Lactobacillus on Recovering of Immune-Damage by Chemotherapy (Cancer Treatment)

We demonstrated the role of natural probiotics 3L, 3 Lactobacilli, in the establishment of a strong and sustainable beneficial healthy gut flora, after chemotherapy through experimental results through in vivo model. Using rat CTX model (immunosuppression induced by cyclophosphamide), we suggested some new adjuvant to chemotherapy as drugs + lactobacillus treament. Further, we proposed a new probiotic formulation (L. acidophilus + L. casei + L. plantarum) to be explored in the prevention of health condition loss by alteration of the general immune system, in numerous studies that reported the use of probiotics involving Lactobacillus in the post-chemo or post-surgical procedures. Here, in our study, Illumina MiSeq sequencing was used to generate sequencing data from microbial genomic DNA libraries, which is appreciable to check for the effects of 3L on bacteria. Microbiome analysis, phylogenetic and classification reports, community data have supported the experiments and the results where 3L had strong beneficial effects on the microbiome. Further, the influence on specific metabolic pathways are assisted in deriving the conclusion of the study (use of 3L for cancer therapy) to the mode of action, mechanistically by correcting microbiota composition and enhancing specific gut metabolic functions.


Introduction
. The idea of turning to complementary and/or alternative more natural medicine to treat cancer is not new, but it may retain a more sustained consideration in the purchase of our modern life [34]. Given the link between the gut and the immune system, gut microbes (or microflora) have become an important target to boost T cells and shape a better efficiency of cancer therapy [35][36]. Using different cancer model studies, it has been shown that a rich gut flora strongly influences the effectiveness of anticancer drugs such as cyclophosphamides (CTX) and immune checkpoint inhibitors (ICIs) [37][38][39][40]. Our present study compares with any of the marketed probiotic formulations such as L. rhamnosus that are known to inhibit the growth of cancer cells in a dose-and time-dependent manner [41]. It also compares to L. rhamnosus GG, as the most studied L model in cancer, to potentiate the gut microbiota and protect against the genesis of tumors [35,42].
In this new and challenging research field, mainly focusing on colon cancer, an universal 'probiotic' approach is expected to prevent patient selection in view of the different treatments and individualized host responses to gut modulation [43]. Accordingly, for medicinal purposes, we have developed a specific 3L Lactobacillus bioproduct (a natural cocktail composed of L. acidophilus, L. casei and L. plantarum), which has been demonstrated to have a very beneficial effect not only on physiological but also on biochemical status of hyperlipidemic mice. In mice, "3L" has been shown not only to induce a beneficial gut flora, but also to significantly reduce cholesterolemia, LDL/HDL ratio, blood lipid concentration and weight gain [44][45].
Rodents are the best models to study human metabolic syndromes and obesity-associated pathologies [46].
They are also very reliable models to investigate human cancer immunology and immunotherapy [47]. Therefore, we used five groups of rats as experimental models to check for the effects of 3L on gut flora and cancer chemotherapy. We analyzed the gut microbiome after chemotherapy (use of cytotoxic drugs such as CTX) for treatment of cancer by using Illumina MiSeq sequencing optimized for full complete microbial genome applications. We showed a microbiome composition significantly altered by chemotherapy. Then, we tested the effects of chemotherapy combined with 3L on gut flora and health status of rats treated for cancer. The overall result of cancer treatment with CTX + Lactobacillus was beneficial to an extent not reported before. Increasing the dose of a Lactobacillus cocktail (acidophilus -casei -plantarum) had a marked positive effect on gut flora. Applying a microecosystem such as Lactobacillus did not cure cancer, but contributed to a better gut health in the chemotherapy for tumor treatment. A pharmacological bioproduct high in Lactobacillus strains is shown to significantly contribute to a healthy gut microbiome despite CTX chemotherapy. We therefore propose a new natural medicine to be applied in cancer treatment, in particular in prevention of an altered gut due to cyclophosphamide therapy.

