Technical field of the invention

The present invention relates to methods for determining the status in a cell. In particular the present invention relates to methods for determining whether a cell population will undergo replication catastrophe in response to a cancer therapy as well as to methods of treatment.

Background of the invention

Cancer represents a complex disease with complex etiologies and is a growing health care concern, as an aging population and lifestyle changes globally lead to increased cancer rates. No single treatment is effective against the various cancers, and the same treatment will even vary in efficacy depending on the individual. Therefore, methods of evaluating the efficacy of cancer therapy, with the potential to adapt methods of treatment to be the most effective for an individual, are highly desired. Summary of the invention

The present invention relates to novel methods for detecting a cells sensitivity to chemotherapy and to methods of diagnosis and/or treatment.

Thus, an object of the present invention relates to give tools which enable to distinguish non-responsive cells from responsive cells and to identify new treatments which may be more effective.

Thus, one aspect of the invention relates to a method for identifying and quantifying the number of cells in a population that are in replication catastrophe state, comprising

a) Providing a cell population

b) Labelling DNA-bound RPA (DB-RPA)

c) Labelling a marker for DNA damage, such as double strand breaks (DSB)

d) Labelling the nucleus e) Quantifying the DNA bound-RPA and the DNA damage marker, such as double strand break (DSB) marker, in the nuclei of the cells of the population by

/. Acquiring images of the cell population,

//. Processing the acquired images by software-based image analysis,

// ^' / ^' . Quantifying the signal from the labelled DB-RPA from b ) and labelled DNA damage marker, such as DNA double strand break marker from c),

iv. Determining the level of the total cell population that displays DNA damage marker, such as DNA double strand break marker and that at the same time also displays maximum levels of DB-RPA, a state defined as replication catastrophe.

Further aspects of the invention are described, and relate to methods of selecting a cancer therapy and methods of treatment of cancer, as well as methods for monitoring cancer therapy.

A further aspect of the invention relates to a method of classifying a compound of interest, comprising quantifying the amount of DNA-bound RPA in a cell nucleus and correlating said amount to the amount of DNA damage, such as DSB, in the same cell nucleus, for the cells in a cell population that has been exposed to said compound of interest; and comparing the resulting correlated data for the compound of interest with correlated data from one or more reference

compounds. Brief description of the figures

Figure 1. ATR inhibition triggers hyper-loading of RPA on S-phase chromatin followed by fork breakage.

(A) Quantitative Image-Based Cytometry single-cell Analysis (QIBC) of immunolabeled U-2-0-S cells. Asynchronous cells were treated with HU (2mM) and ATRi (2microM) for the indicated times and immunostained. Mean nuclear intensities for DAPI, RPA1 (DNA-bound fraction, CB) and gamma-H2AX were determined for each of >5000 individual cells and plotted in a scatter diagram. S- phase cells, with chromatin-loaded RPAl, are labeled in blue.

(B) Right: Mean RPAl and DAPI values from (A) are plotted in a scatter diagram. The blue line indicates the threshold used to discriminate S-phase from Gl and G2 cells based on chromatin-loaded RPAl levels. Left: Average intensities of chromatin-loaded RPA in S-phase cells from the scatter diagram.

Figure 2. Fork breakage occurs when ssDNA exhausts nuclear RPA

(A) QIBC of U-2-OS cells. Mean gamma-H2AXand RPA fluorescence intensities from Figure 1(D, E) are plotted in a scatter diagram.. Untreated cells are depicted in grey.

(B) QIBC of U-2-OS cells treated for the indicated times with HU (2mM) and ATRi (2microM) and immunostained for gamma-H2AXand RPAl after pre- extraction. Mean values are plotted in a scatter diagram. Percentages of ATM- dependent gamma-H2AX-positive cells (containing DSBs) were calculated.

(C) QIBC of the BJ primary fibroblasts exposed to HU and ATRi as indicated, and immunostained for gamma-H2AXand RPAl after pre-extraction. Percentages of cells with DSBs were calculated as in (B).

Figure 3. By limiting origin firing ATR delays exhaustion of RPA and global breakage of active forks

(A) U-2-OS cells treated with the general CDK inhibitor roscovitine (20microM), CDC7 inhibitor (20microM), or transfected with CDC45 siRNA (indicated

concentrations, 72h) were incubated with HU (2mM), ATRi (2microM) for 80 min and analyzed by QIBC.

(B) Quantification of QIBC plots from (A). Right, average values for RPAl in S- phase cells are shown.

Figure4.DNA breakage in replication factories marks a "point of no return

(A) A representative QIBC scatter plot showing the average fraction of cells containing broken forks.

