WO2018229519A1

WO2018229519A1 - Methods for diagnosing breast cancer

Info

Publication number: WO2018229519A1
Application number: PCT/HU2018/050029
Authority: WO
Inventors: Péter BAY; James J. Goedert; Tünde KOVÁCS; Edit MIKÓ; András VIDA
Original assignee: Debreceni Egyetem; U.S. National Cancer Institute
Priority date: 2017-06-12
Filing date: 2018-06-12
Publication date: 2018-12-20
Also published as: EP3638816A1

Abstract

The invention discloses methods for diagnosing early phases of breast cancer by detecting a decrease of the abundance of bacteria capable of producing lithocholic acid. In particular, methods for the diagnosis of early stage breast by the assessment of bacteria capable of producing lithocholic acid under physiological conditions in the human body are provided.

Description

METHODS FOR DIAGNOSING BREAST CANCER

FIELD OF THE INVENTION

The invention relates to methods for diagnosing breast cancer by determining the level of bacteria capable of producing lithocholic acid. In particular, methods for the diagnosis of early stage breast cancer are provided.

BACKGROUND OF THE INVENTION

The human body harbors a vast number of symbiotic, commensal and pathogenic bacteria in the bodily cavities and the body surface. The ensemble of these microbes is referred to as the microbiota and its collective genome as the microbiome. Recent advances pointed out that changes in the composition of the microbiome and certain bacterial metabolites crucially impact the metabolic, behavioral, cardiovascular or immune function of the host and have pivotal role in diseases that were previously not associated with bacteria^1"4. Alterations of the microbiome are associated with certain cancers. Although, the bacterial microbiota may have a widespread role in carcinogenesis, the number of directly tumorigenic bacteria is extremely small, some 10 bacterial species fall into this category⁵. It seems more likely that pathological changes in the microbiota/microbiome (dysbiosis) determine susceptibility to the disease or influence the progression of the disease⁴.

Most of these cancers affect those organs that are/can be directly in contact with microbes such as the urinary tract⁶, cervix⁷, skin⁸, airways⁹, or the colon⁴. These microbiome-host interactions are best characterized in the colon. In the intestine a breach of the biological barrier between the microbes and the underlying tissues enables an adverse physical contact between microbes and host cells that induces the production of paracrine bacterial metabolites⁴. Through these, the microbiome modulates tumorigenesis, tumor promotion, severity of the disease, and chemotherapy effectiveness in colonic tumors⁴. Direct stimulation of the cancer cells by bacteria probably has a role in bacteria-mediated induction of lymphomas^{10 11} and possibly prostate cancer⁶.

Much less is known of the role of the microbiome in the regulation of those tumors that are located in different compartments and are only connected to the microbiome through the circulation. Bacterial metabolites are likely candidates to be transported to tumor by the bloodstream to exert their affects in distant tumors. For hepatocellular carcinoma, lipopolysaccharide¹² and deoxycholic acid (DCA)¹³ have been identified as promoters, while a short chain fatty acid (SCFA), propionate¹⁴ is an inhibitor. Bacterial metabolites may have indirect effects too, for example, DCA alters the secretory profile of hepatic stellate cells and it is this altered secretome that acts in a pro- proliferative fashion¹³.

Numerous bacterial metabolites have been identified that are either the microbes' own metabolites (e.g., short chain fatty acids, lactate, pyruvate) or modified products of the host (e.g., secondary bile acids, metabolites of aromatic amino acids, redox-modified sex steroids)^{15 16}. These bioactive metabolites act through various pathways that involve the modification of gene expression (e.g., activation of histone deacetylases and other lipid-mediated transcription factors) or the modulation of signal transduction in the host.

Bile acids are among the candidate role-players in both the tumorigenic and the antitumor pathways. Primary bile acids (cholic acid (CA) and chenodeoxycholic acid (CDCA)) are formed in the liver, and are converted to secondary bile acids (mainly deoxycholic acid (DCA) and lithocholic acid (LCA), respectively) by the intestinal microbiota. LCA has been shown to kill cultured human neuroblastoma and breast cancer cells as well as rat glioma cells in cell culture (Goldberg 2011, Batta et al. 2018).

WO2014126044 describes a method for assessing the risk of carcinogenesis by detecting an elevated level of e.g. lithocholic acid. A method to reduce the level of LCA by antibacterial agents is also suggested. It is also hypothesized that the increased abundance of bacteria capable of producing secondary bile acids is associated with a higher cancer risk or cancer.

WO2014146202 describes that an alteration in the microbiota of the breast is associated with cancer, e.g. breast cancer. Several bacterium genera were found to show an increased abundance in patients suffering of cancer, a. o. Pseudomonas, Staphylococcus, Acinetobacter, and Bacillus. Further, a method of treating or preventing breast cancer is claimed, comprising the administration of bacteria including e.g. Lactobacillus, Eubacterium and Bifidobacterium.

WO2014130162 describes a method for determining the presence or risk of a hormonally sensitive cancer, wherein the level of microbial DNA in a test sample is compared to a level of bacterial DNA in a control sample. However, WO2014130162 describes that the bacterial DNA to be determined is derived from a bacterium that degrades estrogen.

Goedert and co-workers²⁴ have assessed microbiome changes in breast cancer patients, finding that postmenopausal breast cancer patients had reduced diversity and altered composition of the gut microbiome compared to closely matched control women. In the order Clostridiales, case patients had higher levels of Clostridiaceae, Faecalibacterium, and Ruminococcaceae; and they had lower levels of Dorea and Lachnospiraceae.

SHORT DESCRIPTION OF THE INVENTION

It has been found that the presence of breast cancer may be determined by the assessment of the abundance of bacteria capable of producing lithocholic acid (LCA). In particular, the early stages of breast cancer are characterized by the decreased abundance of such bacteria. It has also been found that not only the abundance of such bacteria, but also the levels of a bacterially produced secondary bile acid, namely LCA, are indicative of breast cancer.

The invention provides a method for diagnosing early stage breast cancer in a human subject, comprising assessment of the abundance of at least one bacterium species, which is capable of producing LCA under physiological conditions in the human body, in a test sample derived from the subject, wherein the test sample comprises human microbiota of said subject, wherein a decrease in the abundance of the at least one species in the test sample compared to a reference value of abundance of the at least one bacterium species typical of the absence of early stage breast cancer,

is indicative of early stage breast cancer in said subject.

The invention provides a method for diagnosing early stage breast cancer in a human subject, comprising assessment of the molar ratio of LCA to LCA precursor bile acid in a test sample of said subject, wherein a decrease in said ratio in the test sample of said subject compared to a reference ratio of LCA to the LCA precursor bile acid typical of the absence of early stage breast cancer,

is indicative of early stage breast cancer in said subject.

The invention provides a method for diagnosing early stage breast cancer in a human subject, comprising assessment of the level of lithocholic acid in a test sample derived from said subject wherein a decrease in the level of LCA in the test sample of said subject compared to a reference level of LCA typical of the absence of early stage breast cancer,

is indicative of early stage breast cancer in said subject.

Preferably, the method for diagnosing early stage breast cancer in a human subject comprises the assessment of the level of lithocholic acid in a test sample derived from said subject wherein a decrease in the level of LCA in the test sample of said subject compared to a reference level of LCA typical of the absence of early stage breast cancer,

is indicative of early stage breast cancer in said subject.

Preferably, the method for diagnosing early stage breast cancer in a human subject comprises the assessment of the molar ratio of LCA to LCA precursor bile acid in a test sample of said subject, wherein a decrease in said ratio in the test sample of said subject compared to a reference ratio of LCA to the LCA precursor cholic acid typical of the absence of early stage breast cancer,

is indicative of early stage breast cancer in said subject.

In a highly preferred embodiment the method comprises assessing the abundance of at least one bacterium species which is capable of producing LCA under physiological conditions in the human body, in a test sample derived from the subject, wherein the test sample comprises human microbiota of said subject, wherein a decrease in the abundance of the at least one species in the test sample compared to a reference value of abundance of the at least one bacterium species typical of the absence of early stage breast cancer is indicative of early stage breast cancer in said subject,

and

assessing the level of lithocholic acid in a test sample derived from said subject wherein a decrease in the level of LCA in the test sample of said subject compared to a reference level of LCA typical of the absence of early stage breast cancer,

is indicative of early stage breast cancer in said subject

and optionally /or

assessing the molar ratio of LCA to LCA precursor bile acid in a test sample of said subject, wherein a decrease of said ratio in the test sample of said subject compared to a reference ratio of LCA to the LCA precursor bile acid typical of the absence of early stage breast cancer,

is indicative of early stage breast cancer in said subject.

In some embodiments the method further comprises assessing the level of any one of the compounds selected from the group consisting of: total bile acids, cholic acid (CA), chenodeoxycholic acid (CDCA), ursodeoxycholic acid (UDCA), deoxycholic acid (DCA) and combinations thereof, in a test sample of said subject, wherein a decrease in said level in the test sample of said subject compared to a reference level typical of the absence of breast cancer, is indicative of breast cancer in said human subject. In preferred embodiments the reference level is typical of the absence of early stage breast cancer.

