The present invention concerns a method for selecting patients suffering from colorectal cancer for anti-PD1 therapy.

Background

Colorectal cancer (CRC) is the third most common malignancy and is the fourth leading cause of cancer-related deaths in men and the third leading cause among women worldwide. It presents a high therapeutic need. Indeed, for most patients with metastatic CRC, chemotherapy is the only viable possibility. However, anti-PD1 immunotherapy was recently approved by the Food and Drug Administration as a second-line treatment of a specific subgroup of patients.

Anti-checkpoint antibodies (such as the anti-PD1 antibody nivolumab and the anti- CTLA4 antibody ipilimumab) are a new class of antibodies that are very promising molecules in cancer treatment of various malignancies such as multiple melanoma, lung, bladder, lymphoma and CRC. However, only a minority of patients is responding well to the treatment. Indeed, the results of the major clinical trials reveal that around 40-60% of patients do not respond satisfactorily to these therapies.

Accordingly, it is crucial to have biomarkers which can efficiently predict the response of a patient suffering from CRC to anti-PD1 therapy.

Some biomarkers have been proposed for predicting efficacy of anti-PD1 therapy. For example, PD-L1 overexpression is an important and widely-explored predictive biomarker for the response to anti-PD1 therapies. However, PD-L1 staining cannot be used to accurately select patients for anti-PD1 therapies due to the low prediction accuracy and dynamic changes of the expression. Tumor-infiltrating immune cells and molecules in the tumor microenvironment, or along with PD-L1 expression, may be important in predicting clinical benefits of anti-PD1 therapies. To the inventors' knowledge however, no simple technology is available so far for selecting CRC patients responding to anti-PD1 treatment, in particular sporadic MSS (Microsatellite Stable) CRC patients. There is thus still an important need of biomarkers for predicting the response of a CRC patient to an anti-PD1 therapy.

The present invention meets this need. Host factors associated with environmental and nutritional factors play an important role in the colorectal carcinogenesis. In the recent two decades, growing attention has been given to the role of intestinal microbiota in CRC-carcinogenesis. Studies showed a reduction of tumorigenesis in the intestines of animals raised in a germ-free environment, in both genetically induced and colitis-associated CRC mouse models. Moreover, colorectal cancer patients often exhibit a distinct microbiota compared to healthy population. However, mechanisms by which intestinal microbiota dysbiosis has been shown to promote colorectal carcinogenesis are not yet elucidated. Recently, Wong et al. (2017) Gastroenterology 153:1621 -1633 showed that gavage with stools from colorectal cancer patients promote intestinal carcinogenesis in germ-free and azoxymethane-(AOM)- carcinogen mice models. Despite the large number of studies, CRC associated dysbiosis is partly defined. However, some bacterial strains, such as Bacteroides fragilis, Fusobacterium nucleatum and Escherichia coli, have a well-established role in CRC development.

Even though E. coli is a commensal bacterium, several strains have acquired some virulence factors including genotoxins, such as colibactin. The colibactin toxin, a hybride polyketide non-ribosomal peptide is encoded by the polyketide synthase {pks) pathogenicity island (Johnson et al. (2008) J. Clin. Microbiol. 46:3906-391 1 ). These E.coli bearing pks island were preferentially detected in CRC samples in comparison to non- neoplasic control (Arthur et al. (2012) Science 338:120-123, Buc et al. (2013) PLoS ONE 8:e56964). The mechanisms by which CRC-associated E. coli promote colorectal carcinogenesis are diverse and somewhat specific to the animal models and the microbial status of the animals (germ-free or SPF). However, modulation of immune response and inflammation seems to play a central role in these mechanisms.

Interestingly, metagenomics whole genome sequencing (WGS) of gut microbiome in patients suffering from melanoma and treated with anti-PD1 therapy showed an enrichment notably in Escherichia coli in non-responders (Gopalakrishnan et al. (2018) Science 359:97-103).

Summary

The present invention arises from the unexpected finding by the inventors that the presence of E. coli bearing pks island in mice subcutaneously implanted with MC38 tumor prevented the activity of an anti-PD1 antibody, whereas said anti-PD1 antibody was efficient in mice implanted with MC38 tumor but having a natural microbiota. Accordingly, the presence of E.coli harboring the pks island in patient stools or colonic biopsies can be predictive of patient resistance to anti-PD1 therapy.

The present invention thus concerns a method of predicting resistance to anti-PD1 therapy in a subject suffering from cancer, said method comprising the steps of:

a) determining, from a biological sample from the subject, in particular a feces or a colonic biopsy sample from the subject, the presence of a pks island,

b) predicting from the result of step a) that the subject is likely to be resistant to anti- PD1 therapy.

In a particular embodiment, the presence of a pks island is determined in step a) by determining the presence of Escherichia coli bacteria positive for pks island.

In a particular embodiment, said cancer is colorectal cancer.

In a particular embodiment, the presence of a pks island is determined in step a) by determining the presence of a pks island DNA.

In another particular embodiment, the presence of a pks island DNA is determined by PCR.

In a particular embodiment wherein the presence of a pks island is determined in step a) by determining the presence of Escherichia coli bacteria positive for a pks island, the presence of Escherichia coli bacteria positive for a pks island is determined by isolation of E. coli bacteria on agar and determination of the presence of a pks island DNA, in particular by PCR.

In another particular embodiment wherein the presence of a pks island is determined in step a) by determining the presence of Escherichia coli bacteria positive for a pks island, the presence of Escherichia coli bacteria positive for a pks island is determined by detecting the presence of an E. coli- specific DNA, in particular by PCR, and determining the presence of a pks island DNA, in particular by PCR.

In a particular embodiment, the determination of the presence of a pks island DNA by PCR is carried out by determining the presence, by PCR, of at least one, in particular two, genes of the pks island.

In still another embodiment, the anti-PD-1 therapy is a therapy with cemiplimab.

Another object of the invention is a method for selecting therapy to treat a subject suffering from cancer, said method comprising the steps of:

a) determining, from a biological sample from the subject, in particular a feces sample or colonic biopsy from the subject, the presence of a pks island, b) when it is determined in step a) that said pks island is present, selecting for said subject an anti-cancer therapy which is not an anti-PD-1 therapy.

In a particular embodiment, the presence of a pks island is determined in step a) by determining the presence of Escherichia coli bacteria positive for pks island.

In a particular embodiment, said cancer is colorectal cancer.

In a particular embodiment, the anti-cancer therapy which is not an anti-PD-1 therapy is selected from the group consisting of irinotecan, oxaliplatin, combination of irinotecan with 5-FU and leucovorin, combination of oxaliplatin with 5-FU and leucovorin, combination of irinotecan with capecitabine, combination of oxaliplatin and capecitabine, anti-EGFR antibodies such as cetuximab and panitumumab, anti-VEGFR antibodies such as bevacizumab, anti-VEGFR2 antibodies such as ramucirumab, aflibercept, zivaflibercept, regorafenib, combination of trametinib and palbociclib, reolysin ^® , combination of dabrafenib, trametiniv and panitumumab, and combinations thereof.