(B) Schematic depiction of the 'point of no return' concept. QIBC from U-2-OS cells treated for 80 minutes with HU (2mM) and ATRi (2microM) and immunostained for chromatin-loaded RPA and gammaH2AX is used as an example. The concept posits that cells that exhaust their RPA pool trespass a point of no return, which is associated with DNA breakage in all active replication factories and thereby results in a permanent proliferation arrest even if the stress insults are removed. Equally important for the concept is the premise that as long as cells keep their RPA surplus, stalled forks remain protected against breakage, ensuring the integrity of the affected loci after reentry into the cell cycle.

Figure 5. RPA exhaustion is an obligatory step before fork breakage

(A) U-2-OS cells were transfected with the indicated concentrations of siRNA against RPA1 for 2 days, treated with HU as indicated, pre-extracted,

immunostained with gammaH2AX and RPA1 antibodies, and analyzed by QIBC. Arrows mark cells with H2AX hyperphosphorylation corresponding to RC.

(B) U-2-OS cells were transfected as in (A), treated with ATRi as indicated, and analyzed as in (A).

Figure 6. RPA exhaustion is an obligatory step before fork breakage

U-2-OS cells were incubated with the drugs (aphidicolin 20microM, cytarabine 50 microM, gemcitabine 1 microM) or exposed to UV (20 J/m2) as indicated, treated or not with ATRi, incubated for indicated times, and analyzed as in (A).

Figure 7

Shows cells behaviour in normal and under replication stress.

Figure 8

Schematic overview of the potential of drug combinations

Figure 9

A: Staining of cells with markers

B Shows QIBC data as scatter diagram, and the amount of cells in replication catastrophe.

1= ssDNA increases and RPA loads onto chromatin (DNA)

2= point of RPA exhaustion

3= Replication catastrophe occurs after RPA exhaustion Figure 10

Shows an illustration of the shortcomings of Western Blot (hypothetical example). Figure 11

Shows reference scatter diagrams for three drugs from different classes. A shows the scatter diagram resulting from incubation with HU and ATRinhibitor

(Replication catastrophe). B and C show the scatter diagrams resulting from incubation with a drug A and and a drug B respectively. As can be seen, the scatter diagrams are distinct both from that in A, as well as from each other. See also Example 13.

Figure 12

Figure 12 shows the scatter diagrams for three drugs in the same class. A shows the scatter diagram for the reference compound (Reference scatter diagram), which is drug A (see also Figure 11 B). B and C show the scatter diagrams for compounds X and Y respectively. As can be seen, compounds X and Y are in the same class as Drug A. See also Example 13.

Detailed description of the invention

Definitions

Prior to discussing the present invention in further details, the following terms and conventions will first be defined :

"Replication Protein A" (RPA) is a protein complex made up of three subunits:

1. RPA1 (also known as RPA 70; Homo Sapiens Gene no: 6117; Protein

accession number P27694; OMIM number 179835),

2. RPA2 (also known as RPA32, Homo Sapiens Gene no: 6118) and

3. RPA3 (also known as RPA14, Homo Sapiens Gene no: 6119).

The term Replication Protein A (RPA) refers to the intact complex unless otherwise stated. RPA binds to ssDNA generated in replication and protects it from

breakage.

The term "Tumours" as used in the context of the present invention shall refer to any kind of tumour such as e.g. benign and malignant tumors, carcinomas, sarcomas, leukemias, lymhomas, carcinomas in situ, neoplasias, dysplasias, etc. Tumours may comprise tumours of the head and the neck, tumours of the respiratory tract, tumours of the gastrointestinal tract, tumours of the urinary system, tumours of the reproductive system, tumors of the endocrine system, tumors of the central and peripheral nervous system, tumors of the skin and its appendages, tumors of the soft tissues and bones, tumors of the lymphopoietic and hematopoietic system, breast cancer, colorectal cancer, anogenital cancer etc.

The term "cancer" as used herein refers to a group of diseases involving abnormal cell growth with the potential to invade or spread to other parts of the body. The term "cancer" as used in the context of the present application shall refer to cancers of any kind and origin and precursor stages thereof, respectively, including benign tumours.

The term "drug" refers to a chemical compound which is intended to be used as a pharmaceutical and/or reagent in academic or industrial research activities, and which has a physiological effect on the body when ingested.

The term "scatter plot" is well known in the art and refers to a useful summary of a set of bivariate data (two variables). The term is used interchangeably with "scatter diagram", both terms referring to the same concept.

As used herein, the words "comprises", "comprising", and similar words, are not to be interpreted in an exclusive or exhaustive sense. In other words, they are intended to mean "including, but not limited to".

In contrast, the word "consists", "consisting", and are intended to mean

"including, and limited to".

The present invention will now be described in more detail in the following.