In highly preferred embodiments the method for diagnosing early stage breast cancer is a method for diagnosing breast cancer stage 0 according to the American Joint Committee on Cancer (AJCC) TNM system. In embodiments for diagnosing stage 0 breast cancer the assessed feature (i.e. the abundance of the at least one bacterium species, the level of LCA, the molar ration of LCA to LCA precursor bile acid, the level of the compound) is compared to a corresponding reference value indicating the absence of stage 0 breast cancer and preferably also stage 1 breast cancer.

In certain embodiments the method for diagnosing early stage breast cancer is a method for diagnosing breast cancer stage 1 according to the AJCC TNM system. In embodiments for diagnosing stage 1 breast cancer the assessed feature is compared to a corresponding reference value indicating the absence of stage 1 breast cancer and preferably also stage 0 breast cancer.

In preferred embodiments at least the abundance of the bacterium species capable of producing LCA and the level of at least one of the aforementioned compounds assessed in the test sample indicate the presence of the breast cancer. In most preferred embodiments the abundance of the bacterium species and the level of LCA in the test sample indicate the presence of the breast cancer.

In preferred embodiments the abundance of at least two or at least three or at least four or at least five bacterium species is detected.

In an embodiment the at least two or at least three or at least four or at least five bacterium species are selected from the genera Clostridium, Eubacterium, Bacteroides, Escherichia, Peptostreptococcus, Pseudomonas, Ruminococcus, Staphylococcus, Dorea and the family Lachnospiraceae.

In a preferred embodiment the at least two or at least three or at least four or at least five bacterium species are selected from the genera Bacteroides, Escherichia, Clostridium, Pseudomonas and Staphylococcus.

In another embodiment the at least one species is or at the least two species or the at least three species or the at least four species are selected from

Bacteroides, Escherichia, Clostridium and Pseudomonas, or Bacteroides, Escherichia, Clostridium and Staphylococcus, or Bacteroides, Clostridium, Pseudomonas and Staphylococcus, or Escherichia, Clostridium, Pseudomonas and Staphylococcus, or Bacteroides, Escherichia and Clostridium, or Bacteroides, Escherichia and Pseudomonas, or Bacteroides, Escherichia and Staphylococcus, or Bacteroides, Pseudomonas and Staphylococcus, or Bacteroides, Pseudomonas and Clostridium, or Bacteroides, Clostridium and Staphylococcus, or Escherichia, Clostridium and Staphylococcus, or Escherichia, Clostridium and Pseudomonas, or Escherichia, Pseudomonas and Staphylococcus, or Clostridium, Pseudomonas and Staphylococcus, or Bacteroides and Escherichia, or Bacteroides and Clostridium, or Bacteroides and Pseudomonas, or Bacteroides and Staphylococcus, or Esch- erichia and Clostridium, or Escherichia and Pseudomonas, or Escherichia and Staphylococcus, or Clostridium and Pseudomonas, or Clostridium and Staphylococcus, or Pseudomonas and Staphylococcus.

In another embodiment the at least two or at least three or at least four or all bacterium species are selected from Staphylococcus haemolyticus, Escherichia coli, Bacteroides thetaiotaomicron, Clostridium sordellii and Pseudomonas putida.

In another embodiment the at least two bacterium species are S. haemolyticus and E. coli, or S. haemolyticus and B. thetaiotaomicron, or S. haemolyticus and C. sordellii, or S. haemolyticus and P. putida, or E. coli and B. thetaiotaomicron, or E. coli and C. sordellii, or E. coli and P. putida, or B. thetaiotaomicron and C. sordellii, or B. thetaiotaomicron and P. putida, or C. sordellii and P. putida. In another embodiment the at least three bacterium species are S. haemolyticus, E. coli and B. thetaiota- omicron, or S. haemolyticus, E. coli and C. sordellii, or S. haemolyticus, E. coli and P. putida, or S. haemolyticus,

B. thetaiotaomicron and C. sordellii, or S. haemolyticus, B. thetaiotaomicron and P. putida, or S. haemolyticus,

C. sordellii and P. putida, or E. coli, B. thetaiotaomicron and C. sordellii, or E. coli, B. thetaiotaomicron and P. putida, or E. coli, C. sordellii and P. putida, or B. thetaiotaomicron, C. sordellii and P. putida.

In another embodiment the at least four bacterium species are S. haemolyticus, E. coli, B. thetaiotaomicron and C. sordellii, or S. haemolyticus, E. coli, B. thetaiotaomicron and P. putida, or S. haemolyticus, E. coli, C. sordellii and P. putida, or S. haemolyticus, B. thetaiotaomicron, C. sordellii and P. putida, or E. coli, B. thetaiotaomicron, C. sordellii and P. putida.

According to an embodiment the abundance of the bacterium species to be assessed is detected by detecting the abundance of the DNA, RNA, protein of (or produced by) or any gene product of any one of the genes or ORFs (open reading frame) of the bai (bile acid inducible) operon.

According to an embodiment the gene or ORF of the bai operon whose abundance is to be detected is the baiH or the gene product is a product of baiH.

A method for diagnosing breast cancer in a human subject, comprising

assessing the level of any one of the compounds selected from the group consisting of: total bile acids, CA, CDCA, UDCA, DCA, LCA and combinations thereof, in a test sample derived from said subject,

wherein a decrease in said level as compared to a reference level typical of the absence of breast cancer is indicative of breast cancer.

In preferred embodiments at least the levels of total bile acids and cholic acid (CA); or total bile acids and CDCA; or total bile acids and UDCA; or total bile acids and DCA; or total bile acids and LCA; or at least the levels of CA and CDCA; or CA and UDCA; or CA and DCA; or CA and LCA; or

at least the levels of UDCA and CDCA; UDCA and DCA; or UDCA and LCA or

at least the levels of DCA and CDCA; or DCA and LCA; or LCA and CDCA are assessed.

In particular at least the levels of total bile acids, CA and CDCA; or total bile acids, CA and UDCA; or total bile acids, CA and DCA; or total bile acids, CA and LCA; or total bile acids, CDCA and UDCA; or total bile acids and CDCA and DCA; or total bile acids and CDCA and LCA; or total bile acids and UDCA and DCA; or total bile acids and UDCA and LCA; or total bile acids and DCA and LCA; or CA, CDCA and UDCA; or CA, CDCA and DCA; or CA, CDCA and LCA; or CA, UDCA and DCA; or CA, UDCA and LCA; or CA, DCA and LCA; or CDCA and UDCA and DCA; or CDCA and UDCA and LCA; or CDCA and DCA and LCA; or UDCA and DCA and LCA are assessed.

In particular at least the levels of total bile acids, CA, CDCA and UDCA; total bile acids, CA, CDCA and DCA; total bile acids, CA, CDCA and LCA; total bile acids, CA, UDCA and DCA; total bile acids, CA, UDCA and LCA; total bile acids, CA, DCA and LCA; total bile acids, CDCA, UDCA and DCA; total bile acids, CDCA, UDCA and LCA; total bile acids, CDCA, DCA and LCA; total bile acids, UDCA, DCA and LCA; CA, CDCA, UDCA and DCA; CA, CDCA, UDCA and LCA; CA, CDCA, DCA and LCA; CDCA, UDCA, DCA and LCA are assessed. In particular at least the levels of total bile acids, CA, CDCA, UDCA and DCA; total bile acids, CA, CDCA, UDCA and LCA; total bile acids, CDCA, UDCA, DCA and LCA; total bile acids, CA, UDCA, DCA and LCA; total bile acids, CA, CDCA, DCA and LCA; CA, CDCA, UDCA, DCA and LCA are assessed.

In most preferred embodiments the levels of total bile acids, CA, CDCA, UDCA, DCA and LCA are assessed.

The method may optionally comprise any one of the following steps

a) assessing the abundance of at least one bacterium species, which is capable of producing LCA under physiological conditions in the human body, in a test sample derived from the subject, wherein the test sample comprises human microbiota of said subject, wherein a decrease in the abundance of the at least one species in the test sample compared to a reference value of abundance of the at least one bacterium species typical of the absence of breast cancer, is indicative of breast cancer in said subject,

b) assessing the molar ratio of LCA to LCA precursor bile acid in a test sample of said subject, wherein a decrease of said ratio in the test sample of said subject compared to a reference ratio of LCA to the LCA precursor bile acid typical of the absence of breast cancer, is indicative of breast cancer in said subject,

c) assessing the level of lithocholic acid in a test sample derived from said subject wherein a decrease in the level of LCA in the test sample of said subject compared to a reference level of LCA typical of the absence of breast cancer, is indicative of breast cancer in said subject,

d) any combination of a)-c), i.e. a) and b) or a) and c) or b) and c) or a), b) and c).

In a preferred embodiment the method of the invention is a method for diagnosing breast cancer stage 0 and/or stage 1 (early stage) according to the American Joint Committee on Cancer (AJCC) TNM system.