The present invention also concerns an anti-cancer therapy which is not an anti-PD- 1 therapy for use in a method for treating a subject suffering from cancer, said method comprising:

a) determining, from a biological sample from the subject, in particular a feces or colonic biopsy sample, from the subject, the presence of a pks island,

b) when it is determined in step a) that said pks island is present, administering to said subject a therapeutically effective amount of said anti-cancer therapy which is not an anti-PD-1 therapy.

In a particular embodiment, the presence of a pks island is determined in step a) by determining the presence of Escherichia coli bacteria positive for pks island.

In a particular embodiment, said cancer is colorectal cancer.

The present invention also concerns an anti-PD1 and/or anti-PDL1 antibody for use in a method for treating a subject suffering from cancer, said method comprising:

a) determining, from a biological sample from the subject, in particular a feces or colonic biopsy sample from the subject, the presence of a pks island,

b) when it is determined in step a) that said pks island is absent, administering to said subject a therapeutically effective amount of an anti-PD-1 and/or anti-PDL1 antibody.

In a particular embodiment, the presence of a pks island is determined in step a) by determining the presence of Escherichia coli bacteria positive for pks island.

In a particular embodiment, said cancer is colorectal cancer. In a particular embodiment, said anti-PD1 or anti-PDL1 antibody is cemiplimab.

The present invention further concerns a method for treating a subject suffering from cancer, said method comprising the steps of:

a) determining, from a biological sample from the subject, in particular a feces or colonic biopsy sample from the subject, the presence of a pks island,

b) when it is determined in step a) that said pks island is present, administering to said subject a therapeutically effective amount of an anti-cancer therapy which is not an anti-PD-1 therapy; or

when it is determined in step a) that said pks island is absent, administering to said subject a therapeutically effective amount of an anti-PD-1 therapy.

In a particular embodiment, the presence of a pks island is determined in step a) by determining the presence of Escherichia coli bacteria positive for pks island.

In a particular embodiment, said cancer is colorectal cancer.

In another particular embodiment, said anti-cancer therapy which is not an anti-PD-1 therapy is selected from the group consisting of irinotecan, oxaliplatin, combination of irinotecan with 5-FU and leucovorin, combination of oxaliplatin with 5-FU and leucovorin, combination of irinotecan with capecitabine, combination of oxaliplatin and capecitabine, anti-EGFR antibodies such as cetuximab and panitumumab, anti-VEGFR antibodies such as bevacizumab, anti-VEGFR2 antibodies such as ramucirumab, aflibercept, zivaflibercept, regorafenib, combination of trametinib and palbociclib, reolysin ^® , combination of dabrafenib, trametiniv and panitumumab, and combinations thereof.

In another particular embodiment, said anti-PD1 therapy is a therapy with cemiplimab.

The present invention also relates to a method of predicting resistance to anti-PD-1 therapy in a subject suffering from cancer, said method comprising the steps of:

a) optionally isolating, on agar, E. coli bacteria from biological sample from the subject, in particular a feces or colonic biopsy sample from the subject or performing PCR on said biological sample using primers specific for E. coli bacteria, to detect whether E. coli bacteria are present,

b) performing PCR on said biological sample from the subject using primers specific for a pks gene island, in particular specific for a gene selected from the group consisting of the ClbH, ClbJ, ClbN, ClbC, Clbl, ClbO, ClbB and ClbK genes, to detect whether a pks island is present, and

c) predicting from the result of step b) that the subject is resistant to anti-PD-1 therapy if a pks island is present.

In a particular embodiment, the method comprises step a) of isolating, on agar, E. coli bacteria from biological sample from the subject, in particular a feces or colonic biopsy sample from the subject or performing PCR on said biological sample using primers specific for E. coli bacteria, to detect whether E. coli bacteria are present, and step c) consists in predicting from the results of steps a) and b) that the subject is resistant to anti-PD-1 therapy if an Escherichia coli bacteria positive for pks island is present.

According to a particular embodiment, the primers specific for a gene selected from the group consisting of the ClbH, ClbJ, ClbN, ClbC, Clbl, ClbO, ClbB and ClbK genes are specific for a ClbN gene of the pks island.

According to another particular embodiment, the PCR is performed using one or both of a primer comprising SEQ ID NO: 1 and a primer comprising SEQ ID NO: 2.

In an alternative embodiment, the method of predicting resistance to anti-PD-1 therapy comprises the steps of:

a) sequencing the microbiota from a biological sample, in particular a feces or colonic biopsy sample from the subject,

b) determining, from the sequencing performed in step a), the presence of a pks island, in particular of an Escherichia coli bacteria positive for pks island, and

c) predicting from the result of step b) that the subject is resistant to anti-PD-1 therapy if a pks island, in particular an Escherichia coli bacteria positive for pks island is present.

According to another embodiment, said cancer is colorectal cancer.

According to another embodiment, the anti-PD-1 therapy is a therapy with cemiplimab.

The present invention further concerns a method for selecting therapy to treat a subject suffering from cancer, said method comprising the steps of:

a) optionally isolating, on agar, E. coli bacteria from biological sample from the subject, in particular a feces or colonic biopsy sample from the subject or performing PCR on said biological sample using primers specific for E. coli bacteria, to detect whether E. coli bacteria are present,

b) performing PCR on said biological sample from the subject using primers specific for a pks island gene in particular specific for a gene selected from the group consisting of the ClbH, ClbJ, ClbN, ClbC, Clbl, ClbO, ClbB and ClbK genes to detect whether a pks island is present, and c) if it is determined in step b) that said pks island is present, selecting for said subject an anti-cancer therapy which is not an anti-PD-1 therapy.

In a particular embodiment, the method comprises step a) of isolating, on agar, E. coli bacteria from biological sample from the subject, in particular a feces or colonic biopsy sample from the subject or performing PCR on said biological sample using primers specific for E. coli bacteria, to detect whether E. coli bacteria are present, and step c) consists in selecting for said subject an anti-cancer therapy which is not an anti-PD-1 therapy if it is determined from steps a) and b) that an Escherichia coli bacteria positive for pks island is present.

According to a particular embodiment, the primers specific for a gene selected from the group consisting of the ClbH, ClbJ, ClbN, ClbC, Clbl, ClbO, ClbB and ClbK genes are specific for a ClbN gene of the pks island.

According to another particular embodiment, the PCR is performed using one or both of a primer comprising SEQ ID NO: 1 and a primer comprising SEQ ID NO: 2.

In an alternative embodiment, the method of predicting resistance to anti-PD-1 therapy comprises the steps of:

a) sequencing the microbiota from a biological sample, in particular a feces or colonic biopsy sample from the subject,

b) determining, from the sequencing performed in step a), the presence of a pks island, in particular of an Escherichia coli bacteria positive for pks island, and

c) if it is determined in step b) that a pks island, in particular an Escherichia coli bacteria positive for pks island is present, selecting for said subject an anti-cancer therapy which is not an anti-PD-1 therapy.

According to another embodiment, said cancer is colorectal cancer.