1. Method for measuring the number of cells in a population with Replication Catastrophe

Replication catastrophe is characterized by the exhaustion of the RPA and the subsequent accumulation of DNA damage (such as DNA double strand breaks) in the cell. In order to determine the proportion of cells in a population that are undergoing replication catastrophe, one aspect of the invention relates to where the level of DB-RPA (DNA-bound RPA, DB-RPA) and the level of DNA damage are assessed. DNA damage may be in the form of DNA breaks alone (double stranded, DSB, or single stranded, SSB) or breaks together with single stranded DNA (ssDNA).

RPA is found in the nucleus of the cell, in either free, soluble form (referred to as free or nucleoplasms RPA), or bound to DNA (DB-RPA). DB-RPA protects single stranded DNA (ssDNA) generated during e.g. DNA replication, DNA repair or transcription.

During replication, ssDNA is generated in great numbers in conditions of replication stress and therefore the proportion of DB-RPA increases. However, at a certain level of ssDNA, the pool of free RPA is exhausted and no more complex can be supplied as DB-RPA. SsDNA generated beyond this point will not be protected, and thus will be more susceptible to double-strand breakage (DSB). The cell will then progress to cell death or cell senescence (an irreversible arrest of cell proliferation) caused by accumulation of DNA damage. This state where free RPA is exhausted and, as a consequence, DNA damage such as DNA double strand breaks accumulates resulting from unprotected ssDNA, is referred to herein as Replication Catastrophe (RC).

By assessing the level of DB-RPA and the appearance of RC in cells in a

population, it can be seen whether cells have exhausted the supply of free RPA, as well as whether this has occurred in all cells or only a subset of cells. When there is still a reserve of free RPA that can be recruited, the amount of DB-RPA can still increase over time and RC is prevented. Beyond this point, when the amount of DB-RPA can no longer increase, the cells enter Replication Catastrophe.

In one wording, the term Replication Catastrophe may be explained as the situation in the cell where the pool of free RPA is exhausted.

In another wording, the term Replication Catastrophe may be explained as the situation in a cell, where the pool of free RPA is exhausted and DNA damage, such as DNA double strand breaks, accumulate.

In another wording, the term Replication Catastrophe may be explained as the situation in a cell, where the pool of free RPA is exhausted, and DNA damage, such as DNA double strand breaks, accumulate beyond a threshold value.

Threshold values may be established by comparison with controls, such as negative controls (for example untreated cells) and/or positive controls. Thus, the assessment of whether cells are in RC or not may comprise comparing the levels of for example one or more of moieties selected from DB-RPA and/or DNA damage such as DSB with a threshold level for the selected moieties, and if the percentage is lower than said threshold, concluding that the cells are not in RC; and if the percentage is equal to or higher than said threshold, concluding that the cells are in RC. Threshold values may be established by comparison with controls, such as negative controls (for example untreated cells) and/or positive controls.

In depth explanation of the term Replication Catastrophe is given in the article "ATR Prohibits Replication Catastrophe by Preventing Global Exhaustion of RPA" Cell Volume 155, Issue 5, 21 November 2013, Pages 1088-1103.

The DNA damage happening in cells which have exhausted the pool of free RPA (i.e. no longer display increasing amounts of DB-RPA) is irreversible and leads ultimately to cell death or cell senescence. Thus, the invention in another aspect relates to a method for identifying and quantifying the number of cells in a population that are in replication catastrophe state, comprising

a) Providing a cell population

b) Labelling DNA-bound RPA (DB-RPA)

c) Labelling a marker for DNA damage, such as double strand breaks (DSB) marker d) Labelling the nucleus e) Quantifying the DNA bound-RPA and the DNA damage marker, such as double strand break (DSB) marker, in the nuclei of the cells of the population by

/. Acquiring images of the cell population,

//. Processing the acquired images by software-based image analysis,

///. Quantifying the signal from the labelled DB-RPA from b ) and labelled DNA damage marker, such as DNA double strand break marker from c), iv. Determining the level (for example the percentage) of the total cell population that displays DNA damage marker, such as DNA double strand break marker and that at the same time also displays maximum levels of DB-RPA, a state defined as replication catastrophe.

In one embodiment, the invention relates to a method for identifying and quantifying the number of cells in a population that are in replication catastrophe state, comprising

a) Providing a cell population

b) Removing free RPA from the cells in the cell population and then fixing the cells

c) Labelling the remaining RPA, being DNA-bound RPA

d) Labelling a marker for DNA damage, such as DNA double strand

breaks (DSB)

e) Labelling the nucleus

f) Quantifying the DNA bound-RPA and the DNA double strand breaks in the nuclei of the cells of the population by

/. Acquiring images of the cell population,

//. Processing the acquired images by software-based image analysis,

// ^' / ^' . Quantifying the signal from the labelled RPA from c ) and labelled DNA double strand break marker from d),

iv. Determining the level, (for example the percentage) of the total cell population that displays DNA double strand break marker and that at the same time also displays maximum levels of DB-RPA, a state defined as replication catastrophe.

