The present invention relates to a method for in vitro diagnosis of Constitutional Mismatch Repair Deficiency (CMMR-D). More specifically, the present invention relates to the use of microsatellite markers for the diagnosis of CMMR-D and for identifying patient requiring monitoring for detection of cancer onset and/or patient at risk of developing a tumor. The present invention also relates to a method of identification of microsatellite markers associated with CMMR-D.

The microsatellite instability (MSI) phenotype occurs specifically in Mismatch Repair (MMR)-deficient cells due to the accumulation of replication errors in repetitive DNA sequences. MSI has been widely reported in the tumor DNA of both inherited and sporadic MMR-deficient neoplasms. However, this type of genetic instability has not been detected in the germline DNA of patients suffering from Hereditary Non Polyposis Colorectal Cancer (HNPCC), also called Lynch syndrome. The Lynch syndrome is due to dominant autosomal mutation in one of the 4 genes MLH1, MSH2, MSH6, PMS2 called "MMR genes" which are involved in the mismatch repair during DNA replication. Patients suffering from Lynch syndrome have a risk of 80% to develop, in their adult years, colon cancer and various cancer affecting endometrium, ovary, stomach, small intestine, liver, superior urinary system, brain and skin.

In contrast to individuals with Lynch syndrome, cases with biallelic deleterious germline mutations in MMR genes leading to constitutional mismatch repair-deficiency (CMMR-D), or Lynch-3 syndrome, have been recognized since 1999 (Ricciardone et al. (1999) Cancer Res. 59: 290-3 ; Wang et al. (1999) Cancer Res. 59: 294-7). The CMMR- D/Lynch 3 syndrome (herein after called CMMR-D) is characterized by the development of childhood tumors and a huge clinical spectrum very different from one patient to another. The tumors are mainly lymphomas, leukemias, astrocyte-derived brain tumors and/or very early-onset colorectal tumors. Other signs, named call signs, such as atypical "cafe-au- lait" spots are frequently observed in CMMR-D patients, being however also present in healthy subjects from the general population. They allow to recognize, even if not specific, at-risk individuals for CMMR-D syndrome in the general population before the emergence of cancers.

To date, when CMMR-D syndrome is suspected in patients, three tests for diagnosing the disease are available.

The first one concerns the identification of a MSI process in healthy tissues, such as blood or skin, of these patients using limiting dilution analysis of DNA, but results are very contradictory in the litterature and the method is neither sensible nor specific enough, besides being complex and expensive (Wimmer and Etzler (2008) Hum. Genet. 124: 105- 22; Gallinger et al. (2004) Gastroenterology. 126: 546-585; Felton et al. (2007) Clin. Genet. 71 : 483-498; Bacon et al. (2001 ) Oncogene. 20: 7467-7471 ). In fact, it has been demonstrated in these publications that the methodology used is critical for identification of MSI process, and that the methodology consisting of using a small-pool PCR, i.e. using limiting dilution of DNA, allow the identification of MSI, whereas the methodology consisting of using conventional PCR, i.e. using un-diluted DNA, does not allow MSI identification.

The second test concerns the search for constitutional mutations of MMR genes in these patients by sequencing the MMR genes MLH1, MSH2, MSH6 and PMS2, but this reference method can be used, as for many genetic diseases, when the diagnosis is strongly suspected, and its implementation in a routine screening, from a call sign, is not possible because this method is too long and too expensive. Moreover, it is often difficult to identify mutations in two allelles of the same gene in these patients and a lot of coding variants can have a significance not yet specified.

The third test concerns the identification of a MSI phenotype in the tumor, but this method allows a diagnosis only after the development of a tumor and it needs to be confirmed in terms of reliability. In fact, there is often a difficulty to establish a MSI phenotype in tumors which are not colorectal cancers, for instance in lymphoma or cerebral cancers, and these patients can also developped cancers which are not due to MSI. Furthermore, identification of a MSI phenotype in a tumor does not necessarily means that the patient is suffering from CMMR-D syndrome since it can be a sporadic MSI cancer (rare in young patients) and/or especially a consecutive MSI tumor in a Lynch patient.

Thus, there is a need to develop a simple method that detects constitutive MSI in germline DNA from normal cells of CMMR-D patients, before the emergence of tumors, allowing an early diagnosis of CMMR-D syndrome.

There is also a need to early identify patient suffering from CMMR-D syndrome in order to establish a monitoring for detection of cancer onset.

The inventors surprinsingly identified that cultured, and optionally immortalized lymphocytes, from CMMR-D patients acquire MSI phenotype during culture and, hence enable for the early detection of constitutive MSI. Moreover, contrary to that was previously accepted, the inventors identified that MSI identification is achievable using conventional PCR, i.e. using un-diluted DNA sample.

Use of microsatellite markers for the diagnosis of constitutional mismatch repair-deficiency

A first aspect of the invention is a method for in vitro diagnosing constitutional mismatch repair-deficiency (CMMR-D), which method comprises the steps of:

a) culturing lymphocytes obtained from a patient likely to be affected by CMMR-D; and

b) detecting the presence or absence of instability on at least one microsatellite marker of CMMR-D in the cultured lymphocytes of step a);

wherein the detection of instability on at least one microsatellite marker in the cultured lymphocytes of the patient indicates that the patient suffers from CMMR-D.

As used herein, the term "CMMR-D" refers to an inherited cancer syndrome, also named Lynch-3 syndrome, involving bi-allelic germline mutations in at least one of the MMR genes MLH1, MSH2, MSH6 and PMS2. CMMR-D is usually characterized by the development of chilhood tumors (before the age of 15) but not necessarily, as a first tumor may occur in CMMR-D patients when adults. The tumors may correspond, without limitation, to lymphomas, leukemias, brain tumors, such as glioblastoma, astrocytoma, or neuroblastoma, and/or very early-onset colorectal tumors. Other signs, named "call signs", are frequently observed in CMMR-D patients. Call signs may correspond to (i) presence of atypical "cafe-au-lait" spots, and/or (ii) parental consanguinity with the development of a tumor in a child, and/or iii) an evocative familial history of Lynch syndrome, and/or iv) a brother or a sister suffering from a cancer in childhood, and/or v) a childhood cancer outside the spectrum of tumors conventionally associated with the neurofibromatosis type 1 (NF1 ). By "an evocative familial history of Lynch syndrome", it is meant signs which can evocate a familial Lynch syndrome or HNPCC according to the Bethesda criteria : a) a colorectal cancer before the age of 50; or b) presence of at least two tumors of the HNPCC spectrum, such as colorectal cancer, stomach cancer, endometrial cancer, bladder cancer, urinary tract cancer, ovary cancer, whatever the age at disease onset; or c) a colorectal cancer, diagnosed before the age of 60, with an evocative histology of an MSI phenotype ("Crohn-Like"); or d) a patient with a colorectal cancer having a first- degree relative suffering from a tumor of the HNPCC spectrum diagnosed before the age of 50; or e) patients with a colorectal cancer having two or more first-degree relatives suffering from a tumor of the HNPCC spectrum, whatever the age at disease onset. As used herein, the patient is a human or a non human mammal, in particular a rodent, a feline, a canine, a bovine or an ovine. In a preferred embodiment, the patient is a human. More particularly, the patient is a child, an adult, a man or a woman. In a particular preferred embodiment, the patient is a child.