According to another embodiment, the anti-cancer therapy which is not an anti-PD-1 therapy is selected from the group consisting of irinotecan, oxaliplatin, combination of irinotecan with 5-FU and leucovorin, combination of oxaliplatin with 5-FU and leucovorin, combination of irinotecan with capecitabine, combination of oxaliplatin and capecitabine, anti-EGFR antibodies such as cetuximab and panitumumab, anti-VEGFR antibodies such as bevacizumab, anti-VEGFR2 antibodies such as ramucirumab, aflibercept, zivaflibercept, regorafenib, combination of trametinib and palbociclib, reolysin ^® , combination of dabrafenib, trametiniv and panitumumab, and combinations thereof.

The present invention further concerns a method for treating a subject suffering from cancer, said method comprising:

a) determining, from a biological sample from the subject, in particular a feces or colonic biopsy sample from the subject, the presence of a pks island, b) if it is determined in step a) that said pks island is absent, administering to said subject a therapeutically effective amount of an anti-PD-1 and/or anti-PDL1 antibody.

In a particular embodiment, the presence of a pks island is determined in step a) by determining the presence of Escherichia coli bacteria positive for pks island.

According to a particular embodiment, said cancer is colorectal cancer.

According to another particular embodiment, said anti-PD1 and/or anti-PDL1 antibody is cemiplimab.

According to another embodiment, the method further comprises administering an additional anti-cancer therapy.

According to another embodiment, the additional anti-cancer therapy is selected from the group consisting of immune checkpoint inhibitors, radiation therapy, surgery, small molecule kinase inhibitors, chemotherapeutic agents including platinum-based chemotherapeutic agents, nucleic acid synthesis inhibitors, cancer vaccines, anti-CD38 antibodies, anti-MUC16 x CD3 bispecific antibodies, anti-CD20 x CD3 bispecific antibodies, granulocyte-macrophage colony-stimulating factor (GM-CSF), qhίί-TΰRb antibodies, an indoleamine-2, 3-dioxygenase (IDO) inhibitor, an IL-6R inhibitor, an IL-4R inhibitor, an IL-10 inhibitor, a cytokine such as IL-2, IL-7, IL-21 , and IL-15, an anti inflammatory drug, and combinations thereof.

According to another embodiment, the additional anti-cancer therapy is radiation therapy which is selected from the group consisting in hypofractionated radiation therapy and stereotactic body radiation therapy.

According to another embodiment, the additional anti-cancer therapy is small molecule kinase inhibitors selected from the group consisting in sorafenib and ceritinib.

According to another embodiment, the additional anti-cancer therapy is chemotherapeutic agents selected from the group consisting in paclitaxel, pemetrexed and gemcitabine.

According to another embodiment, the additional anti-cancer therapy is platinum- based chemotherapeutic agents selected from the group consisting in carboplatin and cisplatin.

According to another embodiment, the additional anti-cancer therapy is a nucleic acid synthesis inhibitor which is decitabine.

According to another embodiment, the additional anti-cancer therapy is a cancer vaccine which is ISA101 b.

According to another embodiment, the additional anti-cancer therapy is an anti- TGFp antibody which is SAR439459. According to another embodiment, the additional anti-cancer therapy is an anti inflammatory drug selected from the group consisting in corticosteroids and non-steroidal anti-inflammatory drugs.

The present invention further concerns a method for treating a subject suffering from cancer, said method comprising the steps of:

a) determining, from a biological sample from the subject, in particular a feces or colonic biopsy sample from the subject, the presence of a pks island,

b) when it is determined in step a) that said pks island is present, administering to said subject a therapeutically effective amount of an anti-cancer therapy which is not an anti-PD-1 therapy; or

when it is determined in step a) that said pks island is absent, administering to said subject a therapeutically effective amount of an anti-PD-1 therapy.

In a particular embodiment, the presence of a pks island is determined in step a) by determining the presence of Escherichia coli bacteria positive for pks island.

According to a particular embodiment, said cancer is colorectal cancer.

According to another particular embodiment, the anti-cancer therapy which is not an anti-PD-1 therapy is selected from the group consisting of irinotecan, oxaliplatin, combination of irinotecan with 5-FU and leucovorin, combination of oxaliplatin with 5-FU and leucovorin, combination of irinotecan with capecitabine, combination of oxaliplatin and capecitabine, anti-EGFR antibodies such as cetuximab and panitumumab, anti-VEGFR antibodies such as bevacizumab, anti-VEGFR2 antibodies such as ramucirumab, aflibercept, zivaflibercept, regorafenib, combination of trametinib and palbociclib, reolysin ^® , combination of dabrafenib, trametiniv and panitumumab, and combinations thereof.

According to another particular embodiment, said anti-PD1 therapy is a therapy with cemiplimab.

Brief description of the figures

Figures 1-4. Lack of anti-tumoral effect of PD-1 based immunotherapy on growth rate of MC38 murine colonic tumors in pks ⁺ E. coli pre-infected mice.

Figure 1 : Level of 1 1 G5-bacterial colonization (in CFU/g of feces) in feces of mice injected with isotype control (1 1 G5 + isotype control, circles) or with RMP1 -14 antibodies (1 1 G5 + Ab, squares). Figure 2: Level of 1 1 G5-bacterial colonization (in CFU/g of tissue) in colonic tissue of mice injected with isotype control (1 1 G5 isotype control) or with RMP1 -14 antibodies (1 1 G5 Ab).

Figure 3: Response to PD-1 based immunotherapy as assessed by the MC38 tumoral volume (mm ³ ) in RMP1 -14 treated animals (PBS + Ab, triangles) versus untreated animals (PBS + isotype CTL, squares).

Figure 4: Lack of response of PD-1 based immunotherapy in mice preinfected with pks+ E. coli 1 1 G5 strain, as assessed by MC38 tumoral volume (mm3) in animals treated (1 1 G5 + Ab, triangles) or not (1 1 G5 + isotype CTL, squares) with RMP1 -14 antibody.

Figure 5: Densities of TILs in tumor and at the invasive margin of patients colonized by pks ⁺ or pks ^~ E. coli, were assessed by digital image analysis. ^* p= 0.05.

Figure 6: The colon cancer-associated E. coli 1 1 G5 strain modulates T cell density in the colon of APC ^Min/+ mice. Immunofluorescence staining for CD3 was performed on colon section at 50 days post infection. Cells density were obtained by digital image analysis of each stained section. The figure shows the density of TILs for each group of mice (non- infected : PBS, and 1 1 G5ACIbQ-infected : 1 1 G5). ^* p= 0.05.

Figure 7-12. Effect of pks+ E. coli preinfection on TILs level in mice grafted with MC38 tumors.

Figure 7: Representative flow cytometry plots to assess percentages of CD3 ⁺ T-cells among live CD45 ⁺ leucocytes in grafted MC38 tumors for each group of mice.

Figure 8: Significant decrease of percentages of CD3 ⁺ T-cells among live CD45 ⁺ leucocytes in tumors of 1 1 G5-infected mice compared to control mice (means plus SEM). Figure 9: Representative flow cytometry plots to assess percentages of CD8 ⁺ CD3 ⁺ T-cells among live CD45 ⁺ leucocytes in grafted MC38 tumors for each group of mice.