Particular embodiments of the methods according to the invention relate to where the assa is an in vitro assay.

Further, it should be noted that the steps of labelling DB-RPA, DNA damage marker and nucleus may be performed in any order, or simultaneously. Cell population

The invention in this embodiment describes the level of DB-RPA in each cell of a population.

In a particular embodiment, the number of cells corresponds to a number where cells will be in exponential growth at the time of performing the method, and will depend on the cell line used. For U20S (human bone osteosarcoma) cells, 20000 cells can be are seeded per well in 100 ul medium.

In some embodiments the cell population according to step a) comprises for example 200 cells or more, 500 cells or more, 1000 cells or more, such as for example 1500 or more, 2000 or more, 3000 or more, 4000 or more, 5000 or more, 6000 or more, 7000 or more, 8000 or more, 9000 or more, or 10000 or more cells. In some embodiments, the cell population has from 200 to 20000 cells, such as from 200 to 15000, 200 to 10000, or for example from 500 to 20000, 800 to 20000, 1000 to 20000, 2000 to 20000 cells; or for example 2000 to 15000, 2000 to 12000, 2000 to 10000 cells. In other embodiments, the cell population may be from 2000 to 500000 cells, for example 5000 to 4000000 cells.

In particular embodiments, the cell population has no less than 200 and no more than 10000 cells, such as no less than 500 and no more than 10000 cells.

In one embodiment of the method, the quantifying step (step f or e) includes measuring labelled DNA-bound RPA in foci of the cells of the population, for example in 100000 foci or more, or such as 250000 foci or more, 300000 or more, 400000 or more, 500000 or more, 600000 or more, 700000 or more, 800000 or more, or 100000 or more foci. In particular embodiments, the quantifying step (step f) or e)) includes measuring labelled DB-RPA in from 50000 to 200000 foci in cells of the population.

The method of the invention gives teaching both on the number of cells which have DB-RPA, as well as the individual level of DB-RPA in those cells. This data can give information on cells with all levels of DB-RPA, like the percentage of cells in a population that has undergone RC, but also on the percentage of cells in the population that has a certain level of DB-RPA and but has not reached RPA exhaustion (i.e., no remaining free RPA).

This teaching cannot be gained for example from a Western blot, where all cells are extracted and pooled. Such a blot would only give the average for the whole population, and would not indicate if there were a sub-set of cells nearing RPA exhaustion (see for example Figure 10).

The cell population which is provided may be any mammalian cell population of interest. For example, the cell population may comprise or consist of mammalian cells, such as rat, mouse, human, dog, cat, horse etc. cells. In one particular embodiment, the cell population consists of human cells.

The cell population may comprise or consist of one or more primary cells, i.e. cells which are non-immortalised. The primary cells are directly obtained from a mammal.

The primary cells may in one embodiment be primary tumour cells. In this context, primary tumour cells means cells directly obtained from a tumour or cancer in a mammal. A primary tumour cell is not considered a cell line in the context of this application.

The primary cells, such as primary tumour cells, may have been grown or expanded in vitro prior to performing the method, for example for 1 to 8 passages, such as for 1 to 6 passages, or for example 1 to 4 passages.

In some embodiments the cell population may comprise or consist of cells from one or more cell lines. A cell line means a cell that has been immortalised.

The cell population that is provided may also be any combination of the above- described cell populations.

In a particular embodiment, the cell population that is provided comprises or consists of human cells. In a further particular embodiment, the cell population that is provided comprises or consists of human cells obtained from a primary tumour, which cells may have been expanded in vitro.

Treated cell population

In one embodiment, the cell population which is provided is treated with a cancer therapy. Examples of cancer therapy are: chemotherapy, radiation therapy and combinations thereof. Particular embodiments of the invention relate to where cancer therapy is a chemotherapy. Chemotherapy as used herein refers to a category of cancer treatment that uses chemical substances, especially one or more anti-cancer compounds. Particular embodiments of chemotherapy comprises compounds which inhibit ATR and/or CHK1 signalling.

In one embodiment, the cell population which is provided comprises or consists of primary cells and/or primary tumour cells, obtained from a mammal that is treated with a cancer therapy prior to the cell population being obtained.

In one embodiment, the cell population which is provided comprises or consists of cells from one or more cell lines, wherein said cell population is treated with a cancer therapy.

Cell culture

The cells may be adherent cells or non-adherent cells. Adherent cells are cultured according to conventional practice in cell culture dishes or the like. Non-adherent cells may be cultured and stained in suspension (see Example 12), and later transferred to for example imaging plates or fixed on microscopy slides for image acquisition.