By "patient likely to be affected by CMMR-D", it meant a patient who presents call signs of CMMR-D, a patient who is suffering from childhood tumors such as leukemia, lymphoma, glioblastoma, neuroblastoma, astrocytoma and/or colorectal cancer, a patient who is suffering from tumors in adulthood, and/or a patient who suffered from a first cancer in childhood and who is suffering from a second cancer in adulthood.

In a particular embodiment, the method comprises a step of immortalization of lymphocytes from a patient affected by CMMR-D prior to step a). Accordingly, the method of the invention may comprise the steps of:

aO) immortalizing lymphocytes obtained from a patient likely to be affected by

CMMR-D,

a) culturing the immortalized lymphocytes obtained at step aO); and

b) detecting the presence or absence of instability on at least one microsatellite marker of CMMR-D in the cultured immortalized lymphocytes of step a);

wherein the detection of instability on at least one microsatellite marker in the cultured immortalized lymphocytes of the patient indicates that the patient suffers from CMMR-D.

In a particular embodiment, the detection of instability on at least three microsatellite markers, in the cultured lymphocytes, optionally immortalized lymphocytes, of the patient indicates that the patient suffers from CMMR-D. In another particular embodiment, the detection of instability on at least five microsatellite markers, in the cultured lymphocytes, optionally immortalized lymphocytes, of the patient indicates that the patient suffers from CMMR-D.

In order to carry out the process according to the present invention lymphocytes (preferably mononuclear lymphocytes) may firstly be isolated from e.g. the blood or plasma according to known methods e.g. by a Ficoll density gradient centrifugation.

According to the method of the invention, in step aO) immortalization of lymphocytes may be performed by conventional methods of immortalization. For example, lymphocytes can be immortalized by the Epstein Barr Virus (EBV).

According to the method of the invention, the step of culturing lymphocytes, optionally immortalized lymphocytes, is performed by conventional method of culture, for instance using RPMI-1640 medium or Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% of Foetal Calf Serum. In an embodiment, culturing lymphocytes in step a) is performed in the presence of proliferative agents, such as Phytohemagglutinine (PHA), Concanavaline A, Pokeweed Mitogen (PWM), Nocardia Water Soluble Mitogen and Lipopolysaccharide (LPS).

In a preferred embodiment, lymphocytes, optionally immortalized lymphocytes, may be cultured during 1 to 150 days, more preferably during 30 to 150 days, 60 to 120 days, 80 to 1 10 days, or 90 to 100 days. For instance, lymphocytes, optionally immortalized lymphocytes, may be cultured during 1 , 2, 3, 4, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 1 10, 1 15, 120, 125, 130, 135, 140, 145 or 150 days.

As used herein, the terms "microsatellite marker of CMMR-D" refers to microsatellites which have been identified as being involved in CMMR-D. Preferably, the microsatellite markers of CMMR-D are selected from the group consisting of BAT26, NR21 , NR27, NR24 and BAT25. Still preferably, said at least one microsatellite marker is BAT26, NR21 and NR27, and more preferably BAT26, NR21 , NR27, NR24 and BAT25.

The BAT26 microsatellite marker is a 26A repeat localized in the intron 5 of the hMSH2 gene, and is found for instance between positions 16298 and 16323 of the NCBI Reference Sequence NG_00071 10.1 . More preferably, the BAT26 microsatellite marker has a repeated pattern comprised between 20A to 26A.

The NR21 microsatellite marker is a 21 T repeat localized in the 5' untranslated region of the SLC7A8 gene, and is found for instance between positions 483 and 503 of the NCBI Reference Sequence NM_012244.2. More preferably, the NR21 microsatellite marker consists of a repeated pattern comprised between 16T to 21 T.

The NR27 microsatellite marker is a 26A repeat localized in the 5' untranslated region of the BIRC3 gene, and is found for instance between positions 1031 and 1056 of the NCBI Reference Sequence NM_001 165.3. More preferably, the NR27 microsatellite marker consists of a repeated pattern comprised between 23A to 26A.

The NR24 microsatellite marker is a 23T repeat localized in the 3' untranslated region of the ZNF2 gene, and is found for instance between positions 3248 and 3270 of the NCBI Reference Sequence NM_021088.2. More preferably, the NR24 microsatellite marker consists of a repeated pattern comprised between 18T to 23T.

The BAT25 microsatellite marker is a 25T repeat localized in the intron 16 of the c- KIT gene, and is found for instance between positions 741 18 and 74142 of the NCBI Reference Sequence NG_007456.1 . More preferably, the BAT25 microsatellite marker consists of a repeated pattern comprised between 19T to 25T.

According to the present invention, in step b) detection of the presence or absence of instability may be performed by any method well-known in the art. In a preferred embodiment, detection of the presence or absence of instability is performed on an un-diluted DNA sample. By "un-diluted DNA sample", it is meant that the DNA has not been diluted down to achieve a limiting dilution of 1 to 3 genomes per sample. In other word, detection of the presence or absence of instability is performed on an un-diluted DNA sample that contains at least 4 genomes per sample.

In a preferred embodiment, detection of the presence or absence of instability is performed by sequencing at least one microsatellite marker of CMMR-D.

In a particular preferred embodiment, detection of the presence or absence of instability is performed by genotyping at least one microsatellite marker of CMMR-D.

The genotyping can be performed by amplifying the microsatellite markers and isolating the amplification products by capillary electrophoresis. Amplification of the microsatellite markers can be performed using conventional polymerase chain reaction (PCR) techniques, i.e. using an un-diluted DNA sample, and more preferably, amplification of microsatellite markers is performed by multiplex PCR. In other words, amplification of the microsatellite markers is not performed using small-pool PCR, i.e. on a DNA sample extensively diluted to 1 to 3 genomes equivalents per PCR reaction (Parsons et al. 1995, Cancer. Res., 55:5548-5550).

In a preferred embodiment, amplification of the microsatellite markers comprises an initial denaturation step followed by cycles of denaturation-annealing-elongation steps and a final extension step.