Figure 10: Significant decrease of percentages of CD8 ⁺ CD3 ⁺ T-cells among live CD45 ⁺ leucocytes in tumors of 1 1 G5-infected mice compared to control mice (means plus SEM). Figure 1 1 : Representative flow cytometry plots to assess percentages of neutrophils (CD1 1 b ⁺ Ly6G ⁺ Ly6C ^_ cells) among live CD45 ⁺ leucocytes in grafted MCD38 tumors for each group of mice.

Figure 12: Significant increase of percentages of neutrophils among live CD45 ⁺ leucocytes in tumors of 1 1 G5-infected mice compared to control mice (means plus SEM). Detailed description of the invention

In the context of the invention, the term "cancer" refers to any cancer for which anti- PD1 therapy can be contemplated. Examples of such cancers include colorectal cancer in particular progressive metastatic colorectal cancer, multiple melanoma, lung cancer in particular non-small cell lung cancer, bladder cancer, lymphoma in particular Hodgkin's lymphoma, B cell lymphoma or follicular lymphoma, renal cell carcinoma, squamous-cell carcinoma of the head and neck, ovarian cancer, Merkel cell carcinoma and urothelial carcinoma.

In a particular embodiment, said cancer is melanoma or colorectal cancer, more particularly colorectal cancer.

In a more particular embodiment, said colorectal cancer is Microsatellite Stable (MSS) colorectal cancer.

“Subject” for the purposes of the present invention includes humans and other animals, particularly mammals, and other organisms. Thus the methods are applicable to both human therapy and veterinary applications. In a particular embodiment the subject is a mammal, and in a more particular embodiment the subject is human.

The methods of the present invention comprise a step a) of determining, from a biological sample from the subject, as defined above, the presence of a pks island.

By "biological sample" is meant herein a sample obtained from the gastrointestinal tract of the subject such as a feces sample or a colonic biopsy sample.

By "pks island" is meant herein a 54 kb genomic island comprising a total of 19 genes ( clbA to clbS) which encode the machinery enabling the production of colibactin. This machinery consists of three non-ribosomal peptide megasynthases (ClbH, ClbJ, ClbN), three polyketide megasynthases (ClbC, Clbl, ClbO), two hybrid NRPS/PKS megasynthases (ClbB, ClbK), and nine accessory, tailoring and editing enzymes.

According to the present invention, the determination of the presence of a pks island is carried out by determining the presence of at least one expression product or expression product derivative of the pks island or by determining the presence of a pks island DNA.

By "at least one expression product of the pks island" is meant herein at least one protein or mRNA encoded by at least one of the 19 genes of the pks island, in particular at least one of the ClbH, ClbJ, ClbN, ClbC, Clbl, ClbO, ClbB and ClbK genes. By "expression product derivative of the pks island" is meant herein the metabolites, intermediate products and toxins, such as colibactin, produced via the machinery encoded by the pks island as defined above, in particular via the enzymes encoded by at least one of the ClbH, ClbJ, ClbN, ClbC, Clbl, ClbO, ClbB and ClbK genes.

The determination of the presence of at least one expression product of the pks island can be performed by any technique well-known from the skilled person, such as by Western Blot, enzyme-linked immunosorbent assay, immunochemistry, RT-PCR or qPCR.

The determination of the presence of at least one expression product derivative of the pks island can be performed by any technique well-known from the skilled person, such as by Western Blot, enzyme-linked immunosorbent assay or immunochemistry.

In a particular embodiment, the determination of the presence of a pks island is determined by comparing the determined level of said pks island with a predetermined threshold value. Statistical methods for determining appropriate threshold values will be readily apparent to the skilled person. The threshold values may have been determined, if necessary, from samples of subjects known to contain (positive control) or not to contain (negative control) pks islands.

In a particular embodiment, the presence of a pks island is determined in step a) by determining the presence of pks island DNA.

By "pks island DNA" is meant herein a nucleic acid included in the pks island, said nucleic acid being of at least 25 bp, in particular of at least 50 bp, at least 100 bp, at least 200 bp, at least 300 bp, at least 400 bp, at least 500 bp, at least 600 bp, at least 700 bp, at least 800 bp, at least 900 bp, at least 1000 bp, at least 2000 bp, at least 3000 bp, at least 4000 bp, at least 5000 bp, at least 6000 bp, at least 7000 bp, at least 8000 bp, at least 9000 bp, at least 10 000 bp, at least 20 000 bp, at least 30 000 bp, at least 40 000 bp, or at least 50 000 bp.

A pks island DNA may be in particular a nucleic acid included in at least one of the 19 genes of the pks island, in particular at least one of the ClbH, ClbJ, ClbN, ClbC, Clbl, ClbO, ClbB and ClbK genes. For instance, the pks island DNA is a nucleic acid included in the ClbN gene of the pks island.

The determination of the presence of a pks island DNA can be performed by any technique well-known from the skilled person, such as by PCR, using specific probes typically on a DNA chip, or by sequencing.

In a particular embodiment, the determination of the presence of a pks island DNA is performed by PCR.

In a particular embodiment, the determination of the presence of a pks island DNA is performed by determining the presence, by PCR, of at least one, in particular two, genes of the pks island, as defined above, in particular at least one, in particular two, genes selected from the group consisting of the ClbH, ClbJ, ClbN, ClbC, Clbl, ClbO, ClbB and ClbK genes.

In a particular embodiment, the determination of the presence of a pks island DNA is performed by PCR using primers specific for a pks island gene, in particular using primers respectively specific for at least one, in particular two, genes of the pks island, as defined above, in particular at least one, in particular two, genes selected from the group consisting of the ClbH, ClbJ, ClbN, ClbC, Clbl, ClbO, ClbB and ClbK genes.

In a particular embodiment, the determination of the presence of a pks island DNA is performed by PCR using primers specific for a pks island gene, in particular using primers specific for a ClbN gene of the pks island, more particularly using one or both of a primer comprising SEQ ID NO: 1 and a primer comprising SEQ ID NO: 2.

Typically, the presence of the pks island can be determined by the following method. Total DNA is preferably extracted from the feces sample to be studied. Then DNA amplification is preferably performed typically using a Taq DNA polymerase with primers, in particular 10 mM primers, typically located in the ClbN gene of the pks island. Suitable primers, disclosed in Johnson et al. (2008) J. Clin. Microbiol. 46:3906-391 1 , are typically of the following sequence:

- Forward primer: GTTTT GCT CGCCAG AT AGT CATT C (SEQ ID NO: 1 )

- Reverse primer : CAGTT CGGGT AT GT GT GG AAGG (SEQ ID NO: 2).

In a particular embodiment, the presence of a pks island is determined by the determining the presence of Escherichia coli bacteria positive for pks island.

Determining the presence of Escherichia coli bacteria positive for pks island may be performed by any suitable method well-known from the skilled person.

In particular, determining the presence of Escherichia coli bacteria positive for pks island may be determined by isolation, on agar, of E. coli bacteria present in the sample, and determination of the presence of a pks island DNA, as disclosed above.