Removal of free RPA and fixing of cells

RPA is present in the nucleus of the cell, and may be present as free

nucleoplasms RPA or as DNA-bound RPA (DB-RPA). DB-RPA protects single- stranded DNA (ssDNA) generated e.g. at sites of replication. The method of the invention requires the measurement of the amount of DNA-bound RPA. Any method in which the DNA-bound RPA can be distinguished from the free RPA can be used. For example, methods where the DNA-bound RPA is separated from non- DNA-bound RPA so the DNA-bound fraction can be specifically labelled and quantified may be used in the methods of the present invention.

In one embodiment, the method of the invention includes removal of free nucleoplasm^ RPA before fixing and before labelling.

Thus, on one embodiment of the invention, the labelling of DNA bound RPA comprises removing free RPA from the cells in the cell population and fixing the cells, and then labelling the remaining RPA, being DNA-bound RPA. In one embodiment the removal is done by permeabilization of the nucleus and cell membranes and a subsequent wash step, resulting in the removal of the free RPA. The remaining RPA is RPA which is bound to DNA, i.e. DB-RPA. The permeabilization may be done by an incubation with a mild lysis buffer. The mild lysis buffer aims to perforate the membranes in order to allow the removal of free nucleoplasms RPA. The mild lysis buffer may comprise one or more detergents which will perforate the membranes. Mild lysis buffers typically use non-ionic detergents, for example Triton X-100, NP-40, Brij 35, n-Dodecyl-beta- D-Maltoside or combinations thereof. In one embodiment, permeabilization is done by incubating the cell population with buffered solution such as for example Phosphate-buffered Saline containing a non-ionic detergent, such as for example Triton X-100; at a temperature from 0°C to about 10°C, such as from 2 to 4°C; for a period of no more than 10 minutes, such as no more than 7 minutes, no more than 5 minutes, no more than 2 minutes, such as about 1 minute. In particular embodiments, the PBS contains Triton X-100 in an amount of no more than 1% (vol/vol), such as no more than 0.8 %, no more than 0.5% such as about 0.2%, all percentages vol/vol.

After removal of free RPA, fixing is done. Fixing leads to the immobilization of the remaining proteins and cell structures. Fixing may for example be done by incubating the cell population with 4% formaldehyde at room temperature for an amount of time from 5 minutes to 2 hours.

Labelling of DB-RPA

The labelling of the remaining RPA in the cells of the cell population, being DB- RPA, may be done in any way which is compatible with the later step of high content microscopy which is used for acquisition of images. In particular embodiments, DB-RPA is labelled with a fluorescent label, for example one or more of fluorophores from AlexaFluors, Seta Fluors, or DyLight Fluors.

Fluorescence is particularly suited for high content microscopy because of the high sensitivity which can be achieved.

The term "labelling" as used herein refers to the attachment of a label to a target. Examples of targets in this invention are RPA, DB-RPA, DNA damage markers such as DSB markers and nuclei. A label refers to a composition detectable by spectroscopic, fluorimetreic, photochemical, biochemical, immunochemical, enzymatic, chemical or other physical means. In particular embodiments of the invention relate to where the label is one or more fluorescent compositions.

Labelling may be direct for example by the binding of DB-RPA to a compound which is directly conjugated to e.g. a fluorophore, e.g., primary antibody conjugated to fluorophore. Labelling may alternatively be indirect, for example via the conventional two-step method of immunological staining which uses a primary antibody to recognize and bind the target, and secondary antibody that is conjugated to a label, such as fluorophore, to bind to the primary antibody.

In one embodiment, one or more antibody or antibody-derived binding agent which binds to RPA is used. Examples of antibodies or antibody-derived binding agents include antibodies (such as monoclonal and polyclonal antibodies), Fab fragments, single chain antibodies, single domain antibodies, microantibodies.

In another embodiment, one or more antibody mimetic which binds to RPA may be used. Examples of antibody mimetics include affibody molecules, affilins, affimers, affitins, anticalins, avimers, DARPins, FYnomers, Kunitz domain peptides and monobodies.

In one embodiment, a primary antibody is used to detect and bind to RPA, and a secondary antibody conjugated to a fluorophore is used to bind the primary antibody. The fluorophore is later detected.

RPA-bindning

In the labelling of RPA, one or more antibody or antibody-derived binding agent which binds to RPA may be used for recognition of RPA. The antibody or antibody- derived binding agent may bind to any of the RPA-subunits, as long as binding reflects intact RPA complex. Only intact RPA complex will bind to DNA.

In one embodiment, the labelling of the remaining RPA, being DB-RPA, is done by labelling of RPA1, for example by using an antibody against RPA1, such as a monoclonal antibody against RPA1.

In another embodiment, the labelling of the remaining RPA, being DB-RPA, is done by labelling of RPA2 and/or RPA3, for example by using antibody against RPA2 or RPA3 respectively.