Industrial preparation of Lactobacillus
In a pioneering study, we developped a three-strain lactobacillus probiotic formula (3L) against cholesterolemia and hyperlipidemia [43]. The same probiotic formula used in this study against immune damages in chemotherapy for cancer (L. acidophilus SD65, L. casei SD07 and L. plantarum SD02) was produced by our Industrial Laboratory platform for natural medicine in Jinan (Shandong Province, P.R. China). The growth of pure cultures of the three bacterial strains was taken in de Man, Rogosa and Sharpe (MRS) agar liquid medium and placed in an anaerobic workstation held at 37°C (industrial platform), following Yue et al. (2014) [43]. For 3L preparation, the bacterial cells of each strain were harvested by centrifugation at 2000 × g for 20 min (4°C). The cell pellet in each strain was resuspended in the ratio of 10 9 CFU/ml in sterile saline water solution and kept at 4°C. The tripartite L probiotic solution (3L) was freshly prepared by mixing the cold suspensions of SD65, SD07 and SD02 in an equal volume and stored in cold conditions (4~10°C) and used on later point of time. Accordingly, rats received every day a dose (0.3 ml) of bioproducts administered intra-gastrically using a stainless-steel needle in addition of chemotherapy (cyclophosphamide, CTX; Figure 1). ml/kg bodyweight (M), and 5) Immune attacked and treated with Lactobacillus, i.e. 5.0 ml/kg bodyweight (H). Total number of studied animals was 50, in each group were 10 rats ( Figure 1). Control rats (CK group) were with continuous gastric perfusion of normal saline (NS). Rats in groups 2-5 were immune attacked by an injection of cyclophosphamide (CTX, 10 mg/ml, intraperitoneally). "Nude" rats were only treated with CTX at day 1,5,8,15 and 22 (IM group; Figure 1). "Covered" rats were with continuous gastric perfusion of NS containing a specific dosage of 3L (1.25 ml/kg, low dose, L group; 2.50 ml/kg, middle dose, M group; 5.0 ml/kg, high dose, H group; Figure 1). The deleterious effects of CTX cyclophosphamide chemotherapy on the rat immune system and the beneficial effects of Lactobacillus preparation on immuno suppression induced by cyclosphosphamide were highlighted in a companion study [48]. We found that from one week to one month (28 days) after administration, the number of white blood cells of the animals treated with Lactobacillus preparation (H-dose: 5 mg/kg) was higher than that of the animals in the model group treated with CTX. The CD4+/CD8+ ratio (ratio of T helper cells to cytotoxic T cells) was also higher in the animals treated by Lactobacillus preparation (L-dose: 1.25, M-dose: 2.5 and H-dose: 5 mg/kg). Lactobacillus-treated animals showed increased serum level of interleukin 6 (IL-6) and high interleukin (IL-6 and IL-2) gene expression, but significant decrease in the rat mRNA expression levels of Tumour Necrosis Factor alpha (TNF-alpha). The effects on leukocytes, T helper cells, interleukins and TNF-alpha were dose-dependent [48], which urged us to test three rather different doses (  The blue dots show the groups that have maintained gut flora close to control healthy conditions (Lactobacillus-treated groups). The dark blue dot shows the group (high 3L) that has gut flora (microbiome) highly similar to control healthy conditions (CK). Time 0 is when chemotherapy starts in +CTX groups.
Feces were collected in each group after twenty-eight days treatment. N= total number of rats in experiment, n= number of rats per group.

Preparation of five fecal microbial genomic DNA samples in rat CTX model
For the preparation of DNA samples from the five groups of rats (CK, IM, L, M and H) in rat CTX model, DNA was extracted using the method previously selected for mice and piglet fecal microbiome analysis [43][44]49]. This method was reliable for testing, quantity, purity and quality control of fecal DNA samples and Illumina sequencing [49]. As described in Yue et al. (2020) [49], 2.0 g of fecal samples from Group CK-H rats were processed for microbial genomic DNA extraction using QIAamp Fast DNA Stool MiniKit (Qiagen GmbH, Hilden, Germany) and used as template (10 ng [50]. The purpose in MiSeq was to add adapter sequences onto the ends of microbial DNA fragments to generate indexed libraries for both single-and paired-end reads [51]. Firstly, fecal microbial genomic DNA amplicons from rats were subject to terminal end repair. The 5'-end of DNA was excised by End Repair module (Mix2), and added with a phosphate group. Meanwhile, the missing base of 3'-end was filled. Adenosine base was added at the 3'-end of each microbial DNA sequence for prevention of self ligation. This also insured that each DNA target sequence was properly linked to the sequencing linker. A sequencer corresponding to a library-specific tag (Index Sequence) was added at the 5'-end of the PCR amplicons to immobilize DNA in flow cell. Self-ligated fragments were removed using BECKMAN AMPure XP Beads (Beckman Coulter™, Illkirch, France) to purify the microbial library system. To enrich the libraries as much as possible for the DNA of interest, PCR amplicons were used as a template in a second-PCR run. The PCR conditions were as described under preparation of five fecal microbial genomic DNA samples in rate cancer model. In MiSeq, PCR amplicons were purified using the magnetic beads screening method (Beckman) and analyzed by 2% agarose gel electrophoresis before high-throughput sequencing.
Prior to high-throughput sequencing, the quality of each rat fecal microbial genomic DNA library was checked by using Agilent High Sensitivity DNA kit on Agilent Bioanalyzer (Agilent Technologies Inc. , Waldbronn, Germany). Each DNA library produced only a single peak and no joints on Agilent check.
Sequence librairies were then quantified on the Promega Quantifluor fluorescence quantitative system using Quant-iT PicoGreen dsDNA assay kit (Promega Corporation, Madison, USA). Library concentration was above 2 nM. After serial dilutions, DNA samples were mixed, denatured by NaOH and sequenced. Two-end sequencing with 2 x 300 bp reading length was performed using Illumina MiSeq Reagent kit v3 on 600 cycles MiSeq Sequencer (Illumina Next Generation). The optimal sequencing length of the target DNA fragment was comprised between 200 and 450 bp.
Sequences with more than one base mismatched with the 5'-end primer and/or more than eight consecutive identical bases were discarded [52][53][54]. Other sequences were classified into Operational Taxonomic Units (OTUs), which were further used for taxonomic identification of bacterial DNA and phylogenetic analysis [55]. The diversity level of each sample was evaluated on the basis of OTU values, while the depth of sequencing (enhanced microbial community analysis) was shown by analysis of rarefaction curves [56]. The composition of the five samples corresponding to the five groups of rats (CK-H) was analyzed at different taxonomic levels: phylum, order, class, family, genus and species (i.e. complete microbiome; Figure 1). The microbiome results (OTU, 100%) were displayed in tables and histograms (R software). Rare OTUs (with an abundance proportion less than 0.001%) were not considered for microbiome analysis [57]. Shared and unique OTUs within each group (calculated by R) were displayed in Venn diagrams in R (Treat*/2.5.1_Venn).