Isolation, on agar, of E. coli bacteria present in the sample, can typically be performed by culturing, in particular aerobically culturing, said sample on suitable culture medium, such as onto MacConkey agar plates, then morphologically characterizing the obtained colonies and further identifying Gram-negative colonies using suitable bacterial identification kits. Alternatively, isolation, on agar, of E. coli bacteria present in the sample, can typically be performed by culturing said sample on Drigalski agar and chromogenic agar chromID CPS3 ^® (bioMerieux), which allow the identification of E. coli, and optionally by confirming said identification using automated Vitek II ^® system (bioMerieux). Typically, isolation, on agar, of E. coli bacteria present in the sample can be performed as disclosed in Buc et al. (2013) PLoS ONE 8:e56964.

Alternatively, the presence of Escherichia coli bacteria positive for pks island may be determined by determination of the presence of an E. coli- specific DNA in the sample, and determination of the presence of a pks island DNA, as disclosed above.

Determination of the presence of an E. coli- specific DNA in the sample can be performed by any method well-known from the skilled person, in particular by PCR.

In particular, determination of the presence of an E. coli- specific DNA can be performed by PCR using primers specific for E. coli bacteria.

Such primers specific for E. coli bacteria are well-known from the skilled person. Typically, primers specific for E. coli 16S DNA can be used, as disclosed in Sabat et al. (2000) Applied Environ. Microbiol. 66:844-849. Alternatively, primers specific for the E. coli uidA gene or primers specific for the E. coli flanking region of uspA gene can be used, as disclosed respectively in Choi et al. (2018) Korean J. Food. Sci. Anim. Resour. 38:829- 834 and Chen and Griffiths (1998) Lett. Appl. Microbiol. 27:369-371 .

Alternatively, determining the presence of Escherichia coli bacteria positive for pks island may be carried out by sequencing, DNA chip, FISH, Western Blot, enzyme-linked immunosorbent assay, immunochemistry, immunofluorescence, or global sequencing of the microbiota.

In the context of the invention, the term "anti-PD1 therapy" refers to any therapeutic treatment inhibiting or antagonizing the PD1 / PDL1 pathway.

Anti-PD1 therapies are well-known from the skilled person and include anti-PD1 antibodies or antibody derivatives such as pembrolizumab, nivolumab, cemiplimab, spartalizumab (also known as PDR001 ; see, e.g., WO 2015/1 12900), tislelizumab (also known as BGB-A317; see, e.g., WO 2015/035606), camrelizumab (also known as SHR- 1210; see e.g., WO 2015/085847), dostarlimab (also known as TSR-042; see, e.g., WO 2014/179664), sintilimab (also known as IBI308; see, e.g., WO 2017/025016), MEDI0608 (formerly AMP-514; see, e.g., WO 2012/145493 and U.S. Patent No. 9,205,148), PF- 06801591 (see, e.g., WO 2016/092419), JS001 (see, e.g., WO 2014/206107), MGA012 (see, e.g., WO 2017/019846), AGEN2034 (see, e.g., WO 2017/040790), and JNJ- 63723283 (see, e.g., WO 2017/0791 12), and anti-PDL1 antibodies or antibody derivatives such as atezolizumab, avelumab, durvalumab, BMS-936559 antibody or CK-301 antibody. In a particular embodiment, the anti-PD1 therapy is a therapy using an anti-PD1 and/or anti-PDL1 antibody.

The terms“antibody”,“immunoglobulin”, or“Ig” may be used interchangeably herein. The term antibody includes, but is not limited to, synthetic antibodies, monoclonal antibodies, recombinantly produced antibodies, multispecific antibodies (including bi specific antibodies), human antibodies, humanized antibodies, chimeric antibodies, intrabodies, single-chain Fvs (scFv) (e.g., including monospecific, bispecific, trispecific etc.), camelized antibodies, single domain antibodies, Fab fragments, F(ab') fragments, disulfide-linked Fvs (sdFv), anti-idiotypic (anti-ld) antibodies, and epitope-binding fragments of any of the above, fusion proteins comprising the antigen-binding domain of an antibody, fusion proteins based on a PD-1 or PD-L1 ligand (i.e.“traps” such as fusion proteins comprising a PD-1 or PD-L1 -binding domain of a natural PD-1 or PD-L1 ligand, fused to the Fc portion of an immunoglobulin) and any combination thereof. In particular, antibodies include immunoglobulin molecules and immunologically active portions of immunoglobulin molecules, i.e., antigen binding domains or molecules that contain an antigen-binding site that specifically binds to a PD-1 or PD-L1 antigen. The anti-PD1 or anti-PDL1 antibodies can be of any origin (human, murine or other), class (e.g., IgG, IgE, IgM, IgD, IgA and IgY) or subclass (e.g., lgG1 , lgG2, lgG2a, lgG2b, lgG3, lgG4, lgA1 and lgA2). In particular embodiments, the anti-PD1 or anti-PDL1 antibodies are humanized, such as humanized monoclonal anti-PD1 and/or anti-PDL1 antibodies.

In a particular embodiment, the anti-PD1 therapy is a therapy with cemiplimab.

By "resistance to anti-PD1 therapy" is meant herein that the anti-PD1 treatment does not improve the health condition of the treated patient with respect to the disease to be treated. For example, the tumor growth is not reduced and/or slowed down, and/or the survival rate is not increased.

By "predicting resistance to anti-PD1 therapy" is meant herein determining the likelihood that an anti-PD1 therapy, as defined above, if administered to a patient, will not improve the health condition of said patient with respect to the disease to be treated.

As used herein, the term“anti-cancer therapy” refers to any protocol, method, and/or agent that can be used in the prevention, management, treatment, and/or amelioration of a cancer. In certain embodiments, the terms“therapies” and“therapy” refer to a biological therapy, supportive therapy, and/or other therapies useful in the prevention, management, treatment, and/or amelioration of cancer known to one of skill in the art, such as medical personnel. By "anti-cancer therapy which is not an anti-PD1 therapy" is meant any therapy preferably known to have a beneficial effect on the particular cancer to be treated, which is not an anti-PD1 therapy as defined above.

Typically, when said cancer is a colorectal cancer, said anti-cancer therapy which is not an anti-PD1 therapy is selected from the group consisting of irinotecan, oxaliplatin, combination of irinotecan with 5-FU and leucovorin, combination of oxaliplatin with 5-FU and leucovorin, combination of irinotecan with capecitabine, combination of oxaliplatin and capecitabine, anti-EGFR antibodies such as cetuximab and panitumumab, anti-VEGFR antibodies such as bevacizumab, anti-VEGFR2 antibodies such as ramucirumab, aflibercept, zivaflibercept, regorafenib, combination of trametinib and palbociclib, reolysin ^® , combination of dabrafenib, trametiniv and panitumumab, and combinations thereof.

When it is determined that the pks island, in particular the E. coli positive for pks island, is absent in a biological sample, in particular a feces or colonic biopsy sample, from a subject, an anti-PD-1 and/or anti-PDL1 therapy may be administered in combination with an additional anti-cancer therapy.