In certain preferred embodiments the method for diagnosing breast cancer is a method for diagnosing stage 1 breast cancer. In embodiments for diagnosing stage 1 breast cancer the assessed feature is compared to a corresponding reference value indicating the absence of stage 1 breast cancer and preferably also stage 0 breast cancer. The sample from the subject to be assessed may be serum, plasma, whole blood, breast duct fluid, breast tumor tissue or feces. The sample in which the abundance of bacteria is to be assessed is preferably feces. The sample in which the level of a bile acid is to be determined is preferably serum. The samples to be compared are corresponding samples, e.g. when the abundance of a bacterium species is determined in a feces sample, the reference value is derived from feces.

Preferably the bacterium species capable of producing LCA under physiological conditions in the human body is capable of said conversion in the human gastrointestinal tract.

American Joint Committee on Cancer (AJCC) (TNM system and its anatomic) staging (stage grouping) system was used according to the 7^th edition.

SHORT DESCRIPTION OF THE DRAWINGS

Figure 1. In early stages of human breast cancer bacterial LCA biosynthesis is suppressed

Serum was pooled from the healthy controls and breast cancer patients of cohort 1. (A) The bile acid composition of these pooled samples was determined. (B) By summing the different bile acid species total serum bile acid content was calculated. (C) Serum CDCA and (D) LCA levels from the samples of healthy controls and breast cancer patients are plotted, LCA/CDCA ratio was calculated from the samples of healthy controls and breast cancer patients.

Figure 2. The abundance of the baiH DNA of different species of bacteria (Staphylococcus haemolyticus, Escherichia coli, Bacteroides thetaiotaomicron, Clostridium sordelli, Pseudomonas putida) was determined in the fecal DNA samples of cohort. * indicates p<0.05

Figure 3. LCA inhibits the proliferation of breast cancer cells and reduces cancer aggressiveness in vivo (A) MCF7 and 4T1 cells and primary fibroblasts were treated with LCA in the concentrations indicated for 48 hours then total protein concentration was determined in SRB assays (MCF7: n=8; 4T1 : n=6; fibroblasts n=5). Values are fold increases, where 1 means protein content in the control cells. (B) MCF7 and 4T1 cells were treated with LCA in the concentrations indicated for 7 days and colonies were stained according to May-Grtinwald- Giemsa that were then counted using the Image J software (MCF7, 4T1 : n=6). (C) MCF7 and 4T1 cells were treated with LCA in the concentrations indicated for 48 hours. Dead cells were stained by propidium iodide (PI) and analyzed by flow cytometry (MCF7, 4T1 : n=3). (D-G) Female Balb/c mice were grafted with 4T1 cells as described and were treated with LCA (15 nmol q.d. p.o.) or vehicle (VEH) (n=8/8) for 18 days before sacrifice. Upon autopsy mice (D) tumor infiltration was scored, (E) the mass and (F) number of metastases, furthermore (G) "Tc-Mibi uptake were determined.

On panel D significance was calculated using the Freeman -Halton extension of the Fisher exact probability test for a 2x3 contingency table. * and ** indicate statistically significant difference between vehicle and treated groups at p<0.05 or p<0.01, respectively.

Figure 4. Other secondary bile acids such as UDCA or DC A were without effect on MCF7 and 4T1 breast cancer cells.

Figure 5. LCA treatment reduces EMT and improves antitumor immune response

A part of the experiments were performed on MCF7 and 4T1 cells treated with LCA in the concentrations indicated for 48 hours, or on female Balb/c mice grafted with 4T1 cells that were treated with LCA (15 nmol q.d. p.o.) or vehicle (VEH) (n=8/8) for 18 days.

(A-B) In 4T1 cells and mouse-derived tumor samples the expression of a set of genes involved in EMT were determined in RT-qPCR reactions and Western blotting. (4T1 - mean + SD; in vivo - mean marked by a line). (C-D) In LCA or VEH-treated (CTL) MCF7 and 4T1 cells (C) cellular morphology was assessed after Texas Red- X Phallodin- and To-Pro-3 staining (representative figure), (D) cellular impedance was measured in ECIS exper- iments (mean + SD). (E) VEGF expression was determined in tumors using RT-qPCR (mean marked by a line). (F) The morphology in LCA or VEH-treated tumors were assessed in hematoxilin-eosine stained 4 μιη histological sections and the number of tumor infiltrating lymphocytes (TIL, marked by arrows, mean marked by a line) was counted in the sections. Stars mark typical tumor cells. (G) The expression of a set of immunological markers was determined in total cDNA prepared from the tumors in RT-qPCR reactions (mean marked by a line).

Scale bar on panel (B) is 10 μιη and 100 μιη on panel (F).

*, ** and *** indicate statistically significant difference between vehicle and treated groups at p<0.05, p<0.01 or p<0.001, respectively.

Figure 6. LCA interferes with multiple anticancer molecular pathways (A) LCA treatment improved total impedance in both MCF7 and 4T1 cells (B) LCA-treated 4T1 cells were slower in moving into a void area in a scratch assay as compared to vehicle-treated ones. * and ** indicate p<0.05 and p<0.01, respectively.

Figure 7. LCA treatment exerts anti-Warburg features (A) induction of mitochondrial oxidative metabolism upon LCA treatment marked by increased mitochondrial membrane potential (Δψ) and (B) succinate and (C) expression of a set of mitochondrial genes in 4T1 and MCF7 cells. (D) LCA not only boosted NRF1 expression but also enhanced its nuclear translocation (E) Using the kmplot.com database we found that while high NRF1 expression is also a negative regulator of breast cancer, higher PGC-Ια expression does not improve survival. *, **_; *** _and **** i_ncjicate p<0.05, p<0.01, p<0.001 and p<0.0001, respectively.

Figure 8. LCA treatment exerts anti -Warburg features

(A-B) MCF7 and 4T1 cells were treated with LCA in the concentrations indicated for 48 hours then the indicated measurements were performed. (A) Extracellular acidification rate (ECAR) (average+SD of a representative measurement) and (C) oxygen consumption rate (OCR) (average+SD of a representative measurement) were performed and data were plotted. (B) Intracellular lactate levels (MCF7, 4T1 : n=2) and (D) citrate levels and (E) citrate/lactate ratio were determined (MCF7, 4T1 : n=2) and were plotted. (F) MCF7 and 4T1 cells were treated with LCA in the concentrations indicated for 48 hours then cells were loaded with 10 mM 13C-acetate for 1 hour that was followed by the determination of the indicated metabolites. (G) MCF7 and 4T1 cells were treated with LCA in the concentrations indicated for 48 hours then cells were loaded with 10 mM 13C-glucose for 1 hour that was followed by the determination of the indicated metabolites. (H) MCF7 and 4T1 cells were treated with LCA in the concentrations indicated for 48 hours then protein extracts were separated by PAGE, blotted onto nitrocel- lulose and probed with the antibodies indicated. (MCF7, 4T1: n=3) (H/in vivo) Female Balb/c mice were grafted with 4T1 cells were treated with LCA (15 nmol q.d. p.o.) or vehicle (VEH) for 18 days. Protein, extracted from the primary tumors, was separated by PAGE, blotted onto nitrocellulose and probed with the antibodies indicated.

* and ** indicate statistically significant difference between vehicle and treated groups at p<0.05 or p<0.01, respectively.

DETAILED DESCRIPTION OF THE INVENTION

To date no direct, causal relationship had been shown between the microbiome and breast cancer, although studies have already suggested an interconnection. In breast cancer gut microbiome had been suggested to facilitate breast cancer progression through deconjugating estrogens making them more prone for reuptake. Goedert and co-workers²⁴ have assessed microbiome changes in breast cancer patients, finding that postmenopausal breast cancer patients had reduced diversity and altered composition of the gut microbiome compared to closely matched control women. Bacteria were identified on the surface of the breast, in the ducts and the microbiome of the breast also changes in breast cancer patients^28"31.

The term "human microbiota" refers to the microbes capable of living in or on the human body. The terms "gut microbiota" and "gut microbiome" refer to species of the human microbiota living in the human gastrointestinal tract. The term "microbiome" and "microbiota" as used in the description may refer to both the human microbiota and human microbiome and the gut microbiota and gut microbiome, preferably to the gut microbiota and microbiome. The causative relationship between the microbiome and breast cancer is further strengthened by the negative association between antibiotic treatment and prevalence or recurrence of breast cancer²²^. Although the literature is not equivocal on all observations, it seems that high doses of antibiotic treatment (over 500 pills) increase the risk for breast cancer. The negative correlation between antibiotic use and breast cancer incidence was also demonstrated in men³⁷. The strongest correlations between antibiotic treatment and breast cancer incidence was found when tetracyclines and macrolides were combined³⁴. Importantly, Wirtz and colleagues³⁵ have shown that fluoroquinolones dosing over 101 days had a tendency to increase the risk for a second breast cancer event and for breast cancer recurrence. These are in line with our finding that CPX treatment shortened life expectancy in breast cancer patients. In mice CPX increased breast cancer risk³⁸ that correlates with our survival experiment. It is also of note that CPX treatment decreases serum LCA in mice.