Using OTU abundance matrix to study complete microbial structure in each group, rarefaction curves were drawn to reflect the microbial diversity among samples, i.e. compare the number of OTUs in the five different groups of rats (CK-H) at the same sequencing depth and justify the diversity level of each sample (QIIME2, alpha rarefaction curve). The length of the curve reflects the number of sample sequencing depth; the longer the curve, the higher the sequencing depth, which strongly increases the possibility to observe increased microbial diversity. The slope of the curve reflects the effect of sequencing depth on microbial diversity in the sample. A flat rarefaction curve (low slope) indicates that the sequencing results are sufficient to reflect microbial diversity, and increasing sequencing depth cannot detect more novel OTUs. A bump rarefaction curve (high slope) indicates that the diversity is not saturated, increasing sequencing depth could help detect more OTUs (Treat*/2.3.2_arare). We also measured "Specaccum" (species accumulation curve) in five groups of rat associated with cancer treatment and 3L. Similar to rarefaction curve, specaccum gives the extend of increase in microbial community richness along with increase in sample size [58]. Using specaccum function in R, we estimated whether the sample size was sufficient to reflect the different underlying bacterial communities of the different groups or samples. The specaccum species accumulation curve was plotted for the total number of OTUs in each sample from the OTU abundance matrix using R in vegan (S3 method; Treat*/2.3.3_specaccum) [59]. In addition, rank abundance curve (RAC) was used to see the number of highly abundant vs rare OTUs in each community [60]. For RAC, OTU values were in descending orders and transformed in Log2 data in R package (Treat*/2.3.4_rabund). Other multiple indices were scored to reflect microbial alpha diversity in rat groups. We included Chao1 index and ACE index to reflect community richness, and Shannon-Simpson indices to reflect both evenness and richness of the bacterial community in each of the five groups of rats related to cancer chemotherapy and lactobacillus treatment (CK-H) using QIIME software (QIIME 2) in R [54][55][56][57][58][59][60].
The differences in gut flora structure and related microbial species between the groups were analyzed by multiple statistical analysis tools (Metastats) in Mothur software (http://metastats.cbcb.umd.edu), providing the sequence difference (or absolute abundance) of two samples/groups on the basis of P and Q values [61].
LEfSe was used to measure community composition differences in the different groups based on linear discriminant analysis (LDA) effect size. LEfSe analysis combines LDA with Krustal-Wallis and Wilcoxon rank sum test to find key biomarkers (i.e, key community members) [62]. Relative abundance matrix was used for LEfSe analysis through the Galaxy Online Analysis Platform for sample group comparison and visual analysis results (http://huttenhower.sph.harvard.edu/galaxy; Treat*/2.5.5_LEfSe).

Phylogenetic and classification analysis
OTU representative sequences were used as taxa to build phylogenetic trees (Newick) in FastTree tool [63].
Using MEGAN [64], abundance and taxonomic composition information of OTUs in each sample was projected to the microbiological classification tree from NCBI (https://www.ncbi.nlm.nih.gov/taxonomy).
Hierarchical trees (GraPhlAn) were constructed with the whole sample population at each taxonomic level.
Using GraPhlan [65], taxonomic units were distinguished by different colors and their abundance distribution was reflected by the node size, i.e. the average relative abundance of the taxonomic unit. Interactive presentations of community taxonomy in groups of rats related to cancer were conducted by KronaTools 2.4 software [66]. The main purpose of ß-diversity analysis was to examine the similarity of community structure between the different groups. Three different types of methods were used to observe the differences between the groups through the natural decomposition of the community data structure and the sample ordination: principal component analysis (PCA), multidimensional scaling (MDS) and clustering analysis (CA). PCA evaluates the similarity between samples based on Euclidean distance, regardless of the possible inter-relation between the original variables (R software). Nonmetric MDS only considers the size of the relationship between samples before to classify the group or sample distances (UniFrac distance matrices of Unweighted and Weighted in R software). Like Nonmetric MDS analysis, CA such as Unweighted Pair-Group Method with Arithmetic Means (UPGMA), single-linkage clustering and complete-linkage clustering uses any distance to evaluate the similary between samples (QIIME-R). Using QIIME, the Weighted and Unweighted UniFrac distance between or within the groups were tested by T-test. The statistical significance was checked by 1000 Monte Carlo permutations. This fully describes the size of the flora structure differences between or within groups of rat fecal samples related to cancer.
Finally, the massive community data produced by high-throughput DNA sequencing in five groups of rats (CK-H) required to use more statistical analyses such as Constrained Ordination and Supervised Learning.
In accordance with a known sample correlation (a sample distribution or grouping information) or a sample test indicator (continuous variable), the commonly used Constrained Ordering and Supervized Learning methods, including redundancy analysis (RDA), canonical analysis (CNA) and partial least squares discriminant analysis (PLS-DA, Variable Importance in Projection or VIP value in R) extracted the pattern characteristics associated with original microbial community data. The larger the VIP value is, the greater is contribution of the species to the differences between groups. In addition, Adonis/PERMANOVA (permutational multivariate analysis of variance) analysis was performed using QIIME software, and 999 substitutions were used to determine whether the differences between groups were statistically different.
Focusing on P value, the smaller the P value, the stronger the difference between groups of rats for cancer.

Comparison of microbiomes in rat groups in relation with chemotherapy and Lactobacillus
In the continuity of our work on Lactobacillus/bacillus on cholesterolemia, lipidemia, diarrhea and scour [see [43][44]49], this new study explores the fecal microbiota of rats subject to chemotherapy (cyclophosphamides) for cancer and significant health improvement when co-treated with a cocktail of three Lactobacillus spp.