The initial denaturation step may be performed under heating conditions ranging from 90 °C to 105^, during 15 sec to 15 min. Preferably, the heating conditions range from 92 ^< C to 102 ^< C, more preferably from 95 °C to 100°C, still more preferably the heating conditions are at 95°C. Preferably, the initial denaturation step is performed during 1 min to 15 min, more preferably during 2 min to 12 min, still more preferably during 5 min to 10 min, and still more preferably the initial denaturation step is performed during 5 min.

In a preferred embodiment, the initial denaturation step is performed at 95 °C during 5 min.

Each cycle of denaturation-annealing-elongation step includes a denaturation phase under heating conditions, followed by an annealing phase performed under conditions which allow the hybridization of the primers to the sequence to be amplified, and an elongation phase performed under conditions which allow the polymerase to synthesizes an extension product from each primer that is annealed to the sequence to be amplified. The denaturation phase may be performed between 90 °C to Ι Οδ'Ό, preferably 92 °C to 100°C, more preferably between 94 °C to 98 Ό, during 10 sec to 4 min, preferably during 10 sec to 2 min, more preferably during 15 sec to 1 min.

The annealing phase may be performed between 35 °C and 70 °C, preferably between 40 °C to 65 °C, more preferably between 45 °C to 60 °C, still more preferably between 50°C to 60°C, during 10 sec to 2 min, preferably during 20 sec to 1 ,5 min, more preferably during 25 sec to 45 sec.

The elongation phase may be performed between 40 °C and 80 °C, preferably between 50 °C to 75 °C, more preferably between 60 °C to 72 °C, during 10 sec to 5 min, preferably during 20 sec to 3 min, more preferably during 25 sec to 1 min, still more preferably during 30 sec to 45 sec.

In a preferred embodiment, the denaturation phase is performed at 95 °C during 30 sec, the annealing phase is performed at 55 °C during 30 sec, and the elongation phase is performed at 72 °C during 30 sec. The denaturation-annealing-elongation step may be repeated during 30 to 60 cycles, preferably during 32 to 40 cycles, more preferably during 35 to 40 cycles. Still more preferably, the denaturation-annealing-elongation step is repeated during 35 cycles.

The final extension step may be performed between 40 °C and 80 °C, preferably between 50 °C to 75°C, more preferably between 60°C to 72°C, during 1 min to 10 min, preferably during 3 min to 8 min, more preferably during 4 min to 6 min.

In a preferred embodiment, the final extension step is performed at 72 °C during 5 min.

Amplification of the microsatellite marker BAT26 may be performed using the forward primer having the sequence 5' - CTG CG GT AATC AAGTTTTT AG -3 ' (SEQ ID NO: 1 ) and the reverse primer having the sequence 5'-AACCATTCAACATTTTTAACCC-3' (SEQ ID NO: 2).

Amplification of the microsatellite marker NR21 may be performed using the forward primer having the sequence 5'-GAGTCGCTGGCACAGTTCTA-3' (SEQ ID NO: 3) and the reverse primer having the sequence 5'-CTGGTCACTCGCGTTTACAA-3' (SEQ ID NO: 4).

Amplification of the microsatellite marker NR27 may be performed using the forward primer having the sequence 5'-AACCATGCTTGCAAACCACT-3' (SEQ ID NO: 5) and the reverse primer having the sequence 5'-CGATAATACTAGCAATGACC-3' (SEQ ID NO: 6).

Amplification of the microsatellite marker NR24 may be performed using the forward primer having the sequence 5'-GCTGAATTTTACCTCCTGAC-3' (SEQ ID NO: 7) and the reverse primer having the sequence 5'-ATTGTGCCATTGCATTCCAA-3' (SEQ ID NO: 8).

Amplification of the microsatellite marker BAT25 may be performed using the forward primer having the sequence 5'-TACCAGGTGGCAAAGGGCA-3' (SEQ ID NO: 9) and the reverse primer having the sequence 5'-TCTGCATTTTAACTATGGCTC-3' (SEQ ID NO:10).

In a particular embodiment, the method further comprises in step b) comparison with a control sample.

As used herein, the control sample may consist of non-cultured cells from the patient likely to be affected by CMMR-D. In a preferred embodiment, said non-cultured cells may be non-cultured peripheral blood cells and/or non-cultured plasma cells. In a particular embodiment, said non-cultured peripheral blood cells and/or non-cultured plasma cells may be non-cultured lymphocytes, and more preferably non-cultured mononuclear lymphocytes. As used herein, non-cultured lymphocytes are lymphocytes, and preferably mononuclear lymphocytes, directly extracted from blood and/or plasma obtained from the patient likely to be affected by CMMR-D. Non-cultured lymphocytes (preferably mononuclear lymphocytes) may be isolated from e.g. the blood or plasma according to known methods e.g. by a Ficoll density gradient centrifugation.

By "instability on at least one microsatellite marker" is meant a deletion or an addition of the repeated pattern of mononucleotides, dinucleotides or trinucleotides.

In an examplary embodiment, instability on the BAT26 microsatellite marker is present when the BAT26 microsatellite consists of a repeated pattern of more than 26A or less than 20A.

In an examplary embodiment, instability on the NR21 microsatellite marker is present when the NR21 microsatellite consists of a repeated pattern of more than 21 T or less than 16T.

In an examplary embodiment, instability on the NR27 microsatellite marker is present when the NR27 microsatellite consists of a repeated pattern of more than 26A or less than 23A.

In an examplary embodiment, instability on the NR24 microsatellite marker is present when the NR24 microsatellite consists of a repeated pattern of more than 23T or less than 18T.

In an examplary embodiment, instability on the BAT25 microsatellite marker is present when the BAT25 microsatellite consists of a repeated pattern of more than 25T or less than 19T. In a particular embodiment, instability is present when a deletion or an addition of at least one nucleotide is detected in the cultured lymphocytes, optionnally in the cultured immortalized lymphocytes, of the patient in comparison to the control sample.

Preferably, instability is present when a deletion or an addition is detected on one allelle, and more preferably on two alleles.

In another particular embodiment, the method comprises a step of determining the resistance or sensitivity of the cultured lymphocytes, optionally cultured immortalized lymphocytes, to DNA-damaging drugs. Said step may be performed simultaneously or consecutively to step b) of the methods as described herein. Accordingly, in one embodiment the method of the invention may comprise the steps of:

a) culturing lymphocytes obtained from a patient likely to be affected by CMMR-D; and

b) detecting the presence or absence of instability on at least one microsatellite marker of CMMR-D in the cultured lymphocytes of step a);

c) determining the resistance or sensitivity of the cultured lymphocytes of step a) to

DNA-damaging drugs,

wherein the detection of instability on at least one microsatellite marker on the cultured lymphocytes of the patient and of the resistance of the cultured lymphocytes of the patient to the DNA-damaging drugs indicates that the patient suffers from CMMR-D.