In a certain embodiment, the additional anti-cancer therapy is selected from the group consisting of immune checkpoint inhibitors, radiation therapy (e.g., hypofractionated radiation therapy, stereotactic body radiation therapy), surgery, small molecule kinase inhibitors (e.g., sorafenib, ceritinib), chemotherapeutic agents (e.g., paclitaxel, pemetrexed, gemcitabine) including platinum-based chemotherapeutic agents (e.g, carboplatin, cisplatin), nucleic acid synthesis inhibitors (e.g., decitabine), cancer vaccines (e.g., ISA101 b), anti-CD38 antibodies (e.g., isatuximab), anti-MUC16 x CD3 bispecific antibodies (e.g., REGN4018), anti-CD20 x CD3 bispecific antibodies (e.g., REGN1979), granulocyte-macrophage colony-stimulating factor (GM-CSF), qhίί-TΰRb antibodies (e.g., SAR439459), an indoleamine-2, 3-dioxygenase (IDO) inhibitor, an IL-6R inhibitor, an IL-4R inhibitor, an IL-10 inhibitor, a cytokine such as IL-2, IL-7, IL-21 , and IL-15, an anti inflammatory drug such as corticosteroids, non-steroidal anti-inflammatory drugs, and combinations thereof.

In a particular embodiment, the immune checkpoint inhibitor is an anti-CTLA4 antibody (e.g., ipilimumab, REGN4659) or an anti-LAG-3 antibody (e.g., REGN3767).

The present invention also concerns an anti-PD1 and/or anti-PDL1 antibody, as defined above, for use in a method for treating a subject suffering from cancer, said method comprising: a) determining, from a biological sample from the subject, in particular a feces or colonic biopsy sample from the subject, the presence of a pks island as defined above, b) when it is determined in step a) that said pks island is absent, administering to said subject a therapeutically effective amount of an anti-PD1 and/or anti-PDL1 antibody as defined above.

Another object of the invention concerns the use of an anti-PD1 and/or anti-PDL1 antibody as defined above in the manufacture of a medicament intended for a method for treating a subject suffering from cancer, said method comprising:

a) determining, from a biological sample from the subject, in particular a feces or colonic biopsy sample from the subject, the presence of a pks island as defined above, b) when it is determined in step a) that said pks island is absent, administering to said subject a therapeutically effective amount of an anti-PD1 and/or anti-PDL1 antibody as defined above.

The present invention further concerns a method for treating a subject suffering from cancer, said method comprising the steps of:

a) determining, from a biological sample from the subject, in particular a feces or colonic biopsy sample from the subject, the presence of a pks island as defined above, b) when it is determined in step a) that said pks island is present, administering to said subject a therapeutically effective amount of an anti-cancer therapy which is not an anti-PD-1 therapy as defined above; or

when it is determined in step a) that said pks island is absent, administering to said subject a therapeutically effective amount of an anti-PD-1 therapy as defined above.

Unless otherwise indicated, “treating” or “treatment” of a disease, disorder, or syndrome, as used herein, means inhibiting the disease, disorder, or syndrome, that is, arresting its development; and relieving the disease, disorder, or syndrome, that is, causing regression of the disease, disorder, or syndrome. As is known in the art, in the context of treatment, adjustments for systemic versus localized delivery, age, body weight, general health, sex, diet, time of administration, drug interaction and the severity of the condition may be necessary, and will be ascertainable with routine experimentation by one of ordinary skill in the art.

The terms“administer” or“administration” refers to the act of injecting or otherwise physically delivering a substance as it exists outside the body (e.g., a formulation of the invention) into a patient, such as by mucosal, intradermal, intravenous, subcutaneous, intramuscular delivery and/or any other method of physical delivery described herein or known in the art. When a disease, or a symptom thereof, is being treated, administration of the substance typically occurs after the onset of the disease or symptoms thereof. When a disease or its symptoms are being prevented, administration of the substance typically occurs before the onset of the disease or symptoms thereof.

The terms “effective amount” or “pharmaceutically effective amount” or “therapeutically effective amount” refer to a sufficient amount of an agent to provide the desired biological, therapeutic, and/or prophylactic result. That result can be reduction, amelioration, palliation, lessening, delaying, and/or alleviation of one or more of the signs, symptoms, or causes of a disease, or any other desired alteration of a biological system. In reference to cancer, an effective amount comprises an amount sufficient to cause a tumor to shrink and/or to decrease the growth rate of the tumor (such as to suppress tumor growth) or to prevent or delay other unwanted cell proliferation. In some embodiments, an effective amount is an amount sufficient to delay development. In some embodiments, an effective amount is an amount sufficient to prevent or delay recurrence.

An effective amount can be administered in one or more administrations. The effective amount of the therapy may: (i) reduce the number of cancer cells; (ii) reduce tumor size; (iii) inhibit, retard, slow to some extent, and preferably stop cancer cell infiltration into peripheral organs; (iv) inhibit (i.e., slow to some extent and preferably stop) tumor metastasis; (v) inhibit tumor growth; (vi) prevent or delay occurrence and/or recurrence of tumor; and/or (vii) relieve to some extent one or more of the symptoms associated with the cancer.

The amount can vary depending on such factors as the size and weight of the subject, the type of illness, or the particular compound of the therapy. The amount also can vary depending on the compound, the disease state and its severity, the age of the patient to be treated, and the like. The effective amount can be determined by one of ordinary skill in the art having regard to their knowledge and to this disclosure.

The terms“comprising” and“including” are used herein in their open-ended and non-limiting sense unless otherwise noted.

The terms“a” and“an” and“the” and similar references in the context of describing the invention (especially in the context of the following claims) are to be construed to cover both the singular and the plural, unless otherwise indicated herein or clearly contradicted by context. Description of the sequences

The present invention will be further illustrated by the examples below.

Example

Colorectal cancer (CRC) is the third most common malignancy and is the fourth leading cause of cancer-related deaths in men and the third leading cause among women worldwide. Host factors associated with environmental and nutritional factors play an important role in the colorectal carcinogenesis. In the recent two decades, growing attention has been given to the role of intestinal microbiota dysbiosis in CRC- carcinogenesis. Studies showed a reduction of tumorigenesis in the intestines of animals raised in a germ-free environment, in both genetically induced or colitis-associated CRC mouse models. Moreover, colorectal cancer patients often exhibit a distinct microbiota compared to healthy population (Nakatsu et at. (2015) Nat. Cornmun. 6:8727). Mechanisms by which intestinal microbiota dysbiosis has been shown to promote colorectal carcinogenesis are not yet elucidated. Association of exposure to antibiotics in early life with colorectal adenoma risk development at the age of 60, suggests the impact of microbiota dysbiosis at the first steps of the carcinogenesis. Recently Wong et al. showed that gavage with feces from colorectal cancer patients promote intestinal carcinogenesis in germ-free and azoxymethane-(AOM)-carcinogen mice models (Wong et al. (2017) Gastroenterology 153:1621 -1633). Despite the large number of studies, CRC associated dysbiosis is partly defined. However, some bacterial strains, such as Bacteroides fragilis, Fusobacterium nucleatum and Escherichia coli, have a well- established role in CRC development.