The topic is of very high significance in pharmacological treatments of cancer, as cyclophosphamides or CTX used against cancerous tumors are commonly found to severely damage the patient's immune system.
The bioproduct for cancer study design includes five different experimental groups (1) healthy control rats and treated rats (CTX chemotherapy), which were either only treated with CTX (2) or treated with CTX and a low (3), middle (4) or high (5)  Middle complement dose (×2, two-times fold) fell in between control (CK) and immune-damaged (IM) groups ( Figure 2). Therefore, increasing Lactobacillus spp. concentration (no more than ×5, five-times fold) was necessary to provoke a strong beneficial effect on the microbiome of rats treated by chemotherapy for cancer ( Figure 2). All samples (three samples per group) and OTU numbers were coherent between the groups CK-H.
About the same quantity of 16S and ITS sequences was obtained in different biological replicates in each group (Table S1). About the same sequence quantity was obtained in five different groups for chemo (Table   S1). No differences were found in the OTU number of samples at each taxonomic level (phylum, class, order, family, genus and species; Tables S2 & S3). All of the OTUs could be classified, but the classification of OTUs at different taxonomic levels in five rat groups related to chemotherapy and Lactobacillus treatment did not show apparent differences in the OTU counting (Tables S2 & S3 (Tables S2 & S3). Therefore, the sample sizes were the same, especially for Family and Genus ( Figure S1). However, using the OTU table for sample diversity in PCA, rank abundance curve, NMDS, principle coordinate analysis (PcoA), Bray-Curtis distance plot (default semimetric), binary Jaccard distance matrix (metric) and UPGMA clearly classified H with CK.
Not only Venn Diagram, but also PCA based on the OTU composition displayed key differences in the five groups ( Figure S2). PCA showed similarity of the chemotherapy + high lactobacillus and control groups at overall community level, with Fusarium, Talaromyces, Sarocladium, Aspergillus and Mucor falling outside a common spectrum of microbe genera ( Figure S2A). Similarly, OTU rank abundance curve (i.e. species richness and species evenness) showed more microbial diversity occurring in control and Lactobacillus (L-H) groups compared to CTX alone immune damaged ill group ( Figure S3). In NMDS, CK-samples tended to group with H-, M-and L-triplicates. IM group fell distantly from the other groups, clearly showing mean dissimilarities between microbiomes from CTX-treated samples and those from controls (no treatment) and rats treated with Lactobacillus in addition of CTX (Bray-Curtis; Figure S4A). The NMDS graph using Jaccard index collapsed very clear information:

Principle Component Analysis and Orthogonal Projections to Latent Structures Discriminant Analysis
CK grouped with H-samples, showing mean similarities between H-microbiomes and controls (Jaccard; Figure S4B). The same grouping was observed using PcoA as a principle analysis ( Figure S5). Both Bray-Curtis (abundance) and Jaccard (0/1 data) indices showed the pair of communities with similar species richness (H and CK; Figure S5AB). Unweighted and non-metric MDS analysis (UPGMA) confirmed the similarity between CK and H samples. On the UPGMA tree, H branches clustered with CK with low distance value (0.005-0.0028; Figure S6A). Unweighted pairs were also found between M and L groups (0.031-0.111, Figure S6A; 0.003, Figure S6B). The branches corresponding to the immune-damaged group (IM) clustered at the bottom of the tree, showing mean distance (or difference) of IM compared to CK, H, M and L ( Figure   S6).

Profiling microbiome in five different groups of rats in relation with cancer chemotherapy treatment
and addition of Lactobacillus pharmacology. Specaccum (number of species vs number of samples) gave the extend of increase in microbial community richness along with increase in sample size. Specaccum estimated that the sample size was sufficient to reflect the different underlying microbial communities of the different rat groups ( Figure S7). Chao1, Simpson, Shannon, Pielou_e, observed species and Goods_coverage indices (alpha-diversity) were scored in grouped and ungrouped samples ( Figure S8 & Table S4). These indices (Chao1, Pielou_e and observed_species) reflected similar community richness and species evenness in CK and H groups ( Figure S8 & Table S4). Goods_coverage index showed marked differences between H and IM. High microbial diversity was found in H, as reflected by Goods_coverage metrics at OTU levels (sample completeness, p = 0.76; Figure S8 & Table S4). Shannon-Simpson indices reflected high evenness and richness of the microbial community in each of the three Lactobacillus groups of rats related to cancer chemotherapy and probiotic treatment ( Figure S8 & Table S4).
In the next step of the study of Lactobacillus for cancer therapy, microbial taxa clustering based on the abundance of each microbe in five rat samples related to cancer and treatment was analyzed by heatmap on group and triplicate samples ( Figure 3AB). Microbe heatmap analysis showed the relative abundance of each taxon in CK, IM, H, M and L. The diverse microbial taxa identified in the rat groups for cancer and 3L are listed in Table 1. There were several broad types of microbes identified in CK group from Acaulium (syn. Scopulariopsis) genus to Xeromyces ( Figure 3 & Table 1 Figure 3 & Table 1).
Therefore, among regulated fungi are species which are not known as commensals or pathogens of animals.