In another embodiment, the method may comprises the steps of :

aO) immortalizing lymphocytes obtained from a patient likely to be affected by

CMMR-D,

a) culturing the immortalized lymphocytes obtained at step aO); and

b) detecting the presence or absence of instability on at least one microsatellite marker of CMMR-D in the cultured immortalized lymphocytes of step a);

c) determining the resistance or sensitivity of the cultured lymphocytes of step a) to DNA-damaging drugs,

wherein the detection of instability on at least one microsatellite marker in the cultured immortalized lymphocytes of the patient and of the resistance of the cultured immortalized lymphocytes of the patient to the DNA-damaging drugs indicates that the patient suffers from CMMR-D.

According to the method of the invention, in step c) determination of resistance or sensitivity of the cultured lymphocytes, optionally cultured immortalized lymphocytes, is performed by conventional methods. Such conventional methods may include any method for studying cell proliferation and viability known by the skilled in the art. Such method may be performed by measuring the metabolic activity of the cells, for example using WST1 , XTT, or MTT reagents, or by measuring the rate of DNA synthesis, for example using ³ H-thymidine or BrdU. For example, said determination may be performed using the WST1 reagent, as described in example 2. In some embodiments, said determination may include the steps of :

(i) incubating the cultured lymphocytes, or the cultured immortalized lymphocytes, of the patient with the DNA-damaging drug or with a vehicle, (ii) determining the proliferation and/or viability level of the cells of step (i), wherein a decrease of proliferation and/or viability of the cultured lymphocytes, or the cultured immortalized lymphocytes, of the patient incubated with the DNA-damaging drug in comparison to the cultured lymphocytes, or the cultured immortalized lymphocytes, of the patient incubated with the vehicle indicates that the cultured lymphocytes, or the cultured immortalized lymphocytes, of the patient are sensitive to the DNA-damaging drug.

In some embodiments, said determination may include the steps of :

(i) incubating the cultured lymphocytes, or the cultured immortalized lymphocytes, of the patient with the DNA-damaging drug,

(ia) incubating control cells with the DNA-damaging drug,

(ii) determining the proliferation and/or viability level of the cells of step (i) and (ia), wherein an increased proliferation and/or viability of the cultured lymphocytes, or the cultured immortalized lymphocytes, of the patient in comparison to the control cells indicates that the cultured lymphocytes, or the cultured immortalized lymphocytes, of the patient are resistant to the DNA-damaging drug.

Step (i) may be performed e.g. for at least 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, or 15 day(s), and/or at most 15, 14, 13, 12, 1 1 , 10, 9, 8, 7, 6, 5, 4, 3, 2, 1 day(s).

Step (i) may be performed e.g. at a temperature of at least 30, 31 , 32, 33, 34, 35,

36, 37, 38, 39, or 40 ^< Ό, and/or of at most 40, 39, 38, 37, 36, 35, 34, 33, 32, 31 , or 30 ^< Ό.

The DNA-damaging drug of step (i) may have a concentration ranging from e.g. 0.01 to 500 μΜ, preferably ranging from 0.1 to 100 μΜ, more preferably ranging from 0.15 to 50 M. DNA-damaging drug at step a) may have a concentration of e.g. at least 0.05, 0.06, 0.07, 0.08, 0.09, 0.1 , 0.15, 0.2, 0.25, 0.30, 0.35, 0.40, 0.45, 0.5, 1 , 1 .5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 40, 45, 50 μΜ, and/or at most 50, 45, 40, 35, 30, 25, 20, 15, 10, 9, 8 7, 6, 5, 4.5, 4, 3.5, 3, 2.5, 2, 1 .5, 1 , 0.5, 0.,45, 0.4, 0.35, 0.30, 0.25, 0.20, 0.15, 0.1 μΜ.

Step (ii) may be performed : - by adding at least 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50 μΙ and/or at most 50, 45, 40, 35, 30, 25, 20, 15, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1 μΙ of WST1 reagent, and

- by incubating the cells with the WST1 reagent during at least 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, or 15 hour(s), and/or at most 15, 14, 13, 12,

1 1 , 10, 9, 8, 7, 6, 5, 4, 3, 2, 1 hour(s),

- reading the absorbance at 450nm.

Step (ii) may be performed at a temperature of at least 30, 31 , 32, 33, 34, 35, 36, 37, 38, 39, or 40 °C, and/or of at most 40, 39, 38, 37, 36, 35, 34, 33, 32, 31 , or 30°C.

Control cells of step (ia) may include cultured lymphocytes, or immortalized cultured lymphocytes from a control subject (i.e. a subject carrying no mutation in any of the 4 genes MLH1, MSH2, MSH6, PMS2), or from a subject suffering from Lynch syndrome (i.e. a subject carrying dominant autosomal mutation in one of the 4 genes MLH1, MSH2, MSH6, PMS2).

DNA damaging drugs may include a DNA-methylating agent, i.e. an agent that is able to methylate the DNA, or a thiopurine such as described in Karran (2001 ) Carcinogenesis 22:1931 -1937. For example, the methylating agent may be the Methylmethane sulfonate (MMS) (CAS Number 62-27-3), the N-Methyl-N'-nitro-N- nitrosoguanidine (MNNG) (CAS Number 70-25-7), the Temozolomide (CAS Number 85622-93-1 ), the N-Methyl-N-Nitrosourea (NMU) (CAS Number 684-93-5). The thiopurine drug may be, for example, the 6-thioguanine (CAS Number 154-42-7), the 6- mercaptopurine (CAS Number 50-44-2), or the azathioprine (CAS Number 446-86-6).

The invention also relates to a method for identifying a patient requiring monitoring for detection of cancer onset, which method comprises the steps of:

a) in vitro diagnosing if the patient suffers from constitutional mismatch repair- deficiency (CMMR-D) according to the method previously described; and b) identifying the patient diagnosed as suffering from CMMR-D in step a) as a patient requiring monitoring for detection of cancer onset.

The types of cancer likely to be monitored for detection of onset may be for instance, none limitatively, leukemia, lymphoma, glioblastoma, neuroblastoma, astrocytoma and/or colorectal cancer.

The invention is further directed to a method for identifying a patient at risk of developing a tumor, which method comprises the steps of:

a) in vitro diagnosing if the patient suffers from constitutional mismatch repair- deficiency (CMMR-D) according to the method previously described; and b) identifying the patient diagnosed as suffering from CMMR-D in step a) as a patient at risk of developing a tumor.

The "tumor" may be, none limitatively, lymphomas, leukemias, glioblastomas, neuroblastomas, astrocytomas and/or colorectal tumors.