Thus, one embodiment relates to a method according to the invention wherein the DB/RPA is detected using an antibody against a target selected from the group comprising RPA1, RPA2 and/or RPA3, preferably RPA1.

Labelling damaged DNA

The labelling of damaged DNA may be done for example in a manner analogous to that done for RPA labelling described above. The labelling of the marker for DNA double strand breaks, may be done in any suitable way which is compatible with the acquisition of labelling data. In one embodiment, one or more antibody or antibody-derived binding agent which binds to DNA double strand breaks is used. Examples of antibody-derived binding agents include antibodies (such as monoclonal and polyclonal antibodies), Fab fragments, single chain antibodies, single domain antibodies, microantibodies.

In another embodiment, one or more antibody mimetic which binds to RPA may be used. Examples of antibody mimetics include affibody molecules, affilins, affimers, affitins, anticalins, avimenrs, DARPins, FYnomers, Kunitz domain peptides and monobodies.

In one embodiment, the DNA double strand break marker is gamma-H2AX.

H2AX is a histone protein, which becomes phosphorylated on serine 139 as a reaction to DNA double-strand breaks (DSB) or Single strand breaks, SSB) and is then known as gamma-H2AX. Kinases of the PI3KK-family (ATM and DNA-PKCs) are mainly responsible for this phosphorylation in response to DSBs. ATR (another PI3KK) phosphorylates H2AX on sl39, as a result of ssDNA. When gamma-H2AX is detected in cells with exhausted free RPA, it is due to DSB (RC).

In one embodiment of this aspect of the invention gamma- H2AX is used as a marker for DNA damage, such as DNA double strand breaks, and is detected by an antibody, such as a primary antibody obtained from Biolegend, 613401.

In another embodiments of this aspect, RPA-phosphorylation, such as one or more of phosphorylation on S4/8 or T21 on RPA2 is a marker for DNA double strand breaks.

RPA2 is a subunit in the RPA complex, and the N-terminal domain of RPA2 is phosphorylated after DNA damage, such as DNA double strand breaks.

Thus, in one embodiment of this aspect of the invention phosphorylated RPA2 is used as a marker for DNA double strand breaks, and is detected by a primary antibody selected from RPA32-pS4/8 (Bethyl, A300-245A), RPA32-pT21 (Abeam, ab61065).

In further embodiments, Phospho-DNA-PK (Ser2056) or Phospho-ATM (Serl981) may be used as DNA damage markers, in particular markers for DNA double strand breaks. Other markers for DNA double strand breaks known in the art may be used.

Labelling the nucleus

According to the method, the nucleus is also labelled. This is done in order to ensure that all cells, even those which might not contain any DB-RPA1, are detected after software- based image analysis.

The nucleus may be labelled by any method which is compatible with the high content microscopy and also with the other fluorescent labels used in parallel (e.g. RPA1 label). In one particular embodiment, the nucleus is labelled with a fluorescent label.

Examples of fluorescent nuclear labels include DAPI, Hoechst33342 dye, BOBO™- 1, YO-PRO-1 dye, SYTOX Green dye, TOTOl.

In one particular embodiment, the nucleus is stained with DAPI (4',6-diamidino-2- phenylindole). DAPI binds strongly to DNA and becomes more fluorescent when bound.

The methods of labelling nucleus is within the person of skill in the art, in particular the method of labelling a nucleus with DAPI.

In a particular embodiment of the methods of the invention, the labelling of DB- RPA, DNA damage marker and nucleus is with fluorescent labels.

Quantifying labelled targets

Quantification of the labelled DB-RPA and DNA damage markers, such as DNA double strand break markers is done in a manner which allows the assessment of which cells in the population contain DB-RPA and/or DNA damage marker such as DSB marker above the background, as well as the assessment of relative amount of DB-RPA and/or DNA damage markers such as DSB marker in those cells.

In one embodiment of the invention, the quantification of the DB-RPA and/or DNA damage (such as DSB marker) in the nucleus of the cells of the population is done by quantitative image-based cytometry (QIBC) method, which method comprises the steps:

/. Acquiring images of the cell population

//. Processing the acquired images for image analysis // ^' / ^' . Quantifying the signal from the labelled species. Particular embodiments relate to quantifying the signal from the labelled RPA and labelled DNA damage marker (such as DNA double strand break marker)

iv. Performing calculations on the data, for example in the case of replication catastrophe, determining the percentage of the total cell population that displays DNA damage marker (such as DNA double strand break marker) and that at the same time also displays maximum levels of DB-RPA, a state defined as replication catastrophe

Particular embodiments relate to the exporting the data of the quantified signal for its visualization, for example in order to generate a scatter diagram.

In other embodiments, for example, Microsoft Excel™ may be used to calculate population percentages.

Acquiring images of the cell population is in other words, acquiring data.

One particular embodiment of the methods of all aspects of the invention relate to wherein step f) iv comprises generating a scatter diagram of the two markers c) and d).