Lecanicillium are known as generalist entomopathogenic fungi [67]. Ustilago is known pathogen of Poaceae plants [68]. Phallus are big saprotrophic mushrooms [69]. The presence of these fungi in the rat microbiota is not necessarily doubtful and deserves attention. The breeding history of rats takes place in Class II animal facility of the Institute of Medicine in SAMS (Specific Free Pathogen/SPF facilities and acute hospital care settings that are designed to maintain organisms in sterile environments). Saprophytic basidiomycetes are known wood-decaying fungi, but they are also reported from penis and urethra of animals, where they play a role in male fertility [70]. In fact, little is known about the fungal floral of the digestive and reproductory tracts in rodents. Species in Lecanicillium are pathogens that parasitize not only insects, but also worms and many other fungi, which may explain their presence in gut fungi associated with rats. Various strains of Lecanicillium are found in gut fungi associated with marmots [71]. As seen in rodents, a large variety of 'forgotten' odd fungi, including Ustilaginales and Ustilago sp. , are found in the human digestive tract [72-   The heatmap analysis also highlighted two genera in particular, Fusarium and Pichia, highly abundant in feces from control and high lactobacillus-treated groups or repeatedly found in control samples but significantly altered by chemotherapy ( Figure 3AB & Table 1). Furthermore, correlation was observed between Rhizopus and Lactobacillus treatment. Rhizopus-levels were significantly altered by chemotherapy, but increased when Lactobacillus was added to phosphamide. High abundance of Rhizopus microbes was detected in H, M and L samples. Higher abundance of Rhizopus was found in M samples (CTX + middle dose/2.5 ml/kg bodyweight of Lactobacillus, Figure 3 & Table 1), suggesting a dose-effect relationship between the dose of bioproduct and the increase of specific microbes. We noted that high-, medium-and low-dose injections of 3L cocktail were beneficial in reducing the levels of Acaulium, Mucor, Olpidium, Penicillium, Periconia, Xerochrysium and Xeromyces ( Figure 3 & Table 1). In heatmap, only one microbial genus (Phallus) was induced by chemotherapy and not cured by high dose of 3L ( Figure 3A). However, medium and low doses of 3L were both able to eradicate Phallus fungi despite CTX treatment ( Figure 3A & Table 1). Phallus infection was not predominant in all IM samples ( Figure 3B). This also showed the importance of dosage and/or a process of gradually decreased dose adjustment over time for a useful therapeutic application of 3L treatment.  Table 2). The dominant microbial phyla were similar between control conditions and specific treatments ( Figure 4A & Table 2). However, CTX induced a significant decrease of Ascomycotalevels, which was not observed using high-doses of 3L during chemotherapy ( Figure 4A). This beneficial effect of 3L injections was observed for many other microbial fungal phyla, including Basidiomycota, Kickxellomycota and Mortierellomycota ( Figure 4A & Table 2). Levels of Mucoromycota and Olpidiomycota in rat fecal microbiomes were increased during CTX chemotherapy, but kept low when adding H, M or L doses of Lactobacillus to anti-tumor treatment ( Figure 4A & Table 2). Similarly, specific beneficial effects of Lactobacillus treatment in complement of chemotherapy for cancer were seen by analysis of the distribution of microbial fungal classes, orders, families and genera in the five rat groups ( Figure 4B-E & Table 2). The most abundant microbial fungal classes of healthy control rats without any other treatment than normal saline injection were Dothideomycetes, Eurotiomycetes, Saccharomycetes, Sordariomycetes and Tremellomycetes. All the five classes were seriously affected by chemotherapy.

Effects of CTX and CTX +
Chemotherapy also induced higher levels of Mucoromycetes in the fecal microbiome. Dothideomycetes, Eurotiomycetes and Sordariomycetes were conserved at normal levels with high dose injections of 3L.
Importantly, Mucoromycetes were maintained at normal levels in all two Lactobacilli-treated samples. Other microbial classes such as Agaricostilbocytes, Leotiomycetes and Tremellomycetes, were also retrieved at normal levels upon probiotic treatments ( Figure 4B & Table 2). The most abundant microbes on order level were Eurotiales, Hypocreales, Saccharomycetales and Trichosporonales not only in CK, but also in H group.  Table 2). Aspergillaceae, Didymellaceae, Nectriaceae and Pleosporaceae were four major microbial families identified in the rat fecal samples related to cancer and specific therapy.
Interestingly, these microbial families were vulnerable to chemotherapy alone, but maintained by adding 3L to CTX treatment. A gradual increase of 3L was necessary to maintain the levels of Aspergillaceae.
Cordycipitaceae, Mycosphaerellaceae, Rhizopodaceae, Thermoascaceae and Trichosporonaceae-levels were also reduced by CTX but maintained when CTX was combined with Lactobacillus injections. It was not the case for all microbial families identified in rat fecal samples. Pichiaceae were an example of microbial families down-regulated after CTX treatment and immune-damage, which could not be reversed by adding  Comparing all the triplicates ( Figure S9), Ascomycota-levels were found to be remarkably high in CK and H triplicates ( Figure S9A). In contrast, particularly low levels of Ascomycota were found in IM021, L103 and M201. Mucoromycota and Olpidiomycota were extremely high in IM015 samples ( Figure S9A). Fusarium as a key biomarker (i.e, a key community member) of CK group ( Figure S11B). Comparative metagenomics and network analysis at phylum level showed a high degree of similarity between control and Lactobacillus-treated rat fecal samples and revealed that Ascomycota were dominant in this network ( Figure   S12). IM samples (in blue) are not associated to this CK-Lactobacillus group ( Figure S12A). Mucoromycota (in orange) were dominant in CTX-immune-attacked ill rat fecal samples ( Figure S12B).

Effects of CTX and CTX + Lactobacillus therapy on metabolic pathways.