Method for identifying microsatellite markers of constitutional mismatch repair- deficiency

The present invention also relates to a method for identifying microsatellite markers associated with constitutional mismatch repair-deficiency (CMMR-D) comprising the steps of:

a) culturing lymphocytes obtained from a patient affected by CMMR-D;

b) detecting the presence or absence of instability on at least one microsatellite in the cultured lymphocytes of step a); and

c) identifying as a microsatellite marker associated with CMMR-D said microsatellite which displays instability in the cultured lymphocytes of the patient affected by CMMR-D but which does not display instability in non-cultured cells of the patient affected by CMMR-D.

In a particular embodiment, the method comprises a step of immortalization of lymphocytes from a patient affected by CMMR-D prior to step a). Accordingly, the method of the invention may comprise the steps of:

aO) immortalizing lymphocytes obtained from a patient affected by CMMR-D;

a) culturing the immortalized lymphocytes obtained at step aO);

b) detecting the presence or absence of instability on at least one microsatellite in the cultured immortalized lymphocytes of step a); and

c) identifying as a microsatellite marker associated with CMMR-D said microsatellite which displays instability in the cultured immortalized lymphocytes of the patient affected by CMMR-D but which does not display instability in non- cultured cells of the patient affected by CMMR-D.

By "patient affected by CMMR-D" is meant an individual who have bi-allelic germline mutations on at least one MMR genes. More preferably, the patient has bi-allelic germline mutations on MLH1, MSH2, MSH6 or PMS2. As used herein, the patient is a human or a non-human mammal, in particular a rodent, a feline, a canine, a bovine or an ovine. In a preferred embodiment, the patient is a human. More particularly, the patient is a child, an adult, a man or a woman. In a particular preferred embodiment, the patient is a child. As used herein, non-cultured cells may be cells directly extracted from blood or plasma of the patient affected by CMMR-D. In a preferred embodiment, said non-cultured cells may be lymphocytes, and preferably mononuclear lymphocytes, directly extracted from blood or plasma of the patient affected by CMMR-D. Non-cultured lymphocytes, and preferably mononuclear lymphocytes, may be isolated from e.g. the blood or plasma according to known methods e.g. by a Ficoll density gradient centrifugation.

FIGURES

Figure 1 depicts clinical presentation and MMR mutation analysis in CMMR-D family A. A partial pedigree for the family is shown with indications for malignancies (filled symbols) and benign lesions (striped symbols) in probands (arrows) and affected members. Stars indicate members for whom blood sample was available and their genotype is shown in slanting characters. WT: wild type; CRC: colorectal cancer; CLS: cafe-au-lait spots.

Figure 2 depicts clinical presentation and MMR mutation analysis in CMMR-D family B. A partial pedigree for the family is shown with indications for malignancies (filled symbols) and benign lesions (striped symbols) in probands (arrows) and affected members. Stars indicate members for whom blood sample was available and their genotype is shown in slanting characters. WT: wild type; CRC: colorectal cancer; CLS: cafe-au-lait spots.

Figure 3 depicts clinical presentation and MMR mutation analysis in CMMR-D family C. A partial pedigree for the family is shown with indications for malignancies (filled symbols) and benign lesions (striped symbols) in probands (arrows) and affected members. Stars indicate members for whom blood sample was available and their genotype is shown in slanting characters. WT: wild type; CRC: colorectal cancer; CLS: cafe-au-lait spots.

Figure 4 depicts electrophoretograms for the BAT26, NR21 and NR27 alleles from AIM and her parents. Allelic sizes are indicated in each case. Bp: base pair.

Figure 5 depicts quantification of ei/-MSI in CMMR-D families A and B. In each case, a cumulative size index (SI) of the BAT26, NR21 and NR27 alleles was calculated. Box-plots correspond to the SI median and quartile values in patients and their parents (experiments performed in 8-plicate). E1/-MSI is illustrated by the aberrant SI values for AIII3-L and BIII2-L (LBL samples) compared to AIII3 and BIII2 (primary lymphocytes) due to deletions within BAT26, NR21 and/or NR27 mononucleotide repeats. Figure 6 depicts time course analysis of ei/-MSI at BAT26, NR21 and NR27 in

LBLs from AIII3 (AIII3-L) and BIII2 (BIII2-L). In each case, the duration of culture is indicated in days. NC: non-cultured, primary blood DNA sample.

Figure 7 depicts comparative analysis of BAT26, NR21 and NR27 microsatellite profiles in LBLs from CMMR-D, HNPCC and MMR-proficient wild-type healthy subjects. The size index (SI) for BAT26, NR21 and NR27 alleles was calculated in additional series of LBLs from HNPCC (H1 -L to H9-L) and healthy wild-type MMR-proficient controls (C1 -L to C13-L). In the absence of primary blood DNA samples from these subjects, data from a previous study by our group on allele sizes for the BAT26, NR21 and NR27 microsatellites was used. SI values below the red line would not be expected and are therefore explained by ei/-MSI arising after long-term culture of LBLs in CMMR-D patients.

Figure 8 depicts cell survival of lymphoblastoid cell lines with MMR-KO genotype (from CMMR-D patients, circles), MMR+/- (from Lynch patients, square) and MMR-wt (from control subjects, triangle) genotype after 5 μΜ MNNG exposure.

EXAMPLE

Example 1

MATERIAL AND METHODS

Subjects, clinical samples, Lymphoblastoid cell lines (LBLs) and mismatch repair gene mutation analyses in CMMR-D families.

Families A and B were from the Lyon Centre Leon Berard and were previously reported as having bi-allelic mutations in MSH6 and PMS2, respectively. Family C was from the Department of Clinical Genetics, Hospital Robert Debre and La Pitie Salpetriere in Paris. The proband CIV1 of Turkish consanguineous parents exhibited a homozygous deletion of exon 8 in the MSH2 gene as detected by multiplex ligation-dependent probe amplification. All HNPCC and FAP patients were sourced from the Department of Clinical Genetics, La Pitie Salpetriere in Paris. Members of families A, B and C as well as HNPCC and FAP patients provided informed consent for their samples to be used in this study. Wild-type healthy controls were provided by CEPH (Centre d'Etude du Polymorphisme Humain), Paris, France. All immortalized lymphoblastoid cells (LBLs) were obtained following a standard EBV infection procedure (Neitzel (1986), Hum Genet. 73: 320-326). LBLs were maintained in DMEM (Life Technologies) containing 10% FCS (InVitrogen).

Microsatellite instability testing.

Constitutional DNA extracted from peripheral blood cells and cultured DNA from LBLs were amplified by PCR at the BAT26, BAT25, NR21 , NR24 and NR27 alleles, as well as at 40 microsatellite sequences contained within coding regions (Table 1 ).