Even if E. coli is a commensal bacterium, several strains have acquired some virulence factors including genotoxins, such as colibactin. The colibactin toxin, a hybride polyketide non-ribosomal peptide is encoded by the polyketide synthase {pks) pathogenicity island (Johnson et al. (2008) J. Clin. Microbiol. 46:3906-391 1 ). This colibactin-p/cs-island was preferentially detected in CRC samples in comparison to non- neoplasic control (Nakatsu et al. (2015) Nat. Commun. 6:8727; Arthur et al. (2012) Science 338:120-123; Buc et al. (2013) PLoS ONE 8:e56964; Prorok-Hamon et al. (2014) Gut 63:761 -770; Eklof et al. (2017) Int. J. Cancer 141 :2528-2536). Genes for colibactin ( clbB gene of the p/cs-island) were highly enriched in familial adenomatous polyposis (FAP) patients’ colonic mucosa compared to healthy individuals. In the same way, Nakatsu et al. detected pks- positive bacteria in adenoma suggesting the presence of these bacteria during early stages of carcinogenesis (Nakatsu et al. (2015) Nat. Commun. 6:8727).

Using a well characterized subcutaneous syngenic mouse model the inventors evaluated the impact of the chronic infection on anti-tumoral response mediated by anti- PD-1 immunotherapy. In parallel, they evaluated the T-cell populations in a collection of human CRC samples previously characterized for their pks island status

Materials and methods

Bacteria strains

The representative CRC-colibactin-producing E. coli strain named 1 1 G5 was isolated from the colonic tissue of a patient with colon cancer and was resistant to ampicillin and kanamycin (Buc et al. (2013) PLoS ONE 8:e56964).

Animals

All studies were approved by local ethical committee (No. CE-2912) and the French Ethical Animal Use Committee (Apafis#5401 and Apafis#13812). Studies were performed using 6- to 7-week-old wild-type (WT) C57BL/6J mice (Charles River Laboratories, L’Abresle, France). All mice were housed in conventional conditions at the animal care facility of Universite Clermont Auvergne (Clermont-Ferrand, France).

11G5 impact on anti-PD-1 mAb efficacy in mice

Transplantable MC38 CRC cells derived from C57BL/6J mice were maintained as monolayers using culture medium consisting of DMEM (Invitrogen, Cergy Pontoise, France) supplemented with 10% FCS (Biowest, Nuaille, France), 1 % glutamin, 1 % hepes, 1 mM non-essential amino acids, 1 mM sodium pyruvate at 37°C in a humidified incubator containing 5% CO2. Experiments were performed in 6-to 8-week old WT C57BL/6J male mice with 8 animals per group. To enhance E. coli strain colonization, the inventors administrated streptomycin (2.5 g/L) for 3 days before oral inoculation of bacteria ( ^« 1 x 10 ⁹ bacteria) or PBS alone (non-infected controls). Nine days after infection, mice were anesthetized by isoflurane inhalation and were inoculated with 1 x 10 ⁶ MC38 cells by dorsal subcutaneous injection at day 0 of the experiment. At 8, 1 1 , 14, 18, 20 and 22 days after tumor cell injection, non-infected or 1 1 G5-infected mice were injected intraperitoneality (i.p) with the PD-1 specific blocking antibody (mu anti-PD-1 ; mlgG1 -a chimeric version of the RPM1 -14 rat monoclonal antibody with a mouse lgG1 Fc domain) or an lgG1 isotype control antibody (mlgG1 , clone 1 B71 1 ) (10 pg/g of mice). The RPM1 -14 rat monoclonal antibody is described in Natalia Martin-Orozco et al. (J Immunol December 15, 2006, 177 (12) 8291 -8295). The chimeric version of the RPM1-14 rat monoclonal antibody with a mouse lgG1 Fc domain has a light chain of SEQ ID NO: 3 and a heavy chain of SEQ ID NO: 4. To evaluate potential toxicity, body weight was measured twice a week. To monitor tumoral growth, tumor volume in mm ³ was calculated twice a week from the measurement of two perpendicular diameters using a caliper according to the formula LxS ² /2 where L and S are the largest and smallest diameters in mm respectively. The mice were sacrificed at day 30 of the experiment. Tumors were removed and weighted.

Chronic infection animal model

Bacterial infection of C57BL/6J-Apc ^Min/+ female was performed as described in Bonnet et al. (2014) Clin Cancer Res 20:859-867. To enhance E. coli strain colonization, streptomycin (2.5 g/L) was administered for 3 days prior to oral inoculation with bacteria ( ^« 1 x 10 ^s bacteria in PBS) or PBS alone (non-infected controls). Three E. coli strains were tested: the pks- positive 1 1 G5 strain, its isogenic mutant 1 1 G5ACIbQ and the K-12 MG1655 commensal strain. Faecal bacterial colonization was periodically quantified as colony-forming unit per stool as described in Bonnet et al. (2014) Clin Cancer. Res. 20:859-867. Fifty days after inoculation, the mice were sacrificed. The colons were removed from the caecum to the rectum. Then, the polyps were counted. Tissues were prepared for bacterial colonization and immunostaining experiments. Mesenteric lymph nodes (MLNs) were harvested and homogenized then bacterial colonization was evaluated by selective culture and immune cell were analysed by flow cytometry. According to the parameters studied, at least 6 animals per group were tested and the experiments were reproduced at least twice. Immunofluorescent staining and quantitation of immune cells

At the end of the experiment the colons were swiss-rolled from the distal to proximal end and fixed for 24 h in 10% formalin (Sigma, Tokyo, Japan) at room temperature. Blocks were embedded in paraffin and cut into 5 pm sections, then the tissue sections were prepared for hematoxylin-eosin-safranin staining or immunofluorescence stainings to analyse colonic immune cells in tumor. Immunostainings were performed in four serial sections to target some immune cell populations: CD4 T cell (CD3 ⁺ CD4 ⁺ ); CD8 T cells (CD3 ⁺ CD8 ⁺ ). All IF stainings were performed using an automated Stainer, Discovery XT processors (Roche, Bale, Swiss), and the tyramide signal amplification (TSA)-conjugated fluorochrome method was used on whole colon slides. Cells quantification was made on the whole colonic mucosa by a specific digital image analysis process.

Immune contexture determination on human CRC biopsies

Immune contexture (CD3 and CD8 densities) were determined on 40 CRC samples from the MiPaCor collection samples (DC-2017-2972). All the samples were previously tested for pks status on colonic tissues by PCR (Gagniere et al. (2017) Clin. Sci. 131 :471 - 485), using as forward primer, the primer of sequence SEQ ID NO: 1 and as reverse primer, the primer of sequence SEQ ID NO: 2. The inventors selected 20 p/cs-negative and 20 pks- positive samples. After colonic resection, fresh specimens were collected, fixed in buffered 4% paraformaldehyde, embedded in paraffin, cut into 5 pm slices. Two tissue paraffin sections of 4 pm were processed for immunohistochemistry. Digital images of the stained tissue sections were obtained at 20x magnification and 0.45 pm/pixel resolution. The densities of CD3 ⁺ and CD8 ⁺ T cells in colon tumor and invasive margin regions were determined.