The relative abundance of each functional category (biosynthesis, degradation/utilization/assimilation/, generation of precursor metabolite and energy, glycan pathways and metabolic clusters) was estimated as pathway abundance and relative abundance of read counts ( Figure S13). Differential abundance was mainly found for respiration, fermentation, fatty acid/lipid/carbohydrate degradation and biosynthetic pathways ( Figure S13A). Similarly, raw counts for metabolic pathways and enzyme, metabolites and reaction orthologs in MetaCyc showed a strong statistical significance of differential abundance mainly for cofactor, prosthetic group, electron carrier, vitamin, fatty acid and lipid biosynthesis ( Figure S13B). Some specific metabolic pathways in MetaCyc database can be labeled even with a light-level bacterial taxon [76]. Therefore, we also used MetaCyc to identify metabolic pathways and/or bacterial taxa specifically related to the five groups of rats for cancer therapy ( Figure 5). MetaCyc L-methionine salvage cycle III was linked to a high Clostridiales-level in CK control group (PWY-7527, Figure 5A). Another pathway (PWY-7839), 6-hydroxymethyl-dihydropterin diphosphate biosynthesis I, which converts GTP into pterin precursors (methanopterin and sarcinapterin) for the biosynthesis of several cofactors in specific bacterial strains, was found to be particularly highly expressed in CK and Lactobacillus-treated samples due to not only to increased Lactobacillus-levels, but also to an increase in S24-7 Muribaculaceae, Prevotella, Clostridiales, Bacteroides and CF231 Paraprevotellaceae ( Figure 5B Figure 5E). In addition, abundance of helicobacterial taxa required for the TCA cycle (tricarboxylic acid cycle) or the Krebs cycle was also (re)-established by Lactobacillus treatment. Not only Helicobacter-levels, but also the levels of Flexispira, Rothia and Halomonas were maintained despite chemotherapy using Lactobacillus treatment ( Figure 5F). Only Bacillales were not maintained by L-, M-or H-doses of Lactobacillus. Higher doses (> 5.0 ml/kg) of 3L bioproduct were probably required to control Bacillales. High dose (5.0 ml/kg) of 3L was required to control Halomonas ( Figure 5F). Importantly, a formaldehyde oxydation peak was noticed in IM samples. This was due to the development of Enterococcus  Figure 5K). In contrast, tRNA charging was strongly affected by the CTX chemotherapy and/or the treatment

Discussion
Cancer as cell disease caused by changes to DNA is a major burden of threat to human health worldwide.
Even more burden of threat to human health comes when chemotherapy is envisioned as the primary or maybe sole way to prevent cancer development. Using one or more anti-cancer chemical drugs such as cyclophosphamide (cytophosphane, CTX) kills lymphoma or any cancer cells, but it kills or seriously alters also the patient immune system, challenging life and health expectancies perhaps just like the disease itself. This is clearly shown in our study when a strong beneficial healthy gut flora is suppressed by cancer and CTX treatment. Here, we have analyzed the microbiome of five groups of rats in relation with cancer, revealing the completely altered microbiome of rats treated by CTX. Very importantly, we show that lactobacillus treatments are particularly efficient to maintain the normal healthy control gut bacterial flora in the rat intestine and therefore able to establish strong healthy conditions in rats despite chemotherapy. Though there are earlier reports regarding the use of Lactobacillus in gut flora and gastrointestinal tract protection [77][78][79][80], we will always seek for the best solution and we still seek for a probiotic remedy with strong and significant impact on the adversive effects developed by CTX therapy, active on benefical microbes, improving microbial balance, activating nutrients and stimulating gut-powered immune systems.
Use of probiotics in cancer is still rare, not applicable to all ages and populations, an emerging field with many contradictory clinical results when comes to interaction with host or consumption by human patient [81][82]. Microbiota manipulation by natural probiotics vs chemical drugs is a constant challenge in human and veterinary medicine, in particular for genetic diseases such as cancer. The safety and stability of chemotherapeutic drugs such as CTX in clinical trials for cancer are doubtful [83][84]. We always look into clinical trials for cancer with high efficiency, with reduced resistance capacity and without affecting the quality of life and patient's health conditions. One has an ability to develop resistance to some chemicals.
Repeated use of the same class of chemicals to control a disease such as cancer can cause many undesirable effects. When one becomes sick and resistant, the chemical (CTX) is used more frequently and the adjuvant must ultimately be added as CTX increases. We have not conducted yet a comprehensive microbiological medical study of the impact of 3L on cancer. We have analyzed the rat response to a new bioproduct (3L: L. acidophilus + L. casei + L. plantarum) in addition to cyclophosphamide CTX. We have demonstrated that the association of acidophilus, casei and plantarum is particularly beneficial to maintain the gut flora and therefore the immune system during chemotherapy for cancer treatment. We present that connection of three Lactobacilli strains has strong beneficial effects on gut microbiota during chemotherapy.
Here, we present MiSeq data for the variations in the microbiome of rat feces under five different conditions: healthy normal, immune-attack (CTX chemotherapy), low, medium and high doses of Lactobacillus (3L), respectively, with special attention for fungal phylum, class, order, family, genus and species associated with cancer and its treatment with one anti-cancer cyclophosphamide drug. In this study, we demonstrate that rats treated with CTX + high doses of a new natural biomedical product (3L) have preserved microbiota of prime importance to sustain strong immune system and healthy condition. Such results are particularly important since it has been clearly established that bacterial dysbiosis accompanies carcinogeneis in malignancies as various as colon, liver and pancreas cancers. In particular, the growth of nocive fungi (Malassezia species) in the gut microbiome can promote oncogenesis via activation of mannose-binding lectins [85], which urges medication to control microbiome and MBL.