Name of primer sequence (dye)5'-3' SEQ ID N °

53BP1 -F (hex)ttccgttgcacctttctctgactg 1 1

53BP1 -R ttctccactccctccaaaggttca 12

ATM-F (fam)ccctgtccagactgttagcttctt 13

ATM-R cgcttctcaaaccaaatagctgg at 14

AURORAB-F (hex)tgttgcaaggtgtcttttggtg 15

AURORAB-R atgagacatacctgagagtggagt 16

BARD1 -F (fam)gccctcgaagtaagaaagtcag 17

BARD1 -R tccagatcttgcagaagcct 18

CARD1 0-F (hex)gaggaagccacagacagtgaa 19

CARD1 0-R gtctactgctatccccagccat 20

CARD1 5/NOD2-F (fam)gaagtacatccgcaccgagttcaa 21

CARD1 5/NOD2-R tgggctgagaacacgtagca 22

CARD6-F (hex)tggcctggttgagataacatggtg 23

CARD6-R agtagctgtgctggcttcaagttc 24

CARD8-F (hex)ttcctgcacagcctatgctatc 25

CARD8-R g acacctccatgg aag aaaacctc 26

CASP10-F (hex)gagtgcagttgcgcaatctcg 27

CASP10-R agacttgatgcagggaaggacaga 28

CASP1 -F (hex)gctgtcataaaccaggaagggaa 29

CASP1 -R ggaccatgtagtatccagcactct 30

CASP3-F (fam)tgtgtgcttctgagccatggtg 31

CASP3-R acccctgcttaatcgtcaataacc 32

CASP4-F (fam)gccacttaaggtgttggaatccc 33

CASP4-R agacactgcttaggccttggagtt 34

CASP5-F (fam)tgtgttattcgctggagacatggt 35

CASP5-R cacgcgattctttcgcaaagag 36

CASP8AP2-F (hex)cgtttatttgaacagcaactaccag 37

CASP8AP2-R tgttggagttgtccacactcct 38

CASP8-F (fam)tcaccatcctgactgaagtgaac 39

CASP8-R aggcaggagaatggcctgaa 40

CDC25C-F (fam)agactgaagcaggtggaaaa 41

CDC25C-R acccggcctatgacaacatt 42

CHK1 -F (hex)atgggaccaacccagtgaca 43 CHK1 -R ttgcagagttctgggactaaagag 44

CTIP-F (fam)ctcgagtgttcatctcctgtatttgg 45

CTIP-R ctgtgatgtgtgaaaagggcata 46

FANCJ-F (hex)gccctggatccagacattgaattg 47

FANCJ-R atcctcagctttcacttctctggc 48

MDC1 -F (hex)ttggcagggagaatgagga 49

MDC1 -R gaagactcaggtgtctgact 50

MDM2-F (fam)ggggaaatctctgagaaagccaaac 51

MDM2-R gcattaaggggcaaactagattcca 52

MRE1 1 -F (fam)gcatttcttaattgtagccccttgt 53

MRE1 1 -R atcccctag acctatgg actg act 54

P63-F (hex)ggtgctggtaattaagttactcaag 55

P63-R ggaggttaggctgtttgtatgg 56

PAK3-F (fam)caccctctgctgaaaatgccaa 57

PAK3-R ggcttatgagaaacactcccgtgta 58

RAD1 7-F (fam)ggagaaaacaacttacggccaagga 59

RAD1 7-R ccaccaatagcttggacctcttga 60

RAD51 APT-F (hex)atccagaacagcaccaaaggag 61

RAD51 APT-R gctcctgttttctgactgcaccta 62

RAD54-F (fam)gacaaacaccaatctctcaaaggc 63

RAD54-R ggcattgtatcctcaggcattc 64

RAD9-F (fam)ccagtgctatcattgtcatccca 65

RAD9-R ccactcttttcattgcagggc 66

REV3-F (hex)ctggctgagtttgagggagacttt 67

REV3-R actccagttggcttggttttaggc 68

RF1 -F cctgcggcagtatctcattct 69

RF1 -R cgggccctgggtgtaattagtt 70

RIF1 -F (fam)ttcaaggcgacgttcagaag 71

RIF1 -R acaatgtcctgggtaccatc 72

SG01 -F (fam)agaggggacccttttacagatttg 73

SG01 -R atgagctagggtcctgtcaaga 74

ATR-F tgaaagcaagttttactggactagg 75

ATR-R (hex)tcttctgtaggaacttgaaagcc 76

BAX-F agttcgtccccgatgcgcttgaga 77

BAX-R (fam)ttcatccaggatcgagcagggcga 78

BLM-F acagcagtgcttgtgagaac 79

BLM-R (fam)ctctgccaccaggaagaatc 80

CBF2-F tcccttactttgtcatcacc 81

CBF2-R (hex)ccatgaagaaagtgaattgg 82

MSH3-F actcccacaatgccaataaaaa 83

MSH3-R ( h ex) ag atg tg aatcccctaatcaagc 84

MSH6-F cgtaatgcaaggatggcgtg 85

MSH6-R (hex)gggtgatggtcctatgtgtc 86

TGFBR2-F cacatgaagaaagtctcaccaggc 87 TGFBR2-R (fam)ctttattctggaagatgctg 88

RAD50-F caagtcccagcatttcatca 89

RAD50-R (hex)aactgcgacttgctccagat 90

TFDP2-F attcctgagcagaattggta 91

TFDP2-R (hex)cagaagaacattaggcgaag 92

GRB14-F gcctgtaccttaaagcagaa 93

GRB14-R (hex)ggcaatagtgatatttatgt 94

GRK4-F tttagagccatagcttcacc 95

GRK4-R (fam)tgctgtttaaactaggtttgc 96

RBBP8-F ccaagactgtgatgtgtgaa 97

RBBP8-R (hex)gtgtcatctcctgtatttgga 98

MBD4-F cag aacaaaaatttg atcctg aactc 99

MBD4-R (fam)gatgctggagcatgtggt 100

Table 1 . List of forward and reverse primers used for amplifying 40 microsatellite sequences contained within coding regions.