It has been surprisingly found that the abundance of bacteria capable of producing LCA may be indicative of certain stages of breast cancer.

We investigated how bile acid and LCA metabolism changes in breast cancer in humans. LCA is produced through deconjugation of chenodeoxycholic acid (CDCA) conjugates, followed by a dehydroxylation on carbon 7 by the action of the enzyme 7α/β hydroxysteroid dehydrogenase (7-HSDH). The enzymes involved in the 7- dehydroxylation of bile acids are organized into one operon called the bile acid-inducible (bai) operon wherein the baiH ORF codes for 7-HSDH in most bacterial species.

baiH abundance was assessed by amplifying baiH ORF from fecal DNA using specific primers. To validate this mode of measurement, we treated mice with ciprofloxacin (CPX), an antibiotic that specifically kills aerobic bacteria, while leaving the anaerobic bacteria intact. When mice were treated with CPX (200mg/kg q.d. for two weeks) the abundance of staphylococcal, escherichial and pseudomonal baiH (aerobic bacteria) decreased, while the ratio of the baiH of the anaerobic bacteria {Bacteroides fragilis, Clostridium scindens) did not change, as expected from the biology of the antibiotic that supports this approach.

The term reference value of abundance of a bacterium species in a corresponding sample indicating the ab- sence of breast cancer (e.g. stage 1 breast cancer) refers to a sample of the same type (e.g. blood, feces) from an individual or a group of individuals not having the breast cancer (e.g. stage 1 breast cancer).

The reference value is derived from individual(s) not having the cancer (stage) to be diagnosed. Preferably, the reference value is derived from healthy individual(s).

The bile acids from which LCA may be produced are CDCA and UDCA. The ratio of LCA to the bile acid from which LCA may be produced, may be the ratio of LCA to CDCA or the ratio of LCA to CDCA+UDCA. The term " (an) LCA precursor bile acid" as used herein refers to CDCA, UDCA and preferably to CDCA and UDCA.

The term LCA or lithocholic acid refers to LCA produced by bacteria.

Early stage breast cancer, as used herein, refers to stage 0 and/or stage 1 breast cancer according to the AJCC TNM staging system.

Bacteria capable of producing LCA from another bile acid are bacteria capable of a step of the conversion of CDCA and/or UDCA into LCA, such as 7a-dehydroxylation of CDCA, 7 -dehydroxylation of UDCA. Both 7a- dehydroxylation and 7 -dehydroxylation pathways include multiple enzymatic steps (See e.g. Ridlon and Bajaj: The human gut sterolbiome: bile acid-microbiome endocrine aspects and therapeutics. Acta Pharmaceutica Sinica B 2015;5(2):99-105 and Jason M. Ridlon, Spencer C. Harris, Shiva Bhowmik, Dae-Joong Kang & Phillip B. Hylemon (2016) Consequences of bile salt biotransformations by intestinal bacteria, Gut Microbes, 7: 1, 22-39.)

The capability of a bacterial species/strain/genus to convert CDCA to LCA or CDCA and/or UDCA into LCA under physiological conditions may be tested in vitro. In an appropriate in vitro assay bacterial lysates and CDCA are mixed and at the end of the assay LCA, CDCA and optionally, UDCA are detected by e.g. mass spectrometry and the LCA/CDCA (optionally or preferably LCA/(CDCA+UDCA)) ratio is calculated. Although culturing (gut) microbes has its challenges, the skilled person will find guidance in e.g. Sommer Advancing gut microbiome research using cultivation Current Opinion in Microbiology 2015, 27: 127-132; Lagier et al. Culture of previously uncultured members of the human gut microbiota by culturomics NATURE MICROBIOLOGY DOI: 10.1038/NMICROBIOL.2016.203; Browne et al Culturing of 'unculturable' human microbiota reveals novel taxa and extensive sporulation Nature 533 26 May 2016. To mimic the natural growing conditions of the gut microbiota, one might use specialised testing conditions, e.g the i-screen platform by TNO. To determine whether a bacterium species is capable of coexisting with a human subject or is capable to grow in the gut, one might consult and search by enrichment of sequence databases such as those of the NIH Human Microbiome Project (as of 06 June 2017) or the Human Pan-Microbe Communities (http : //www . hp mcd .org/ as of 06 June 2017). It is also possible to sequence and compare a characteristic genetic portion of the bacterium of interest with reference samples of healthy humans.

LCA levels decreased in breast cancer patients as compared to age and sex matched healthy individuals (Fig. 1 A, Table 1). Total bile acid and LCA levels decreased in breast cancer patients as compared to age and sex matched healthy individuals (Fig. IB, Table 1) and we have observed a similar trend in all other bile acids we examined (Fig. 1, Table la).

Table 1 secondary bile acid levels (μιηοΙ/L)

Table la primary bile acid levels (μιηοΙ/L)

Since both primary and secondary bile acid levels were lower in breast cancer patients the ratio between che- nodeoxycholic acid (the substrate for LCA synthesis) and LCA in human serum was assessed. A slight decrease in the LCA/CDCA ratio between healthy individuals and breast cancer patients was found that was further aggravated when only stage 1 patients were assessed (Fig. ID). At later stages LCA/CDCA ratio normalized and even increased above the ratio of healthy individuals in stage 3 patients.

The term cholic acid is intended to encompass the conjugates of cholic acid (e.g. with taurin, i.e. taurocholic acid, TCA and with glycin, i.e glycocholic acid, GCA). Similarly, the terms chenodeoxycholic acid, ursodeoxycholic acid, deoxycholic acid and lithocholic acid may encompass the naturally occuring conjugates of these bile acids, e.g. and in particular taurochenodeoxycholic acid (TCDCA), glycochenodeoxycholic acid (GCDCA), tau- roursodeoxycholic acid (TUDCA), glycoursodeoxycholic acid (GUDCA), taurodeoxycholic acid (TDCA), gly- codeoxycholic acid (GDCA), taurolithocholic acid (TLCA), glycolithocholic acid (GLCA).

The term "total bile acids" refers to the sum of bile acids (primary, secondary, conjugates, salts, etc) which may be detected in a sample derived from a human subject. The term "total bile acids" specifically refers to the level of CA, CDCA, DCA, UDCA and LCA taken together. The term "total bile acids" may also specifically refer to the level of C A, GCA, TCA,CDC A, GCDCA, TCDCA, DCA, GDCA, TDCA, UDCA, GUDCA, TUDCA, and LCA, GLCA, TLCA taken together.

The "stage" of a cancer in this description is to be understood as a stage determined using the American Joint Committee on Cancer (AJCC) TNM system anatomic staging.

To get an insight how intestinal LCA biosynthesis changes in breast cancer we assessed the abundance of the baiH ORF in human fecal DNA from the experimental cohort described in²⁴. The term "abundance" in general is meant as a proportion of a given specimen in a given pool. The "abundance" of a bacterium species may refrer to the concentration or the number of cells measured in a sample (e.g. a fecal sample). The term "abundance of the baiH ORF" relates to the proportion of the special segment of bacterial DNA making up the baiH ORF in a DNA pool or isolate.

In order to do that we searched for bacterial species where the ORF for baiH was annotated. We identified the baiH ORF of Gram positive and Gram negative species and measured the abundance of the baiH DNA in fecal DNA samples using qPCR assays. The primers used are shown in Table 2.

Table 2 Primers used for the determination of baiH abundance using qPCR

The primers set up to assess baiH abundance were tested on the fecal DNA derived from the CPX study. The abundance of the baiH ORF DNA did not change when fecal DNA was probed with the thetaiotaomicron or Clostridium scindens-specific primers, in line with the CPX -resistance of these species. In contrast to these, the abundance of baiH ORF DNA decreased when fecal DNA was amplified using probes for Escherichia coli, or Staphylococcus haemolyticus that are CPX-sensitive species pointing out the applicability of our approach.

When all patients were compared to healthy controls the abundance of baiH of Clostridium sordellii, Staphylococcus haemolyticus, Escherichia coli and Pseudomonas putida was lower in breast cancer patients (Fig. 2) in line with the lower LCA levels and LCA/CDCA ratio. A more pronounced decrease in the abundance of the baiH of Bacteroides thetaiotaomicron, Clostridium sordellii, Staphylococcus haemolyticus, Escherichia coli and Pseudomonas putida were observed in stage 0 and stage 1 patients.

Taken together, the bacterial LCA biosynthesis machinery in the intestine is downregulated in breast cancer patients that is very pronounced in the early phase of the disease. Lower capacity to synthesize LCA then contributes to lower LCA levels. These findings point towards the involvement of bacterial LCA metabolism in human breast cancer pathogenesis.

EXAMPLES

Chemicals

All chemicals were from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise stated.