Other examples of performing calcluations on the data include determining levels, ratios, relationships, portions, proportions, amounts etc of DNA damage marker, such as DSB marker, and/or of DB-RPA.

Acquiring images- automated acquisition

The images may be acquired by a microscope, which may be manual or may be automated. Particular embodiments relate to automated image acquisition, for example automated unbiased image acquisition using a microscope. For example, the microscope may have automated autofocus, automated adjustment in x,y axes and/or automated image acquisition controlled by software. One example of an automated acquisition software is ScanR, from Olympus.

The magnification is typically lOx, or 4x.

Where the labelling is done with fluorophores, the microscope is a fluorescence microscope, which can be a wide-field or a confocal.

The images may in an alternative embodiment be acquired by flow cytometry. Processing images- image analysis

The images are processed using image analysis software.

In particular embodiments dynamic background correction is performed which increases signal to background resolution.

In particular embodiments a mask is generated based on the staining of nuclei and is used for segmentation of the nuclei. This mask identifies each nucleus as an individual object. The mask may serve to exclude signal from areas outside the nucleus.

Thus, one embodiment relates to the method according to the invention, wherein step f) ii comprises generating a mask based on the staining in step e ).

Quantifying signal

The signal from the labels is quantified as pixel intensities in the different channels for each individual cell/foci, using image analysis software.

Scatter image or Scatter diagrams

The quantification data may be exported for further analysis and generation of diagrams. In one particular embodiment, the data is exported to a software which generates scatter diagrams, also known as scatter plots. One example of such software is Microsoft Excel or TIBCO SpotFire ® (TIBCO Spotfire, Boston, MA, US).

Several time points

The method according to the invention may further be repeated, and the results compared with each other to document any progression or change in levels of DB- RPA and/or DNA damage (e.g. DSBs) in response to therapy over time.

In one particular embodiment of the method, the quantification comprises labelling and/or acquiring signal at several time points, for example before treatment, and one or more time points after treatment, such as at one or more time points from 15 min to 48 hours after treatment, such as at one or more time points about 15 min, about 30, 45, 60, 90, 120, 150, 180, 220, 240, 260, 280, 300, 320, 360 min after treatment; or for example at one or more time points about 1, 2, 4, 6, 8, 10, 12 14, 16, 18, 20, 22, 24, 26, 28, 30, 32, 34, 36, 28, 40, 42, 46, 48 hours after treatment. Applications of the method

An illustration of the power of the methods of the invention is give in Figures 7 to 10.

Intensities for RPA and gamma H2AX are obtained for individual cells by software- based image analysis, giving rise to the scatter diagrams that allows us to identify the levels of RC coupled to the point of exhaustion of free RPA (see Fig 7).

Referring to Figure 10, whereas a classical biochemical analysis (Western blot) would discard Tumor C as resistant to ATR/CHK1 inhibitors, QIBC reveals that , while there is no detectable DNA damage, the tumor is very close to exhaust its RPA pool and suffer RC, which opens the possibility to treat it with an additional drug.

Referring now to Figure 8, it can be seen that different combinations of drugs which each efficiently elicit RC can be identified and used as alternative therapies to overcome pharmacological barriers or acquired resistance.

Figure 8 and 10 describes a hypothetical case in which an inefficient therapy is revealed by QIBC to not elicit lethal DNA damage. In a similar manner, addition of another drug is shown to greatly enhance the therapeutic response.

Steps in common

The embodiments, features and considerations presented above in the context of the aspect of the invention relating to a method for measuring the number of cells in a population in Replication Catastrophe, also apply to all the other aspects of the invention presented further herein below.

2. Method for determining effectiveness of a cancer therapy

Not all cancer therapies will induce replication catastrophe. If they don't, the cancer cells can be just subjected to non-irreversible DNA damage or replication stress, leading to an eventual regrowth of the cell population. Whether they do or not may depend both on the mode of action of the therapy, as well as on the cells which are being treated. By determining whether a therapy will induce replication catastrophe, as well as to what degree, it can also be determined what additional or alternative therapy can be used in order to elicit to RC in this cell population.