Amplification of the microsatellite marker BAT26 is performed using the forward primer having the sequence 5'-CTGCGGTAATCAAGTTTTTAG-3' (SEQ ID NO: 1 ) and the reverse primer having the sequence 5'-AACCATTCAACATTTTTAACCC-3' (SEQ ID NO: 2). Amplification of the microsatellite marker NR21 is performed using the forward primer having the sequence 5'-GAGTCGCTGGCACAGTTCTA-3' (SEQ ID NO: 3) and the reverse primer having the sequence 5'-CTGGTCACTCGCGTTTACAA-3' (SEQ ID NO: 4). Amplification of the microsatellite marker NR27 is performed using the forward primer having the sequence 5'-AACCATGCTTGCAAACCACT-3' (SEQ ID NO: 5) and the reverse primer having the sequence 5'-CGATAATACTAGCAATGACC-3' (SEQ ID NO: 6). Amplification of the microsatellite marker NR24 is performed using the forward primer having the sequence 5'-GCTGAATTTTACCTCCTGAC-3' (SEQ ID NO: 7) and the reverse primer having the sequence 5'-ATTGTGCCATTGCATTCCAA-3' (SEQ ID NO: 8). Amplification of the microsatellite marker BAT25 is performed using the forward primer having the sequence 5'-TACCAGGTGGCAAAGGGCA-3' (SEQ ID NO: 9) and the reverse primer having the sequence 5'-TCTGCATTTTAACTATGGCTC-3' (SEQ ID NO:10).

PCR products were separated by capillary electrophoresis on an ABI 3100 genetic analyser and analyzed using GeneMapper software v3.7. Experiments with BAT26, BAT- 25, NR21 , NR-24 and NR27 were performed in 8-plicate to allow true instability at these microsatellite loci to be distinguished from background.

RESULTS AND DISCUSSION

Samples from 4 CMMR-D index cases in 3 distinct families (A, B, C) were collected, together with samples from their first degree relatives. Families A and B display previously reported bi-allelic mutations in MSH6 and PMS2, respectively. In family C, the proband of consanguineous parents exhibited a homozygous deletion of exon 8 in the MSH2 gene. Details concerning the clinical features and mutations in these families are shown in Figure 1 to 3. We established LBLs of CMMR-D probands following a standard EBV infection procedure. MMR wild-type subjects with familial adenomatous polyposis (FAP) due to germline APC mutation were used as controls, together with HNPCC patients consulted in our hospital. To detect MSI, we used a panel of 5 non-coding microsatellites (BAT25, BAT26, NR21 , NR24, and NR27) in which instability is almost systematically observed in MMR-deficient tumors.

We observed aberrant BAT26, NR21 and NR27 allelic profiles due to MSI in the cultured LBLs of 3 CMMR-D patients (AIII3-L, BIII2-L and BIII4-L; Figures 4 and 5). MSI progressively increased at these markers with longer duration of culture under standard conditions (Figure 6).

In contrast, all of the control and HNPCC cases displayed stable allelic profiles in culture with similar BAT26, NR21 and NR27 alleles in primary lymphocytes and related LBLs (Figure 5).

We also investigated MSI in microsatellite sequences contained within coding regions (Table 2).

CCKOOOSSH _EPI NT A _PP T _I Total Total

Total DNA LBL Sub-clones DNA DNA

RELATED

GENES MSH67- PMS2-/- PMS2-/- Λ/ = 8 N = 2 PATHWAYS

AIII3-L BIII2-L BIII4-L AIII3-L BIII2-L

53BP1 wt wt wt wt wt

ATM wt wt wt wt wt AURORA B wt wt na wt wt BARD1 wt wt wt wt wt CDC25 na wt wt wt wt

CHK1 na wt na wt na

CTIP wt wt wt wt wt FANCJ wt wt wt wt wt

MDC1 na wt wt wt wt

MDM2 wt wt na wt wt

PAK3 wt wt wt wt wt

RAD9 wt wt wt wt wt RBBP8 wt wt wt wt wt

CASP1 wt wt wt wt wt CASP3 wt wt wt wt wt CASP4 wt wt wt wt wt CASP5 wt wt na wt na CASP8 wt wt wt wt wt CASP82AP2 wt wt na wt wt CASP10 wt wt na wt wt CARD6 wt wt na wt wt CARD8 wt wt na wt na CARD 10 wt wt na wt wt CARD15 wt wt na wt wt RF1 wt wt wt wt na BAX wt wt wt wt wt/+1 (1 )

MRE11 wt wt wt wt wt

P63 wt wt wt wt wt

REV3 wt wt wt wt wt cc

< RAD50 wt wt na wt na

Q_

LU RAD51AP wt wt wt wt wt

CC

<f RAD54 wt wt na wt wt z _Q MSH3 wt wt wt wt wt

MSH6 wt/-1 wt wt wt/-1 (8) wt

BLM wt na wt wt/+1 (1 ) wt

MBD4 wt wt wt wt wt Table 2. Screening for MSI in 40 repetitive sequences contained within coding regions. These microsatellites were found not to be mutated in the blood, although sub-clones with heterozygous mutations in BAX (N = 1 ), BUM (N = 1 ), TFDP2 (N = 1 ), GRK4 (N = 2) and/or TGFBR2 (N = 1 ) were found. One sub-clone was found with a homozygous mutation in TGFBR2 (N = 1 ). As expected, all sub-clones (N = 8) from AIII3 showed a heterozygous MSH6 mutation, representing one of the two germline mutations in this family. These results indicate mosaicism for MSI in repetitive sequences from coding regions in CMMR-D LBLs. Whether these mutational events also occur in vivo is not known.

Since normal blood lymphocytes are heterogeneous, we screened LBL sub-clones from AIII3-L and BIII2-L to improve the detection of putative microsatellite alterations in these coding sequences. We could not detect MSI in the coding sequences of LBLs from CMMR-D patients, but did find rare sub-clones with alterations (Table 2). Due to the low frequencies, this was not considered to be of interest for the diagnosis of CMMR-D.

Overall, we concluded that ex-vivo acquisition of the MSI phenotype in LBLs from CMMR-D patients was readily detectable by amplifying consensual, non-coding microsatellite markers such as BAT26, NR21 and NR27. We have termed this 'ei/-MSI'. We further characterized the specificity of ei/-MSI in CMMR-D individuals by collecting additional, independent LBL controls from MMR-proficient, wild-type healthy subjects, together with additional HNPCC cases from other clinical centers in France. Due to the lack of primary blood DNA samples for these controls, we used as a reference the polymorphisms of BAT26, NR21 and NR27 amplicons previously obtained by our group in a study of 1 ,206 wild-type healthy individuals encompassing 55 different populations worldwide (Buhard et al. (2006) J. Clin. Oncol. 24 : 241 -51 ). In AIII3-L, BIII2-L and BIII4-L, the BAT26, NR21 and NR27 markers displayed aberrant, shortened profiles that fell outside the polymorphic zone defined by our previous study (Figure 7). In contrast, the size of these markers for LBLs from HNPCC and healthy subjects was within the polymorphic zone, demonstrating that ei/-MSI was indeed a specific feature of CMMR-D (Figure 7).

