyweight loss caused by DSS, and LGG administration in fasting as well as along with the chow diet in DSS-induced colitis could enhance protection from the bodyweight loss ( Figure 2B). Also, DSS caused colon shortening, and LGG mitigated its effect; in particular, a signi cant enhancement in the protective role on the intestine was observed when LGG was administered in fasting ( Figure 2C). H&E staining also con rmed more protection of histopathologic changes by LGG in fasting conditions ( Figure  2D). Our ndings supported that LGG could protect mice from DSS-induced colitis, and its administration in fasting enhanced the protective effect probably due to the easier passage of LGG to the intestine in fasting conditions.
LGG protects against chemotherapeutic agent-induced colitis Chemotherapy-induced colitis may manifest in different clinical settings with serious sequelae, impacting patient care and outcomes. We sought to determine whether LGG could play a protective role in chemotherapeutic agent-induced colitis, as seen in DSS-induced colitis ( Figure 3A). The results indicated that 21 d after administration of docetaxel (20 mg/kg weekly) in mice colitis could be reduced by LGG (Figures 3B-3D). Furthermore, mice fed LGG with fasting had a superior protective effect against the taxane-based chemotherapeutic agent docetaxel-induced colitis compared with mice fed LGG without fasting ( Figures 3B-3D). Mice fed with LGG administration without fasting had a moderate bodyweight loss, but with fasting, LGG mitigated the body weight loss even more ( Figure 3B). The length of the colon was shortened by docetaxel. However, LGG administration prevented colonic shortening and LGG with fasting exhibited further prevention ( Figure 3C). The enhanced protection by LGG with fasting was also con rmed by H&E staining ( Figure 3D). These results indicated that LGG was protective against taxaneinduced colitis, and fasting could enhance the LGG effect.
LGG alleviates gut permeability by up-upregulating the expression of tight junction genes Dysfunction of the gut epithelial barrier is a hallmark of in ammatory intestinal diseases. The intestinal epithelial barrier is maintained by tight junctions that connect adjacent epithelial cells and seal the paracellular space 23 . We tested whether LGG may enhance the gut barrier function to protect the intestine from docetaxel-induced colitis. FITC-conjugated dextran was administered to mice by oral gavage. The uorescence signal in the serum indicated that LGG could reduce the permeability of the intestine, and administration of LGG in fasting signi cantly improved gut permeability protection ( Figure 4A). To address the molecular mechanism underlying gut permeability protection by LGG, we performed tight junction gene expression analysis using the mouse Tight Junctions RT 2 Pro ler TM PCR Array (Qiagen, PAMM-143). Eighty-four cell surface receptors involving tight junctions and cell adhesion in the intestine were analyzed. We found that 20 genes were upregulated and 8 genes were downregulated by LGG in docetaxel-treated mice ( Figure 4B). These results were con rmed by performing qPCR of selected genes. The results suggested that LGG upregulated the expression of claudin 5 (Cldn5), tight junction protein 2 (Tjp2) and integrin beta-5 (Itgb5), consistent with the qPCR array results ( Figure 4C). The expression of these genes at the protein level was con rmed by Western blotting ( Figure 4D). Given the potential in uence of LGG on the chemotherapeutic agent-induced alteration, we assessed the proin ammatory cytokines in the intestine using ELISA. The results showed that LGG reduced the expression levels of TNFα, IL-1β. and IL-6 ( Figure 4E).
LGG protects the intestine from colitis by bio lm formation Given the bio lm release from microbial species against extreme environments 24 , we hypothesized that Lactobacillus protected the intestine from colitis through the bio lm ( Figure 5A). We, therefore, investigated bio lm formation by LGG using the crystal violet dye. The optical density measured at 550 nm indicated that LGG generated the bio lm in a dose-dependent manner (Supplementary Figure 3). The speci city of the optical density signal for the bio lm was con rmed by adding 10 µM of aminoimdazole 25 , the bio lm inhibitor 2, to the LGG medium. As expected, the bio lm formation was repressed by 2-aminoimadzole (Supplementary Figure 4). The protective effect of LGG on colitis through bio lm formation was examined by treating mice with docetaxel at 20 mg/kg weekly. LGG reduced colitis via bio lm formation; however, the bio lm inhibitor 2-aminoimdazole abolished the LGG effect ( Figure  5B), indicating that docetaxel-induced colitis could be relieved by LGG-mediated bio lm formation.
We sought to de ne the role of Lactobacillus in the chemotherapy of tumors by determining the potential LGG bene ts in breast cancer treatment with docetaxel. We rst generated the breast cancer model using 4T1 cells and found that increasing the docetaxel dose from 10 mg/kg weekly (Lo) to 100 mg/kg weekly (Hi) signi cantly reduced the primary breast tumor size ( Figure 5C), volume ( Figure 5D), and metastasis nodules in the lung. However, the high dose of docetaxel was detrimental to the survival of tumor-bearing mice due to severe colitis ( Figure 5E). In the parallel group, LGG administration along with docetaxel reduced the tumor size and metastasis and signi cantly improved the survival of tumor-bearing mice treated with docetaxel ( Figure 5E). Collectively, these observations suggested that LGG did not affect tumor size and metastasis directly but alleviated the colitis side effect of chemotherapy and improved the survival of tumor-bearing mice.