It is known that chemotherapy during cancer treatment adversely affects the composition of the gut flora, alters immune, metabolic and physiological functions, and potentially stimulates invasive fungal infection [86][87][88]. So it is not very surprising to find that CTX alters gut flora and provokes an increase in mucor or xerochrysium infection in groups of rats (see Figs. 3-4, S9-S12 & Tables 1-2). The striking finding of our study in five groups of rats in relation with cancer and chemotherapy is the beneficial regulation of the gut flora after injection of different doses of Lactobacillus. High doses of 3L are shown to maintain normal gut flora in rats treated with CTX (see Figs. 2-5, S9-S12 & Tables 1-2), urging to apply the method or lactobacillus treatment in human cancer therapy. We show here that high doses of L. acidophilus + L. casei + L. plantarum are particularly relevant to regulate the levels of Ascomycota and Capnodiales. This may be important discovery for cancer treatment due to the finding that ascomycetes are known to be used in medicine with the antibiotics penicillin and cephalosporin [89], while endophytic sooty mold fungi (Dothideomycetes) can be important for tissue health, environmental adaptation and stress tolerance [90]. of these fungi are plant and human pathogens, many of them are known as biodegraders and biocontrol agents which could be exploited in medical applications [91]. For instance, this could be exploited for cancer treatment if we consider that these ascomycete fungi can help degrade residual CTX and control cyclophosphamide toxicity [92]. Another important aspect of using lactobacillus for cancer treatment is that 3L (high dose, 5.0 ml/kg) is shown to keep pathogens such as mucormycetes (Mucormycota, Mucoraceae) and chytridiomycetes (Olpidiomycota, Olpidiaceae) at bay (see Figs. 4, S9-S12 & Tables 1-2). High mucorales usually invade the blood vessels and are related to emerging infectious diseases such as mucormycosis beyond other zygomycoses [93]. Chytrid fungi cause chytridiomycosis, an emerging disease in amphibians, and a subcutaneous phycomycosis in humans [94]. Therefore, this is a complete pattern of many various fungal infections which can be regulated by 3L. Furthermore, 3L can control the magnitude of beneficial fungal components of the intestinal microbiota, i.e. the gut mycobiome in human health. Some fungi coming from the diet or the environment are important to mediate interaction in the gut bacterial communities and regulate metabolic homeostasis [see [72][73]. This is the case for Eurotiales, Hypocreales, Saccharomycetales and Trichosporonales that are crucial parts of healthy mycobiome in control and lactobacillus-treated rats (see Figs. 4, S9-S12 & Tables 1-2). It could be important to use 3L to maintain the mycobiome balance, microbial community interactions, bacterial-fungal interactions, fungal-fungal interactions and host-fungal interactions during cancer treatment [95]. Similarly to diabetis [96], CTX injections resulted in significant changes in Mortierellomycota-levels, which could be reversed by adding L3 (see Fig. S9A). So, using L3 seems to be extremely useful to target specific components of the mycobiome that plays a key role in the development of many various diseases from cancer to diabetis.
In this prospect for curing or optimizing health, it is probably important to note that the effect of 3L can be dose-dependent. Low dose of 3L cocktail (1.25 ml/kg) was required to treat the rats for Mallasseziomycetes (see Fig. 4C & Table 2), which could be important in medications for specific pathologies such as colorectal cancer. Fungi included in the class Mallasseziomycetes increase with late-stage colorectal cancer [97].
During the phases of chemotherapy treatment in rats, we also find that 3L doses work differently on the mycobiome. For threeL to be effective, quantities of lactobacillus ingredients eventually need to be gradually increased during cancer treament for a specific target fungal group, including Aspergillaceae,  4D & Tables 1-2). Regulation of the fungal mycobiome is an important aspect of cancer treatment and chemotherapy. The gut mycobiome is implicated in microbiome assembly and immune functionality, which prompts the modulation of specific fungi to regulate both gut microbiome and immune system in therapeutic approaches for cancer [98]. However, while 3L is effective to modulate yeasts in the order Saccharomycetales (Ascomycota; see Fig. 4 & Table 2), Pichia is one of the rare examples of fungal genera that are affected by CTX chemotherapy but do not respond to lactobacillus treatments (see Fig. 4E & Tables 1-2). A possible explanation might be that 3L rather controls gut fungi, but not the oral fungi or genera such as Pichia and Candida [99]. Accordingly, a specific oral fungal medication needs to be prescribed to target Pichia and/or Candida. Perhaps the formulation of 3L can still be improved to efficiently treat both oral and gut mycobiomes. Using one or more Lactobacillus strain such as L. reuteri in addition of 3L may be very effective to control the complete human mycobiome on oral and gut diseases (see Jørgensen et al., 2017 [100] & our study).