Cell culture

MCF7 cells were maintained in MEM (Sigma-Aldrich) medium supplemented with 10 % FBS, 1 % penicillin/streptomycin and 2 mM L-glutamine at 37 °C with 5 % C02.

4T1 cells were maintained in RPMI-1640 (Sigma-Aldrich) medium containing 10 % FBS and 1 % penicillin/streptomycin, 2 mM L-glutamine and 1 % pyruvate at 37 °C with 5 % C02.

Primary fibroblasts cells were maintained in DMEM (Sigma-Aldrich, 1000 mg/L glucose) medium supple- meted with 20 % FBS, 1 % penicillin/streptomycin, 2 mM L-glutamine and 10 mM HEPES at 37 °C with 5 % co₂.

Sulphorhodamine B assay

Cells were seeded in 96- well plate (MCF7 - 5000 cells/well; 4T1 - 2000 cells/well; Fibroblast - 5000 cells/well). Cells were treated with different concentration of LCA, DCA, UDCA for two days. Control cells were cultured in the same media and treated with vehicle (DMSO, dilution factor: 10,000 x). At the end of treatment cells were fixed with 50% trichloroacetic acid (TCA) at 4 °C and stained with sulphorhodamine B (SRB) solution (0.4% in 1 % acetic acid) for 10 minutes. Unbound dye was removed by washing the plate with 1 % acetic acid four times. Bound stain was solubilized with 10 mM TRIS base solution and absorbance was read on plate reader (Thermo Labsystems Multiskan MS) at 540 nm.

Each concentration was tested with eight replicates. The experiment was repeated three times.

Colony formation assay

Five hundred cells were seeded in a 6-well plate in complete medium and were cultured with the indicated concentrations of LCA for 7 days. At the end of the assay plates were washed twice in PBS. Colonies were fixed in methanol for 15 minutes, dried and stained according to May-Grtinwald-Giemsa for 15 minutes. Plate was washed with water and the colonies were counted using Image J software. The experiment was repeated three times.

Detection of cell death

LCA- induced cytotoxicity was determined by propidium iodide (PI) uptake. Cells were seeded in 6-well plate (MCF7 - 200,000 cells/well; 4T1 - 75,000 cells/well) treated with LCA for two days and stained with 100 μg/mL propidium iodide for 30 min at 37 °C, washed once in PBS, and analyzed by flow cytometry (FACSCalibur, BD Biosciences). The experiment was done in triplicate and repeated three times.

Scratch assay and video microscopy

Scratch assays were performed similarly as described in El-Hamoly et al., 2014. Briefly, cells were grown in 6-well plates until cell confluence reached about 70-80 %. The plates were manually scratched with sterile 200 μΐ pipette tip, followed by washing the cells twice with PBS. Then cells were incubated with vehicle or LCA (0.3 μΜ) in a thermostate. Cell densities were monitored every hour for one day using JuLi Br Live cell movie analyzer (NanoEnTek Inc., Seoul, Korea). The experiment was repeated three times.

Electric Cell-substrate Impedance Sensing (ECIS)

ECIS (Electric cell-substrate impedance sensing) model ΖΘ, Applied BioPhysics Inc. (Troy, NY, USA) was used to monitor transendothelial electric resistance of MCF7 and 4T1 cells seeded (MCF7 - 40,000 cells/well; 4T1 - 20,000 cells/well) on type 8W10E arrays. Cells were treated with vehicle or 0.3 μΜ LCA after 20 hours and total impedance values were measured for additional 48 hours. Multifrequency measurements were taken at 62.5, 125, 250, 500, 1000, 2000, 4000, 8000, 16000, 32000, 64000 Hz. Modeling tool of ECIS was used to evaluate the Rb (barrier resistance) values of each of the wells at fix 180 s interval. The reference well was set to a no-cell control with complete medium. Each condition was tested in four replicates.

DNA and mRNA preparation and quantitation; EMT screen

DNA was extracted from fecal samples using PowerSoil DNA Isolation kit (MO BIO Laboratories, Inc. Carlsbad, California) according to the manufacturer's instructions.

Total RNA from cells and tumor samples were prepared using TRIzol reagent (Invitrogen Corporation, Carlsbad, CA).

For the assessment of the expression of individual genes two micrograms of RNA were reverse transcribed using High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA). The qPCR reactions were performed with qPCRBIO SyGreen Lo-ROX Supermix (PCR Biosystems Ltd, London, UK) on Light-Cycler 480 Detection System (Roche Applied Science). Geometric mean of 36B4 and cyclophyllin was used for normalization. Primers are listed in table 3. Table 3 Primers used in the RT-qPCR reactions

CCTGTGCAAGAAGCAGAGTG

Cd4

GCTCTTGTTGGTTGGGAATC

TGCTGTCCTTGATCATCACTCT

Cd8a

ACTAGCGGCCTGGGACAT

CCCCAGACTTTTGCTACTGG

Cdlla

TAGGCCGTGTGTCCAGGT

CAGAAACGCCTAAGCCTAGTTG

Cd45

AGGCAAGTAGGGACACTTCATAG

CGTCCCTCAGTCAAGAGGAG

Pdl

GTCCCTAGAAGTGCCCAACA

For the screen of EMT genes the mouse Epithelial to Mesenchymal Transition (EMT) RT2 ProfilerTM PCR Array (Qiagen Sciences, Maryland, USA) was used according to the manufacturer's instructions.

For the assessement of the abundance of the baiH ORF 10 ng of DNA (from fecal samples) was used for qPCR reactions. Primers are listed above (Table 3). Specificity of the qPCR reactions were verified by sequencing PCR products with the primers used for the amplification.

Metabolomics, pulse-chase metabolomics

Our previously described technique (Jeney et al., 2016) was adapted and established based on other methods (Jaitz et al., 2011 ; Szoboszlai et al., 2014). Briefly, after quenching in liquid nitrogen the labelled (in D5030 medium for 1 hour with 10 mM [U-13C] -glucose or [2- 13C] -acetate - Cambridge Isotope Laboratories, Andover, MA, USA) and unlabelled cells were extracted in methanol-chloroform-HiO solution at 4 °C. The supernatant was separated by centrifugation (15 000 g for 10 min at 4 °C) and stored at -80 °C till further analysis. The derivatization was performed according to Jaitz et al. with some modifications. Drying and sonicating samples in 3-nitrobenzyl alcohol-trimethyl-chlorosilane solution followed 80 °C incubation. The reaction was stopped by adding ammonium bicarbonate. The samples were diluted with acetonitril-water solution and the derivate metab- olites were separated by reversed-phase chromatography in Waters Acquity LC system. For the measurements Waters Micromass Quattro Micro triple quadrupole mass spectrometer (Waters Corporation, Milford MA, USA) was operated with an electrospray source in positive ion mode. Standards (L-lactic acid, L-malic acid, succinic acid and citric acid) and the other chemicals except labelled substrates were purchased from Sigma-Aldrich for these measurements.

Measurement of mitochondrial membrane potential

Mitochondrial membrane potential was determined by 3,3'-dihexyloxacarbocyanine iodide (DioC6(3)) staining. Cells were seeded in 6-well plate (MCF7 - 200,000 cells/well; 4T1 - 75,000 cells/well). After two days treatment cells were stained with 40 nM DioC6 for 30 minutes at 37 °C. Then the cells were washed with PBS and harvested by trypsin/EDTA. Cells were subjected to flow cytometry analysis (FACSCalibur, BD Biosci- ences). The experiment was repeated three times.

Measurement of oxygen consumption and extracellular acidification rate

Oxygen consumption rate (OCR) and changes in pH, extracellular acidification rate (ECAR) were measured using an XF96 oxymeter (Seahorse Biosciences, North Billerica, MA, USA). Cells were seeded in 96-well Seahorse assay plate (MCF7 - 3000 cells/well; 4T1 - 1500 cells/well) and treated with vehicle and LCA for two days. Then oxygen consumption was recorded every 30 minutes to follow the LCA effect. Data were normalized to protein content.

SDS-PAGE and Western blotting

Cells were lysed in RIPA buffer (50 mM Tris, 150 mM NaCl, 0.1 % SDS, 1 % TritonX 100, 0.5 % sodium deoxycolate, 1 mM EDTA, 1 mM Na3V04, 1 mM NaF, 1 mM PMSF, protease inhibitor coctail). Protein extracts were separated on 10% SDS polyacrylamide gels and transferred onto nitrocellulose membranes by electroblot- ting. Membranes were blocked with 5 % BSA, and incubated with primary antibodies for overnight at 4 °C. The membranes were washed with IX TBS-TWEEN and probed with IgG HRP conjugated secondary antibodies (Cell Signaling Technology, Inc. Beverly, MA, 1 :2000). Bands were visualized by enhanced chemiluminescence reaction (SuperSignal West Pico Solutions, Thermo Fisher Scientific Inc, Rockford, USA). Antibodies used in this study are listed in Table 4.