Discussion
Probiotics have long been used for maintaining enteric homeostasis and preventing diseases. However, stomach/gastric acid is known to kill germs and bacteria within 15 minutes when pH is less than 3. This study shows that Lactobacillus can tolerate pH 2.0 and grows well in the rst hour at the acidic pH, but the acid tolerance cannot last more than one hour. In contrast, some pathogenic strains of E. coli cannot survive in gastric acid independent of exposure.
LGG of intestinal origin are considered intrinsically resistant to the acid environment and are often employed in fermented foods as probiotics, such as L. acidophilus, L. casei, and L. bulgaricus 26 . Different brands of yogurts contain different strains of Lactobacillus (Table 1). Although there are differences between species and strains, these organisms generally exhibit increased sensitivity at pH values below 3.0 27,28 . In this study, we observed that L. acidophilus, L. casei, and L. bulgaricus from different yogurt brands grew faster in the rst hour, and no in uence on growth was observed for 2 hours at pH 2.0 compared to the Lactobacillus growing in the pH 6.8 medium. Due to the gradient between the extracellular and cytoplasmic pH, cellular functions are inhibited and the cells die when the internal pH reaches a threshold value 29 . In this case, fasting is necessary to reduce the retention time of food in the stomach after consumption.
Our results have been con rmed by another lactobacillus, LGG, widely used in clinical therapies, including colitis 30 . LGG, a signi cant probiotic strain with proven health bene ts, showed better growth in acidic conditions in the rst 1 h of incubation but restrained growth after 2 h. E. coli growth, on the other hand, was signi cantly inhibited in the pH 2.0 medium compared to the pH 6.8 medium independent of the incubation time. Our study showed acid tolerance of LGG consistent with the previous ndings of the highest survival rate of LGG in human gastric and duodenum juice 31 . However, there are also few E.coli strains that survive in acidic conditions because these bacteria use a range of physiological, metabolic, and proton-consuming acid resistance mechanisms to survive acid stress in as low as pH 2.0 32 .
The intestinal tract consists of a diverse microenvironment with more than 500 species of bacteria. A single layer of epithelium separates these commensal microorganisms and pathogens from the underlying immune cells. Thus the epithelial barrier function is a key component in the defense arsenal required to prevent infections and in ammation 33 . A su cient number of probiotics may inhibit pathogenic bacterial adhesion, enhance barrier function, and interact with Toll-like receptors expressed on the intestinal epithelial cells and dendritic cells to produce cytokines/chemokines and further modulate T cells 33 . Probiotics can also produce bioactive metabolites, affect the nervous system, modulate gut motility, reduce pain, and are involved in gut-brain function 33 .
Chemotherapy is a mainstay of primary treatment for advanced breast cancer. However, the side effects are major problems limiting the choice of speci c chemotherapeutic drugs and doses. Paclitaxel and docetaxel, chemotherapy drugs used for solid tumors, including breast cancer, may cause ischemic colitis 34 . In patients with breast carcinoma, colitis events were encountered with paclitaxel-based chemotherapy in both the neoadjuvant setting and treatment for metastatic breast carcinoma 1 . Growing evidence suggests that probiotics may modify the intestinal microenvironment resulting in a decline in proin ammatory cytokines and reduction of colitis 35 . However, the e cacy of probiotic treatment in chemotherapy-induced colitis is still debated. Histologic changes representing ischemic colitis and bowel perforation secondary to transmural necrosis have been reported 34 . Our study has also shown the same histologic changes in colitis caused by docetaxel.
The underlying mechanism of colitis caused by paclitaxel-based chemotherapy is still not known. Both paclitaxel and docetaxel bind to the tubulin subunit, resulting in the formation of stable, nonfunctional microtubule bundles. and interfere with mitosis 36,37 . Therefore, it was postulated that gastrointestinal necrosis or bowel perforation after paclitaxel administration is a direct drug effect because of the transient mitotic arrest due to nonfunctional microtubule bundles 38,39 . The paclitaxel-induced colitis has histological features similar to ischemic colitis, characterized by a thinned attenuated surface epithelium, increased brosis, neutrophil in ltration, and focal hemorrhage [40][41][42] . Colitis of paclitaxel-based chemotherapy can lead to cancer treatment discontinuation, intensive care unit (ICU) admission, and even colonic perforation.
Our study showed that LGG could relieve colitis caused by DSS and chemotherapeutic agent docetaxel, and fasting with LGG could enhance the effect. Reduction of colitis by LGG might allow increased chemotherapeutic dose, leading to enhanced therapeutic e cacy and better survival. Our ndings demonstrated that LGG regulated the expression of tight junction genes, Cldn5, Tjp2, and Itgb5, providing insight into the molecular mechanisms by which LGG regulates intestine permeability. Both LGG groups signi cantly improved mRNA expression of Cldn5, Tjp2, and Itgb5, as evidence of the restoration of the epithelial barrier. The improvement of the barrier function was con rmed by the in vivo FITC-dextran 4000 permeability assay, showing statistical differences with the untreated control group. Previous studies of other probiotics have also reported that the improvement of the epithelial barrier function could contribute to the recovery of colonic damage 43,44 .
As major proin ammatory cytokines, TNF-α and IL-1β, induce apoptosis of epithelial cells, disrupt the epithelial barrier, and prolong in ammation 45 . In our study, mice were euthanized at the end of LGG treatment to examine changes in intestinal response. We observed modest pathological changes, and colon length was signi cantly decreased by DSS or docetaxel in mice without LGG treatment. Additionally, TNF-α was reduced in the colon of mice treated with oral-gavaged LGG, further con rming that colitis was relieved in the LGG-administered mice. The colonic IL-6 expression increased in the colitis model and its antibody showed bene cial effects 46 . In the present study, LGG administration signi cantly reduced the expression of colonic TNF-α, IL-1β, and IL-6 in docetaxel-induced colitis, probably caused by the reduction in in ltration of activated monocytes and macrophages. One previous study also demonstrated that LGG improved colitis with goblet cells by reducing MUC2 expression 47 .
LGG may accelerate ulcerative colitis recovery of the animals 47 , which is consistent with our study. Downregulation of intestinal trefoil factor TFF3 is caused by repression of transcription through TNFalpha and NFκB activation in vitro. This result is consistent with the view that probiotics have to survive the gastrointestinal environment in large quantity and adhere to intestinal cells to exert heath functions 48 .
In general, lactic acid bacteria, especially the species belonging to the genus Lactobacillus, such as L. acidophilus, L. rhamnosus, L. gasseri, L. fermentum, and L. plantarum employ a variety of mechanism for protecting the intestine from in ammation, including phagocytosis regulation 49 , cell surface hydrophobicity adaption, and metabolic inhibitory activities against pathogens 50 that are bene cial to health. However, these endogenous mechanisms may be ineffective when cells are exposed to a wide variety of chemical agents. Therefore, the addition of exogenous protectants is essential for probiotics to endure highly toxic chemical agents. Bio lm can protect microorganisms from harsh environmental conditions such as extreme temperature and pH, high salinity and pressure, poor nutrients, antibiotics, etc., by acting as a barrier 51 . Our analysis of bio lm formation elucidated its role in the LGG anti-colitis effect. Furthermore, the bio lm inhibitor 2-Aminoimidazoles could abolish the probiotic effect on the intestine. 2-Aminoimidazoles are an emerging class of small molecules that inhibit and disperse bio lms in bacteria 25 . Herein we provided a novel bio lm function and an alternative mechanism to explain colitis protection by probiotics. Notably, our ndings demonstrated that LGG alleviated the colitis side effect of chemotherapy and extended the survival of mice with advanced breast cancer following high-dose chemotherapy.