However, one major striking finding of our study about Lactobacillus on cancer disease was the plethora of metabolic functions maintained by 3L treatment in five groups of rats in relation with CTX chemotherapy (see Figs. 5 & S13). Not only cancer disease, but also chemotherapy heavily alters cell metabolism [101][102][103][104]. Therefore, targeting of cancer (and chemotherapy) metabolism, as a complementary strategy, is a rather promising approach not only for therapeutic intervention on the disease, but also for preservation of physiological function in the patient immune system. We show here that a preparation of L. acidophilus (SD65), L. casei (SD07) and L. plantarum (SD02) is extremely efficient not in curing cancer, but in maintaining and/or stimulating many various metabolic pathways via a beneficial effect on the gut flora (see Figs. 5 & S13). After treatment with our preparation of three lactobacilli cocktail (3L), the gut flora of rats subject to chemotherapy is rich in Firmicutes-Clostridia-Clostridiales-S24-7, as found for healthy control conditions (see Fig. 5). Overall, the detail composition of the bacteriome with the MetaCyc data show shared patterns of microbial strains and metabolic activities in healthy and 3L-treated groups for many various systems. Bacteroidales-Muribaculaceae-S24-7are important components of the microbiome for carbohydrate metabolism, while genes for amino acid and vitamin metabolism are upregulated in Deferribacteraceae [105].
on the use of lactobacillus to stimulate amino acid, biotin, carbohydrate, glucose, iron, nitrogen, oxygen, phosphorus, protein, pyruvate, sulfide and vitamin metabolism as alternative or complement to chemotherapy for cancer (see Fig. 5). However, the most striking effect of Lactobacillus treatment on the microbiome is found for Enterococcus (see Fig. 5G-M). Enterococcus infection (bacteraemia, peritonitis, endocarditis and urinary tract infection) is significantly increased by chemotherapy (see Fig. 5G), which becomes a risk of ulcerative colitis or even a risk of colorectal cancer [121]. To raise bacteroides and clostridiales and keep Enterococcaceae at bay in order to balance the gut flora to normal healthy conditions during chemotherapy, we find that probiotics such as lactobacilli are very suitable (see It is known that different formulations of lactobacilli can be used to modify the gut flora and thereby general metabolism and behavior not only in fishes and rodents, but also in humans [79,[122][123]. It is also known that lactobacilli or other general natural bacterial probiotics can be used as an adjuvant treatment during anticancer therapy to maintain the gut flora and stimulate the patient' Lactobacillus necessary to counteract specific aversive or inhibitory effects of one-drug anticancer therapy.
In particular, a systematic analysis of the gut microbiota in groups of rats in relation with cancer disease and cyclophosphamide therapy reveals the infectious bacterial pathogens that are induced by the treatment, but also the key beneficial bacterial families maintaining the immune system and protecting effects of 3L.
L. casei complements the growth of L. acidophilus, a producer of carbohydrate-digesting enzymes, while L.
plantarum, a more flexible and more versatile strain, produces a battery of antimicrobial substances that will help them to survive in the gastrointestinal tract in any condition. Our results in CTX-rats might pave the way to future attempts in cancer chemotherapy for humans. Fragile health conditions and microbial infections are often associated with a weakened immune system as a main consequence of "chemo". Hence, complementary and/or alternative clinical medicine must be provided for cancer prevention and/or treatment.
One major finding in our study is not that cancer can be cured by 3L, but that the body can be rescued from chemo by a tritherapy of lactobacilli via many beneficial effects on gut flora (mycobiome and microbiome) and energy metabolism. Body composition (gut flora and metabolism) is strongly linked to the immune system that relies on energy to break down the risk of chronic diseases. Importantly, it is very likely that 3L works as additive to chemical drugs not only for cancer, but also for many various metabolic diseases because the tritherapy has significantly strong beneficial effects on many various bacteria and metabolic systems (see Fig. 6). It boots the immune system and depletes specific invading agents, such as fungal pathogens (mucor) and infectious bacteria (enterococcus). Also importantly, the panoply of bacteria that are upregulated by 3L treatment (see Fig. 6) is such that our probiotic formula (L. acidophilus SD65, L. casei SD07 and L. plantarum SD02) can be easily improved to target specific metabolic systems in humans. One merit of our work in rats is that different doses of lactobacillus have differential effects on the gut flora (see Fig. 6). So, further research should be undertaken to test the effects of different SD65-SD07-SD02 ratio in 3L mixtures or in combination with other mixtures of lactobacilli or beneficial bacteria (see Fig. 6). Here, we deliver a proof of research feasibility for immunization or improved immune systems in cancer. 3L may serve as a basis for the development of a large family of medicinal microbial bioproducts to be used not on chemo-animal models, but on cancer patients subject to cyclophosphamide therapy.

Conclusion
The specific benefit of the formulated 3L probiotic is a strong relevant action on the gut microbiome in chemotherapy conditions. While CTX and anticancer drugs have multiple potential adverse effects, which is also documented here, adding 3L in anticancer therapy is proven to be beneficial to maintain gut flora and thereby health conditions. Fungal-bacterial interactions control health and disease. 3L maintains fungi and bacteria linked to health. Results on bacteria are limited to metabolic pathways, which seems to be the way 3L deals with chemotherapy. The results in rats are such that the efficiency, frequency and reliability of high-            Comparative analysis of fungal abundance between metagenomes from control (CK) and IM groups in rats.
Trichosporonales (Basidiomycota) are differentially abundant fungi between two groups (CK and IM) of multiple samples, as found for 3L (high-and middle-dose) Lactobacillus therapy (see Fig. S9). the relative abundance of the mean and median reflect the size of the differences between groups (see Fusarium*). B) Intergroup differential classification unit display chart based on classification tree (1e-04, 1e-05). Hierarchical relationships of all taxa from phylum to genus (i.e. from the inner circle to the outer circle) in the sample set. The node size corresponds to the average relative abundance of the taxon. The letter a identifies the name of the classification unit where there is a significant difference between groups. Fusarium has a higher abundance in the sample or group (CK) represented by green (see *, R(v3.1.1), BGI Co., Ltd).