Table 4 Antibodies used

Immunocytochemistry

Cells were grown on coverslips, washed with PBS, fixed with 4 % paraformaldehyde for 15 minutes and permeabilized using 1 % Triton X-100 for 5 minutes. Then cells were blocked with 1 % BSA for one hour and incubated with TexasRed-X Phalloidin (Invitrogen, Oregon, USA) for 45 minutes for the analysis of cellular morphology.

For cellular localization of NRF1 protein cells were incubated overnight with NRF1 primary antibody at 4 °C. After washing steps, cells were incubated with secondary antibody (1 :600, anti -rabbit Alexa 488, Life technologies) for 1 hour at room temperature. Cell nuclei were visualized with TO-PRO-3 iodide (1 : 1000, Life technologies). Coverslips were rinsed and mounted in Mowiol/Dabco solution. Confocal images were acquired with Leica SP8 confocal microscope and LAS AF v3.1.3 software.

Transfection and antagonist treatments

Silencer Select siRNA targeting TGR5 (GPBAR1 - cat.no. 4392420) and Negative control siRNA #1 (cat.no. 4390843) were obtained from Thermo Fisher Scientific. Cells were seeded in 24-well plate (MCF7 - 50,000 cells/well) and on the next day cells were transfected with TGR5 siRNA and negative control at a final concentration of 30 nM using Lipofectamine RNAiMAX transfection reagent (Invitrogen). Cells were incubated with transfection complexes in medium containing LCA (0.3 μΜ) for 48 h.

Cells were treated with U73343 (phospholipase C inhibitor), NF449 (Gsa-selective antagonist), CINPA1 (CAR antagonist), DY268 (FXR antagonist), GSK2033 (LXR antagonist) at a final concentration of 5 μΜ in medium containing LCA (0.3 μΜ) for 48 h. Except for U73343 which was used at a final concentration of 1 μΜ. The inhibitor and antagonists were obtained from Tocris Bioscience.

Animal study

All animal experiments were authorized by the local and national ethical board (reg. 1/2015/DEMAB) and were performed to conform the relevant EU and US guidelines. Experimental animals were female BALB/c mice between 8-10 weeks of age. For all experiments animals were randomized. Animals were bred in the "specific pathogen free" zone of the Animal Facility at the University of Debrecen, and kept in the "minimal disease" zone during the experiment.

4T1 tumor injection

4T1 cells were suspended (2xl0⁶/mL) in ice cold PBS-matrigel (1 : 1, Sigma-Aldrich) at 1 : 1 ratio. From this suspension female BALB/c mice received 50 μL· injections to their 2nd inguinal fat pads on both sides (105 cells/injection). Tumor growth and animal wellbeing was monitored daily.

LCA treatment

Animals received daily oral LCA treatment. LCA stock was prepared in 96% ethanol at lOOx concentration (7.5 mM) for storage at -20°C. LCA stock was diluted each day to a working concentration of 75 μΜ in sterile PBS immediately before the treatment. Ethanol vehicle (1 % in PBS) was prepared and diluted similarly. Animals received a daily oral dose of 200 μΕ/30 g bodyweight from the LCA solution or the vehicle. Researchers administering LCA and vehicle solutions were blinded. Treatment was administered every day at the same time during the morning hours between 8 am and 10am.

Infiltration score

During autopsy tumors were visually assessed and scored based on their infiltration rate into surrounding tissues. If the tumor mass remained in the mammary fat pads without any detectable attachment to muscle tissues then it was classified as a "low infiltration" tumor. In case the tumor mass attached to the muscle tissue below the fat pad but hasn't penetrated it then it was classified as a "medium infiltration" tumor. Finally, if the tumor mass grew into the muscle tissue and penetrated the abdominal wall then it was scored as a "high infiltration" tumor. Researchers involved in scoring primary tumors for their infiltration rate were blinded.

9^9mTc-Mibi uptake assays

Control and tumor-bearing experimental animals were anaesthetized by 3% isoflurane (Forane) with a dedicated small animal anesthesia device and 3.7+0.2 MBq of ^99mTc-MIBI was injected via the lateral tail vein. After the incubation time (60 min) mice were euthanized with 5% Forane and blood samples were taken from the heart. Tissue samples were taken from each organ or tissue (fat, muscle, lung and tumor(s)) and their activities were measured with a calibrated gamma counter (Perkin-Elmer Packard Cobra, Waltham, MA, USA). The weight and the radioactivity of the samples were used to determine the differential absorption ratio (DAR). DAR value was calculated as: [accumulated radio activity/g tissue]/ [total injected radioactivity/body weight].

Ciprofloxacin treatment

Ciprofloxacin was dissolved in PBS (pH 6.8) at 50 mg/mL. Animals received a daily oral dose of 200 mg/kg of ciprofloxacin solution or the corresponding volume of vehicle. Treatment was administered every day at the same time during the morning hours between 8am and 10am.

TIL calculation

Tumor infiltrating lymphocytes (TIL) content of tumors was expressed as the number of TILs per 100 tumor cells.

Human studies

The study in which human feces samples were collected from healthy subjects and breast cancer patients was developed by collaborators at the National Cancer Institute (NCI), Kaiser Permanente Colorado (KPCO), the Institute for Genome Sciences at the University of Maryland School of Medicine, and RTI International. The study protocol and all study materials were approved by the Institutional Review Boards at KPCO, NCI, and RTI International (IRB number 11CN235).

The study in which human serum samples were collected from healthy subjects and breast cancer patients was developed by collaborators at the University of Debrecen (Hungary). The study protocol and all study materials were approved by the Institutional and Hungarian Review Boards (3140-2010).

Cohort for fecal DNA and serum studies are listed in Table 5 and Table 6.

Serum bile acid determination

Bile acids in serum were determined as described in Sakakura, H. et al. : Simultaneous determination of bile acids in rat bile and serum by high-performance liquid chromatography. J. Cromatogr. 621, 123-131 (1993).

Database search

The kmplot.com database was used to study the link between gene expression levels and breast cancer survival in humans. The association of known mutations with breast cancer was retrieved from www.intogen.org/. Gene expression profiles were retrieved from the Gene expression omnibus (www.ncbi.nlm.nih.gov/geoprofiles/). The sequence of the baiH ORF and the bai operon was retrieved from the KEGG (www.genome.jp/kegg/) and the PATRIC databases (www.patricbrc.org/).

Statistical analysis

We used two-tailed Student's t-test for the comparison of two groups unless stated otherwise. Fold data were log2 transformed to achieve normal distribution. For multiple comparisons one-way analysis of variance test (ANOVA) was used followed by Tukey's honestly significance (HSD) post-hoc test. Data is presented as average + SD unless stated otherwise. Statistical analysis was done using GraphPad Prism VI software. Patient cohorts

Table 5 Cohort for serum studies

Patients were recruited at the Medical Center of the University of Debrecen. Patients were age and sex- matched, and the staging was according to AJCC. Patients with other cancers, inflammatory diseases, diseases affecting the GI tract and the liver or receiving antibiotics were excluded from the study.

Table 6 Cohort for fecal DNA studies

This cohort was published in Goedert, J.J., et al.: Investigation of the association between the fecal microbiota and breast cancer in postmenopausal women: a population-based case-control pilot study. J Natl Cancer Inst 107, djvl47 (2015).

Results

LCA, a metabolite of the microbiota, is expressed locally in breast ducts or transferred through the bloodstream to the breast where it may be an important player in bringing about an anti-cancer tumor microenvironment. In our studies LCA inhibited the proliferation of breast cancer cells, while it did not interfere with primary fibroblast cells (Fig. 3A).

Other secondary bile acids, deoxycholic acid (DCA) and ursodeoxycholic acid (UDCA) were investigated in concentrations corresponding to their normal (reference) concentrations in human serum and breast tissue (10 nM - 10 uM) first in short term proliferation assays. LCA reduced cellular proliferation of MCF7 and 4T1 breast cancer cells but did not affect primary fibroblasts (10 nM - 10 uM) (Fig. 3A). Other secondary bile acids such as UDCA or DCA were without effect on MCF7 and 4T1 breast cancer cells (Fig. 4). In the subsequent assays LCA concentrations higher than the reference serum concentration of LCA (-30-50 μΜ) were used as higher LCA concentrations (100-1000 nM) were reported in the breast—. The cytostatic effect of LCA was verified in longer colony forming assays (Fig. 3B). The percent of propidium-iodide positive cells did not change upon LCA treatment suggesting that LCA did not induce cell death (Fig. 3C).

The cytostatic property of LCA was tested in mice that were grafted with 4T1 cells and were treated with LCA (15 nmol LCA p.o. q.d.) or vehicle for 18 days. At the time of the sacrifice neither the weight, nor ^99mTc-Mibi uptake of the primary tumors differed between the vehicle and LCA-treated groups (data not shown). However, the infiltration capacity of the primary tumor to the surrounding tissues markedly decreased upon LCA treatment (Fig. 3D), similarly to the number, mass and ^99mTc-Mibi uptake of the metastases (Fig. 3E-G).