Conclusion
In conclusion, LGG exhibits accelerated growth at pH 2.0 in the rst hour but fails to do so later, suggesting that LGG must be consumed on an empty stomach and not with food to exploit the probiotic activity.
LGG is bene cial in conjunction with the chemotherapeutic agent docetaxel in breast cancer because of bio lm formation and regulation of gut tight junctions. This study provides a new therapeutic strategy for improving the chemotherapeutic e cacy in tumors. Declarations Availability of data and materials

Abbreviations
The datasets used and/or analyzed during the current study can be obtained from the corresponding author upon reasonable request.
Informed consent Not applicable.

Competing interests
All the authors declare that they have no competing interests.

Additional information
Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional a liations.

Supplementary Information
The document of supplementary materials is the supplementary data.  LGG-mediated protection against mouse colitis caused by DSS (A) Schematic representation of the treatment schedule for DSS-induced colitis. BALB/C mice were given 2.5 % DSS in autoclaved water throughout the experiment. Seven days after DSS administration, LGG (5x1010/kg) was administered to BALB/C mice by oral gavage twice a week with/without 2 h fasting. (B) Representative photographs of the colon (left panel) and quanti cation of colon length (right panel). (C) Body weight; (D) H&E-stained colon sections (400x magni cation) from BALB/C mice treated as labeled in the gure. Data are representative of three independent experiments (error bars, SD). * P < 0.05, ** P < 0.01 (two-tailed t-test). LGG-mediated protection against mouse colitis caused by the chemotherapeutic agent docetaxel (A) Schematic representation of the treatment schedule for docetaxel-induced colitis. BALB/C mice (n=5 mice per group) were administered with 6mg/kg docetaxel via intraperitoneal injection once four days. 7 days after docetaxel administration, LGG (5x1010/kg) administered to BALB/C mice by oral gavage twice a week with/without 2 h fasting. 28 days after docetaxel administration, mice were euthanized. (B) Representative colons (left panel); quanti cation of colon length (right panel); (C) Body weight. (D) H&Estained sections of the colon (400x magni cation) from BALB/C mice treated as labeled in the gure.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.