To date, CMMR-D has been diagnosed in more than 90 patients from at least 54 families. The epidemiology of CMMR-D is characterized by an unusually low incidence of HNPCC in first degree relatives. Because of this apparent sporadic presentation of tumors, the incidence of CMMR-D could be significantly underestimated in children who develop tumors. We have shown here that MSI is not detectable in the normal blood of either HNPCC or CMMR-D patients, but can be unmasked by culture of LBLs from CMMR-D but not HNPCC patients, presumably due to the presence of a functional MMR allele in the latter. These findings demonstrate that routine detection of CMMR-D in at-risk individuals can be achieved by immortalizing LBLs and then evaluating the size of specific microsatellites after a period of culture.

This approach is highly specific since we did not detect ei/-MSI in any of the MMR- proficient controls. Our experience in genotyping several MMR-proficient cancer cell lines at the BAT26, NR21 , NR27 and other microsatellite loci corroborate this observation: no changes have been noted in the profile of these markers after many passages in culture. We therefore speculate that MSI occurs specifically in CMMR-D LBLs because the additional cell divisions during culture lead to the accumulation of unrepaired replication errors at mononucleotide repeats. In contrast, the same DNA alterations are recognized and repaired in MMR-proficient LBLs.

Our method seems to be highly sensitive, and after 90 days of culture, we observed a clear ev-MS\ phenotype in LBLs from AIII3 (AIII3-L), BIII2 (BIII2-L) and BIII4 (BIII4-L). The difference in allele size compared to wild-type, as measured by the SI index (Figure 6), indicates that CMMR-D could well be diagnosed after a shorter culture period of one month or less.

However, due to the quasi-monomorphic nature and well-established allelic profiles of BAT26, NR21 , NR27, analysis of LBLs alone is sufficient to detect ev-MS\ that manifests beyond the polymorphic zone in cases where normal blood sample is not available.

Finally, ei/-MSI was detected in LBLs from patients with bi-allelic MSH6 or PMS2 germline mutations. Importantly, inactivation of these MMR genes in tumors has been associated with a 'mild' MSI phenotype characterized by short deletions of DNA repetitive sequences. Since most CMMR-D patients reported so far have been diagnosed with bi- allelic PMS2 mutations, this indicates that detection of ei/-MSI in LBLs may be sensitive enough to diagnose all CMMR-D cases including MSH6- and PMS2-deficient patients. Similar to sporadic MSI tumors, bi-allelic defects in MLH1 or MSH2 would be expected to show higher levels of ei/-MSI in LBLs and therefore be more readily detectable.

In conclusion, we report ei/-MSI as a novel phenomenon that occurs specifically in LBLs from CMMR-D patients. The clinical interest of our finding is that it provides an opportunity for large-scale screening of this hereditary tumor syndrome, thus allowing the incidence of CMMR-D to be evaluated in humans. As an initial clinical indication for our method, we suggest the systematic screening of children with early malignancies and/or multiple, atypical cafe-au-lait spots, observed in almost all CMMR-D patients. This should occur even in the absence of a familial history of Lynch syndrome. Our findings could greatly improve the clinical surveillance and genetic counseling of CMMR-D families. It is worth noting that neither family A or B fulfilled either the Amsterdam or Bethesda criteria. Improved detection of CMMR-D in children could also contribute to the early diagnosis of HNPCC in their parents before they develop MSI colorectal cancers or other Lynch- related tumors. This is of particular interest for the detection of PMS2 ^+/" carriers that may currently be under-diagnosed in the population due to the development of late onset tumors after the age of 60 years. Finally, because CMMR-D represents a severe hereditary cancer syndrome, our method could in future be applied for prenatal diagnosis in at-risk families. Example 2

MATERIAL AND METHODS

The lymphoblastoid cell lines were established at the tumor bank of Cochin hospital. Cells were suspended in complete medium at appropriate density and dispensed in 100-ul aliquots into 96-well plates. After 24h, cells were either treated with 0.15- 5μΜ MNNG and incubated for 9 days at 37°C, or treated with 0.15-20μΜ 6-thioguanine for 24h, rinsed and incubated in complete medium for 6 more days at 37°C. All steps of MNNG treatment were performed in presence of 06-benzylguanine (final concentration of 20 μΜ) and 1 , 2 or 3 pulses of MNNG treatment were performed. Control groups included cells incubated in complete medium without drug. The effect of the drug on the cell lines was evaluated by the WST1 method (Roche diagnostic). Ten microliters of the WST1 stock solution was added, and the cells were incubated at 37°C for an additional 3 h and the absorbance was read at 450 nm. All samples were tested at least in quadruplicate and at least 2 independent experiments were performed. RESULTS AND DISCUSSION

In order to facilitate the diagnosis of CMMR-D patients and in addition to the ev-MSI method described, the development of a functional method based on the resistance of MMR deficient cells (MMR-KO cells in which one of the MMR genes, such as MLH1 , MH2 or MSH6, is inactivated) to particular drugs was provided.

It has been demonstrated that functional MMR is required for the lethality of certain DNA lesions and that its inactivation is accompanied by tolerance to the cytotoxic effect of some drugs. This is a widely acknowledged mechanism underlying resistance to methylating agents and to thiopurine drugs which produce structurally similar types of DNA damage. Tolerance of MMR-KO cells, to thiopurines and methylating agents has been demonstrated primarily with epithelial cells. Here, lymphoblastoid cell lines (LLB) that could be easily established in culture from a blood sample were used. Two drugs that induce DNA damages supported by the MMR system, i.e. 6-thioguanine, a thiopurine drug, and N-Methyl-N'-nitro-N-nitrosoguanidine (MNNG) a methylating agent were used. The impact of these drugs on cell survival is evaluated by proliferation assays (WST1 test and reading of the absorbance spectroscopy). The thiopurine/methylation tolerant phenotype of MMR-KO cells was first validated, whatever the MMR gene that was inactivated (MLH1 , MH2 or MSH6), in a model of epithelial cells using colon cancer cell lines. Then, the experiments were started on a collection of LLB from control subjects (MMR-wt, n = 20), from Lynch patients (MMR+/-, n = 20) and from CMMR-D patients (MMR-KO, n = 6) through the establishment of a collaborative network including the Centre for the Study of Human Polymorphism and the oncogenetic consultation of the Saint-Antoine hospital. Results show that under certain experimental conditions, methylation tolerance is clearly observed for the LLB issued from CMMR-D patients (average survival rate of 90% for MMR-KO LLB vs 40% for MMR-proficient (MMR+/- or MMRwt) LLB, p <0.0001 Student's t test) which supports the possibility to develop a screening test for CMMR-D patients (Figure 8).

Validation of these results on a larger number of lymphoblastoid cell lines is performed. Furthermore, thesec preliminary results can also indicate that MMR+/- lines present an intermediate drug response under other experimental conditions which allow the development of a test for the diagnosis of Lynch patients bearing an heterozygous MMR gene mutation.