LCA interferes with multiple anticancer molecular pathways

After finding the cytostatic property of LCA in breast cancer, we investigated how LCA modulates the different features of breast cancer through assessing classical hallmarks of cancer¹⁹.

LCA inhibited tumor infiltration and metastasis formation (Fig. 3B, D-G), implicating modulation of the epi- thelial-mesenchymal transition (EMT). In order to assess EMT, we performed an RT-qPCR-based screening of EMT marker genes in 4T1 cells and in tumors derived from the in vivo model. This identified genes involved in wnt-signaling (Ctnnbl, Wnt5b, Fzd7, Tcf712, Vegf) (Fig. 5 A) implicating ligands (Wnt5b, Vegf), and downstream signaling molecules (Ctnnbl, Fzd7) suggesting decreased wnt-signaling. Indeed, LCA-treatment inhibited β- catenin signaling as evidenced by lower GSK-3a and GSK-3 phosphorylation and lower β-catenin protein content both in cell lines and in vivo (Fig. 5B). LCA treatment improved cell-to-cell connections, an epithelial feature, as reflected by cobblestone-like morphology in cells (Fig. 5C) and improved barrier function (Rb, Fig. 5D) and total impedance (Fig. 6 A) that provide functional evidence of better cell-to-surface and cell-to-cell adhesion. Furthermore, LCA-treated 4T1 cells were slower in moving into a void area in a scratch assay as compared to vehicle- treated ones (Fig. 6B).

Histological analysis of the primary tumors revealed that LCA treatment increased the number of the tumor infiltrating lymphocytes (TILs) (Fig. 5G). We assessed TIL markers in the primary tumors and found that the expression of CD8a, CD3, CD4, CD1 la, CD45 and PD1 were induced in the LCA-treated mice (Fig. 5G).

Breast cancer depends on Warburg metabolism ^20,21. Therefore we assessed LCA-induced changes in cellular metabolism. LCA treatment induced glycolysis in 4T1 and MCF7 cells as evidenced by increases in extracellular acidification rate (ECAR) and intracellular lactate levels (Fig. 8A, B). Furthermore, we observed the induction of mitochondrial oxidative metabolism upon LCA treatment marked by increased oxygen consumption rate (OCR) (Fig. 8C), mitochondrial membrane potential (Δψ) (Fig. 7A), levels of intracellular citrate (Fig. 8D) and succinate (Fig. 7B), and expression of a set of mitochondrial genes in 4T1 and MCF7 cells (Fig. 7C). The citrate/lactate ratio in 4T1 cells and citrate/lactate ratio in MCF7 cells (Fig. 8E) pointed towards a trend for the dominance of mitochondrial oxidative function over glycolysis.

Pulse-chase metabolomics experiments were performed in MCF7 and 4T1 cells treated with 300 nM LCA.

When cells were charged with ¹³C-acetate, a metabolite that can directly enter the TCA cycle, LCA treatment enhanced the incorporation of ¹³C into succinate and malate (Fig. 8F). Next, we fed cells ¹³C-glucose from which radioactive carbons must enter glycolysis to subsequently feed the TCA cycle or to form lactate. LCA treatment in both cell lines enhanced the amount of ¹³C-labelled citrate and lactate. Moreover, the ratio between ¹³C-citrate and ¹³C-lactate increased, providing further evidence towards mitochondrial dominance of the LCA-induced hy- permetabolic switch of breast cancer cells (Fig. 8G). The administration of LCA did not change 2-hydroxyglu- tarate levels in cells (data not shown). Taken together, LCA treatment made breast cancer cells hypermetabolic and reversed certain features of Warburg metabolism. We assessed components of the cellular energy sensor web and mitochondrial transcriptional regulators to find the roots of the above metabolic changes. LCA induced expression of positive regulators of mitochondrial oxidative phosphorylation. One such factor was nuclear respiratory factor-1 (NRF1) (Fig. 8H). LCA not only boosted NRF1 expression but also enhanced its nuclear translocation (Fig. 7D). In addition, other regulators, such as AMPK was activated (marked by phosphorylation of ACC) and expression of PGC-la, PGC-Ιβ and FOXOl were also induced by LCA (Fig. 7C, 7E). In the in vivo experiments we also observed the LCA-mediated induction of AMPK activity (marked by increased phospho-ACC and phospho-AMPK levels) and enhanced expression of FOXOl as well, although neither NRF1, nor PGC-Ιβ expression was induced by LCA.

Subsequently, we assessed how these metabolic regulators affect the behavior of breast cancer in humans. In a previous study we found that high AMPK and FOXOl expression are important negative regulators of breast cancer in humans²⁰. Using the kmplot.com database we found that while high NRF1 expression is also a negative regulator of breast cancer, higher PGC-la expression does not improve survival (Fig. 7E). Thus, PGC-la expression was not analyzed further. To evaluate the involvement of PGC-Ιβ we assessed the GEO and the Intogen databases. Although no frequent (driver) mutations were found in PGC-Ιβ, it did appear that the expression of PGC- 1 β was associated with breast cancer, including reduced expression in tumors as compared to healthy tissues and in metastases as compared to the primary tumors. Taken together, the modulation of AMPK, FOXOl, PGC- 1β or NRF1 may have (patho)physiological relevance in modulating LCA -evoked effects in humans.

Reijnders, D., et al. Effects of Gut Microbiota Manipulation by Antibiotics on Host Metabolism in Obese Humans: A Randomized Double-Blind Placebo-Controlled Trial. Cell Metab. 24, 63-74. (2016).

Smits, L.P., Bouter, K.E., de Vos, W.M., Borody, T.J. & Nieuwdorp, M. Therapeutic potential of fecal micro- biota transplantation. Gastroenterology 145, 946-953 (2013).

Krishnan, S., Alden, N. & Lee, K. Pathways and functions of gut microbiota metabolism impacting host physiology. Curr Opin Biotechnol. 36, 137-145 (2015).

Garrett, W.S. Cancer and the microbiota. Science 348, 80-86 (2015).

de Martel, C, et al. Global burden of cancers attributable to infections in 2008: a review and synthetic analysis. Lancet Oncol. 13, 607-615. (2012).

Yu, H., et al. Urinary microbiota in patients with prostate cancer and benign prostatic hyperplasia. Archives of medical science : AMS 11, 385-394 (2015).

Chase, D., Goulder, A., Zenhausern, F., Monk, B. & Herbst-Kralovetz, M. The vaginal and gastrointestinal microbiomes in gynecologic cancers: a review of applications in etiology, symptoms and treatment. Gynecol Oncol 138, 190-200 (2015).

Yu, Y., Champer, J., Beynet, D., Kim, J. & Friedman, A.J. The role of the cutaneous microbiome in skin cancer: lessons learned from the gut. Journal of drugs in dermatology : JDD 14, 461-465 (2015).

Gui, Q.F., Lu, H.F., Zhang, C.X., Xu, Z.R. & Yang, Y.H. Well-balanced commensal microbiota contributes to anti -cancer response in a lung cancer mouse model. Genetics and molecular research : GMR 14, 5642-5651 (2015).

10. Yamamoto, M.L. & Schiestl, R.H. Lymphoma caused by intestinal microbiota. International journal of environmental research and public health 11, 9038-9049 (2014).

11. Yamamoto, M.L. & Schiestl, R.H. Intestinal microbiome and lymphoma development. Cancer J. 20, 190-194.

(2014).

12. Dapito, D.H., et al. Promotion of hepatocellular carcinoma by the intestinal microbiota and TLR4. Cancer Cell 21, 504-516 (2012).

13. Yoshimoto, S., et al. Obesity-induced gut microbial metabolite promotes liver cancer through senescence se- cretome. Nature. 499, 97-101. (2013).

14. Bindels, L.B., et al. Gut microbiota-derived propionate reduces cancer cell proliferation in the liver. Br J Cancer 107, 1337-1344 (2012).

15. Ridlon, J.M., Kang, D.J. & Hylemon, P.B. Bile salt biotransformations by human intestinal bacteria. J Lipid Res 47, 241-259 (2006).

16. Wikoff, W.R., et al. Metabolomics analysis reveals large effects of gut microflora on mammalian blood me- tabolites. Proc Natl Acad Sci U S A 106, 3698-3703 (2009).

17. Smith, J.L., et al. Endogenous ursodeoxycholic acid and cholic acid in liver disease due to cystic fibrosis.

Hepatology 39, 1673-1682 (2004).

18. Raju, U., Levitz, M. & Javitt, N.B. Bile acids in human breast cyst fluid: the identification of lithocholic acid.

J Clin Endocrinol Metab. 70, 1030-1034. (1990).

19. Hanahan, D. & Weinberg, R.A. Hallmarks of cancer: the next generation. Cell 144, 646-674 (2011).

20. Fodor, T., et al. Combined Treatment of MCF-7 Cells with AICAR and Methotrexate, Arrests Cell Cycle and Reverses Warburg Metabolism through AMP -Activated Protein Kinase (AMPK) and FOXOl . PLoS One. 11, e0150232 (2016).