Analysis of MC38 Tumor Infiltrating immune cells by flow cytometry

Single-cell suspensions from MC38 tumors were prepared by using Mouse Tumor Dissociation kit and gentleMACS Dissociator (Miltenyi-Biotec) according to the manufacturer’s instructions. All antibodies used in this study are described in Table 1 below. The cell suspensions were washed in PBS and stained with fixable Viability Dye eFluor450 (eBioscience) according to the manufacturer’s instructions. The cells were then washed and incubated at 4°C with anti-CD16/CD32 before the surface staining of the CD45, CD3, CD4, CD8, CD1 1 b, GR-1 +, Ly6G+, and Ly6C+ immune receptors. Data were acquired on a BD LSR II flow cytometer (Becton-Dickinson), and analysis was performed using FlowJo™ (TreeStar) and BD FACSDIVA™ software (BD Biosciences). Table 1 : Antibodies used in immunofluorescence for staining of whole colon sections

(IHC-IF) or flow cytometry (FC).

Used Target Species Clone Manufacturer Fluorochrome / for coupled Protein

IHC-IF CD45 Rabbit NA Abeam NA

Ly6G Rat 1A8 Biolegend NA

CD3 Rabbit SP7 ThermoFisher NA

CD4 Rat 4SM95 ThermoFisher NA

CD8 Rat 4SM15 ThermoFisher NA

Rat IgG Goat NA Jackson NA

Rabbit NS NS Roche HRP

IgG

Rat IgG NS NS Roche HRP Goat IgG NS NS Roche HRP

FC CD45.2 Mouse 104 ThermoFisher PE-efluor610

TCR-b Armenian H57-597 ThermoFisher APC-eFluor780 CD3 Rat 17A2 ThermoFisher APC-eFluor780 CD4 Rat RM4-5 ThermoFisher APC CD8 Rat 53-6.7 ThermoFisher PerCP-eFluor 710 CD25 Rat 3C7 ThermoFisher A488 FoxP3 Rat FJK-16S ThermoFisher PE CD1 1 b Rat M 1/70 ThermoFisher A700 GR-1 Rat RB6-8C5 ThermoFisher A488 Ly6G Rat 1A8 ThermoFisher PerCPefluor 710

Ly6C CD16/CD32 Rat 93 ThermoFisher NA

APC: Allophycocyanin, HRP: Horseradish Peroxidase, NA: Not Applicable, NS: Not Specified, PE: Phycoerythrin, Per-CP: Peridinin-chlorophyll proteins.

Statistical analyses

One-way ANOVA followed by Tukey's post-test or unpaired Student t test were used for statistical analyses. Correlations were determined by Spearman test. All tests were performed using Graph Pad Prism 7 (StataCorp, College Station, TX, USA)

p-values < .05 were considered statistically significant. Results

Infection with the colon cancer-associated E. coli 11G5 strain induces a resistance to the anti-PD- 1 immunotherapy.

The inventors investigated the impact of a chronic infection of the gut with 1 1 G5 strains on the anti-PD-1 therapy efficacy. For this purpose, they chose the MC38 murine tumors grafts on C57BI6 mice as animal model. Animals were firstly orally infected with the 1 1 G5 strain and 9 days post-infection MC38 cells were subcutaneously inoculated. Then MC38 tumors-bearing mice were injected with anti-PD-1 treatment or the isotype antibody control at 8, 1 1 , 14, 18, 20 and 22 days after MC38-cells inoculation. Figures 1 and 2 showed that anti-PD-1 treatment did not impact the bacterial colonization of the gut. In Figure 3 the inventors observed a significant response of MC38 tumors to anti-PD-1 treatment. Indeed, tumoral growth was significantly slowed, demonstrating the anti- tumoral efficacy of the anti-PD-1 treatment in the experimental conditions. In contrast, no anti-tumoral effect was observed in animals infected with the 1 1 G5 strain and subjected to the same treatment (Figure 4).

The effect of colibactin-producing E. coli chronic infection on MC38 murine graft model was investigated using flow cytometry.

A significant decrease in CD3 ⁺ tumor infiltrating-lymphocytes (TILs) was observed in MC38 tumors of 1 1 G5 infected animals in comparison to control animals (Figures 7 and 8). A global decrease of CD8 ⁺ TILs was observed only in 1 1 G5-infected mice group (Figures 9 and 10). No effect was noticed for the CD4 ⁺ TILs (data not shown). Finally, a significant increase of neutrophils population was measured in 1 1 G5-infected samples (Figures 1 1 and 12).

The colon cancer-associated E. coli 11G5 strain induces a decrease of CD3 ⁺ cells population in Min mice colon tumors.

T cell populations were investigated by analyzing the total T cells population by CD3 staining at 50 days p.i., on colonic sections of tumors from infected and control Min mice (non-infected and 1 1 G5ACIbQ-infected). Representative CD3 staining of tumors was obtained. A significant decrease of CD3 ⁺ cells was observed in the tumors (Figure 6). It was observed that intra-tumoral repartition of CD3 ⁺ cells was heterogeneous and the decrease was particularly observed in the margin of the polyps. pks-positive-E. coli colonization is associated with a decrease of tumor infiltrating- iymphocytes T cells (TILs) at the invasive margin of human CRC samples.

In order to make a link with colorectal carcinoma in human and see if the presence of T cells as Tumor infiltrating lymphocytes or in the tumor invasive margin could be correlated with the presence of pks-positive E. coli, CD3 and CD8 stainings were performed on 40 CRC tumors samples at the immunomonitoring Platform of the Hopital Europeen Georges Pompidou (Paris).

The densities of CD3 ⁺ and CD8 ⁺ T cells in colon tumor and invasive margin regions were determined and correlated with the E. coli colonization and the presence of pks- positive bacteria. No significant correlation was observed between colonization of the tissue by E.coli and CD3 ⁺ or CD8 ⁺ cell populations (data not shown). Figure 5 showed a significant decrease of CD3 ⁺ cells in pks- positive tumors only in the invasive margin.

Discussion

In the course of studying immune micro environment, we focused our work on the T- cell populations in the colon of Apc ^Min/+ mice as this population was shown in humans to be of importance for cancer prognosis. It appeared that presence of E.coli positive for pks island in these mice microbiota decreased the number of T cells in the various immune regions that we looked at.

The inventors hypothesized that the presence of pks+ E.coli could affect the efficacy of anti-tumoral treatments in particular aimed at T cell activation such as checkpoint inhibitors. To monitor easily the tumoral growth, the inventors chose to test the anti-PD1 immunotherapy on the MC38-subcutaneaous graft in mice infected orally by pks+ E.coli. Experimental therapeutic response of this model was shown to be sensitive to the microbiota environment. Moreover it was used in different studies to evaluate the impact of several micro-organisms on the efficacy of anti-PD-1 antibodies (Gopalakrishnan et al. (2018) Science 359:97-103; Sivan et al. (2015) Science 350:1084-1089). With this model, the inventors showed that pks+ E. coli infection induces a resistance of the MC38 tumors to anti-PD-1 immunotherapy.

The inventors further showed a significant decrease of CD3 ⁺ cells in human tumors colonized by some pks- positive E. coli suggesting a relation between the presence of this bacteria and CRC prognostic.

This study shows that E. coli bacteria positive for pks island, can be a new biomarker to predict anti-PD-1 response in CRC patients.