21. Martinez -Outschoorn, U.E., Peiris-Pages, M., Pestell, R.G., Sotgia, F. & Lisanti, M.P. Cancer metabolism: a therapeutic perspective. Nature reviews. Clinical oncology, doi: 10.1038/nrclinonc.2016.1060. (2016).

22. Toda, T., et al. Ciprofloxacin suppresses Cyp3a in mouse liver by reducing lithocholic acid-producing intestinal flora. Drug metabolism and pharmacokinetics 24, 201-208 (2009).

23. Tang, X., et al. A joint analysis of metabolomics and genetics of breast cancer. Breast Cancer Res 16, 415 (2014).

24. Goedert, J.J., et al. Investigation of the association between the fecal microbiota and breast cancer in postmenopausal women: a population-based case-control pilot study. J Natl Cancer Inst 107, djvl47 (2015).

25. Goldberg, A. A., et al. Lithocholic bile acid selectively kills neuroblastoma cells, while sparing normal neuronal cells. Oncotarget. 2, 761-782. (2011).

26. Goldberg, A.A., Titorenko, V.I., Beach, A. & Sanderson, J.T. Bile acids induce apoptosis selectively in andro- gen-dependent and -independent prostate cancer cells. PeerJ. 1, el22 (2013).

27. Plottel, C.S. & Blaser, M.J. Microbiome and malignancy. Cell Host Microbe 10, 324-335 (2011).

28. Xuan, C, et al. Microbial dysbiosis is associated with human breast cancer. PLoS One. 9, e83744. doi:

83710.81371/journal.pone.0083744. eCollection 0082014. (2014). 29. Hieken, T.J., et al. The Microbiome of Aseptically Collected Human Breast Tissue in Benign and Malignant Disease. Sci Rep. 6:30751., 10.1038/srep30751. (2016).

30. Chan, A.A., et al. Characterization of the microbiome of nipple aspirate fluid of breast cancer survivors. Sci Rep. 6:28061., 10.1038/srep28061. (2016).

31. Urbaniak, C, et al. The Microbiota of Breast Tissue and Its Association with Breast Cancer. Appl Environ Microbiol. 82, 5039-5048. doi: 5010.1128/AEM.01235-01216. Print 02016 Aug 01215. (2016).

32. Velicer, CM., et al. Antibiotic use in relation to the risk of breast cancer. JAMA. 291, 827-835. (2004).

33. Velicer, CM., Heckbert, S.R., Rutter, C, Lampe, J.W. & Malone, K. Association between antibiotic use prior to breast cancer diagnosis and breast tumour characteristics (United States). Cancer Causes Control. 17, 307- 313. (2006).

34. Friedman, G.D., et al. Antibiotics and risk of breast cancer: up to 9 years of follow-up of 2.1 million women.

Cancer Epidemiol Biomarkers Prev. 15, 2102-2106. (2006).

35. Wirtz, H.S., et al. Frequent antibiotic use and second breast cancer events. Cancer Epidemiol Biomarkers Prev.

22, 1588-1599. doi: 1510.1158/1055-9965.EPI-1513-0454. Epub 2013 Jul 1585. (2013).

36. Tamim, H.M., Hanley, J. A., Hajeer, A.H., Boivin, J.F. & Collet, J. P. Risk of breast cancer in relation to antibiotic use. Pharmacoepidemiol Drug Saf. 17, 144-150. (2008).

37. Satram-Hoang, S., et al. A pilot study of male breast cancer in the Veterans Affairs healthcare system. J Environ Pathol Toxicol Oncol 29, 235-244. (2010).

38. Rossini, A., et al. Influence of antibiotic treatment on breast carcinoma development in proto-neu transgenic mice. Cancer Res 66, 6219-6224 (2006).

39. Batta et al. (2016) A litokolsav hatasa a 4T1 tumorsejtek attetkepzo kepessegere egermodellben. http://hdl.han- dle.net/2437/227200

Claims

1. A method for diagnosing early stage breast cancer in a human subject, comprising

assessment of the abundance of at least bacterium species which is capable of producing lithocholic acid (LCA) under physiological conditions in the human body, in a test sample derived from the subject, wherein the test sample comprises human microbiota of said subject, and

wherein a decrease in the abundance of the at least one species in the test sample compared to a reference value of abundance of the at least one bacterium species typical of the absence of early stage breast cancer, is indicative of early stage breast cancer in said subject.

2. The method according to claim 1, further comprising

assessment of the level of lithocholic acid in a test sample derived from said subject, wherein a decrease in the level of LCA in the test sample compared to a reference level of LCA typical of the absence of early stage breast cancer is indicative of early stage breast cancer in said subject, and optionally

assessment of the molar ratio of LCA to an LCA precursor bile acid in a test sample of said subject, wherein a decrease in said ratio in the test sample compared to a reference ratio of LCA to the LCA precursor bile acid typical of the absence of early stage breast cancer, is indicative of early stage breast cancer in said subject.

3. The method according to claim 1 or 2, further comprising

assessment of the level of any one of the compounds selected from the group consisting of: total bile acids, cholic acid (CA), chenodeoxycholic acid (CDCA), ursodeoxycholic acid (UDCA), deoxycholic acid (DCA) and combinations thereof, in a test sample of said subject,

wherein a decrease in said level in the test sample compared to a reference level typical of the absence of breast cancer, is indicative of breast cancer in said human subject.

4. The method according to claim 3, wherein the reference level of the compound is typical of the absence of early stage breast cancer.

5. The method according to any one of the preceding claims, wherein the at least one bacterum species is selected from the genera Clostridium, Eubacterium, Bacteroides, Escherichia, Peptostreptococcus, Pseudomonas, Ruminococcus, Staphylococcus, Dorea and/or the family Lachnospiraceae.

6. The method according to claim 5, wherein the at least one bacterium is selected from the species Staphylococcus haemolyticus, Escherichia coli, Bacteroides thetaiotamicron, Clostridium sordellii and Pseudomonas putida.

7. The method according to any one of the preceding claims, wherein the abundance of the bacterium species to be assessed is detected by detecting the abundance of the DNA sequence or gene product of any one of the genes or open reading frames (ORFs) of the bile acid inducible (bai) operon of the bacterium species.

8. The method according to claim 6, wherein the gene or ORF is the baiH.

9. The method according to any one of the preceding claims, wherein the method is for diagnosing breast cancer stage 0 according to the American Joint Committee on Cancer (AJCC) TNM system and the test sample is compared to a reference value typical of the absence of stage 0 breast cancer.

10. The method according to any one of claims 1 to 8, wherein the method is for diagnosing breast cancer stage 1 according to the AJCC TNM system and the test sample is compared to a reference value typical of the absence of stage 1 breast cancer.

11. A method for diagnosing early stage breast cancer in a human subject, comprising

assessment of the level of lithocholic acid in a test sample derived from said subject wherein a decrease in the level of LCA in the test sample compared to a reference level of LCA typical of the absence of early stage breast cancer, is indicative of early stage breast cancer in said subject.

12. The method according to claim 11, further comprising

assessment of the molar ratio of LCA to LCA precursor bile acid in a test sample of said subject, wherein a decrease in said ratio in the test sample compared to a reference ratio of LCA to the LCA precursor bile acid typical of the absence of early stage breast cancer, is indicative of early stage breast cancer in said subject.

13. The method according to claim 11 or 12, wherein the method is for diagnosing breast cancer stage 1 according to the American Joint Committee on Cancer (AJCC) TNM system and the test sample is compared to a reference value typical of the absence of stage 1 breast cancer.

14. A method for diagnosing breast cancer in a human subject, comprising

assessing the level of any one of the compounds selected from the group consisting of: total bile acids, cholic acid (CA), chenodeoxycholic acid (CDCA), ursodeoxycholic acid (UDCA), deoxycholic acid (DCA), lithocholic acid (LCA) and combinations thereof, in a test sample derived from said subject,

wherein a decrease in said level as compared to a reference level indicating the absence of breast cancer is indicative of the presence of breast cancer.

15. The method according to claim 14, further comprising

a) assessment of the abundance of at least bacterium species which is capable of producing LCA under physiological conditions in the human body, in a test sample derived from the subject, wherein the test sample comprises human microbiota of said subject, wherein a decrease in the abundance of the at least one species in the test sample compared to a reference value of abundance of the at least one bacterium species typical of the absence of breast cancer, is indicative of breast cancer in said subject,

b) assessment of the level of lithocholic acid in a test sample derived from said subject wherein a decrease in the level of LCA in the test sample compared to a reference level of LCA typical of the absence of breast cancer, is indicative of breast cancer in said subject,

c) assessment of the molar ratio of LCA to LCA precursor bile acid in a test sample of said subject, wherein a decrease in said ratio in the test sample compared to a reference ratio of LCA to the LCA precursor bile acid typical of the absence of breast cancer, is indicative of breast cancer in said subject or

d) any combination of a) - c).

16. The method according to claim 15, wherein in any one of steps a) - c),

the reference value is typical of the absence of early stage breast cancer.