Thus, the invention in another aspect relates to a method for determining the effectiveness of a cancer therapy comprising a) Providing a cell population, and exposing the cell population to a cancer therapy

b) Labelling the DNA-bound RPA;

c) Labelling a marker for DNA damage (such as DSB marker), d) Labelling the nucleus,

e) Quantifying the DNA bound-RPA and the DNA damage marker in the nucleus of the cells of the population by

/. Acquiring images of the cell population,

//. Processing the acquired images by software-based image analysis,

// ^' / ^' . Quantifying the signal from the labelled DNA bound RPA from b) and labelled DNA damage marker from c),

iv. Determining the level of the total cell population that displays DNA damage marker and that at the same time also displays maximum levels of DB-RPA, a state defined as replication catastrophe.

f) Assessing whether the therapy is effective by comparing the level of cells in Replication catastrophe in iv. with a threshold level, and if the level in iv. is lower than said threshold, concluding that the therapy is ineffective; and if the level in iv. is equal to or higher than said threshold, concluding that the therapy is effective

A further embodiment relates to a method for determining effectiveness of a cancer therapy comprising

a) Providing a cell population, and exposing the cell population to a

cancer therapy

b) Removing free RPA from the cells in the cell population and fixing cells

c) Labelling the remaining RPA, being DNA-bound RPA;

d) Labelling a marker for DNA damages,

e) Labelling the nucleus, f) Quantifying the DNA bound-RPA and the DNA damages in the nucleus of the cells of the population by

/. Acquiring images of the cell population,

//. Processing the acquired images by software-based image analysis,

// ^' / ^' . Quantifying the signal from the labelled DB-RPA from c) and labelled DNA damage marker from d),

iv. Determining the percentage of the total cell population that displays DNA damage marker and that at the same time also displays maximum levels of DB-RPA, a state defined as replication catastrophe.

g) Assessing, based on the number of cells that are in Replication Catastrophe, that the therapy is effective if any cells are in Replication Catastrophe, and where no cells are in Replication catastrophe that the therapy is not effective.

In an alternative embodiment, the assessment of step g) instead comprises comparing the percentage in iv. with a threshold level, and if the percentage in iv. is lower than said threshold, concluding that the therapy is ineffective; and if the percentage in iv. is equal to or higher than said threshold, concluding that the therapy is not effective.

One embodiment relates to the method according to this aspect, wherein step f) /V or step e) iv comprises generating a scatter diagram of the two markers DB- RPA and DNA damage (e.g. c) and d)).

A further embodiment relates to the method according to this aspect of the invention, wherein the provided cell population is incubated with one or more of ATR inhibitor, a Chkl inhibitor, a Weel inhibitor and/or an antimetabolite.

Cell population

The provided cell population may in some embodiments be a cell population obtained from a primary tumor, for example as a tissue sample from a cancer patient. In order to ascertain whether a therapy will be successful in driving a cell population to replication catastrophe, the provided cell population may be exposed to a cancer therapy. Exposure to a cancer therapy means that the cell population is incubated with one or more cancer treatments, such as one or more chemotherapy compounds or potential drugs, for example one or more pharmaceuticals indicated for cancer, such as one or more of a pharmaceutical from the class ATR-inhibitors,

The cell population may be in some embodiments be cultured in vitro prior to incubation in order to expand the cells.

In some embodiments of this aspect, the cell population may be a sample from a patient who at the time the sample was taken, was not undergoing cancer therapy. For example, the patient may have never undergone cancer therapy, or may have previously undergone but, at the time the sample was taken had negligible or no detectable amounts of cancer therapy in blood.

3. Method for selecting a cancer therapy

The invention in another aspect relates to a method for evaluating different cancer therapies and selecting a therapy for a cancer patient, prior to a cancer treatment of this patient. The method aims to examine the whether the sensitivity of the particular cancer to one or more different cancer therapies, in order to identify the most promising therapy/-ies.

Thus, the invention relates in one embodiment to a method of selecting a cancer therapy for a patient having a cancer comprising

a) Providing a cell population from the patient, obtained prior to a cancer treatment of the patient wherein the cell population is a primary tumour cell population, and incubating said primary tumour cell population from the patient with a cancer therapy for a period of time

b) Labelling the DNA-bound RPA, c) Labelling a marker for DNA damage, such as DSB marker d) Labelling the nucleus e) Quantifying the DNA bound-RPA and the DNA damage marker in the nuclei of the cells of the population by

/. Acquiring images of the cell population, //. Processing the acquired images by software-based image analysis,

// ^' / ^' . Quantifying the signal from the labelled DB-RPA from b ) and labelled DNA damage marker from c),

iv. Determining the level of the total cell population that displays

DNA damage marker and that at the same time also displays maximum levels of DB-RPA, a state defined as replication catastrophe,

f) Assessing whether the therapy is effective by comparing the level of cells in Replication catastrophe in iv. with a threshold level, and if the level in iv. is lower than said threshold, concluding that the therapy is ineffective; and if the level in iv. is equal to or higher than said threshold, concluding that the therapy is effective

In a further embodiment, the invention relates to a method of selecting a cancer therapy for a patient having a cancer comprising

a) Providing a cell population from the patient, obtained prior to a

cancer treatment of the patient wherein the cell population is a primary tumour cell population, and incubating said primary tumour cell population from the patient with a cancer therapy for a period of time

b) Removing free RPA from the cells in the cell population and fixing cells

c) Labelling the remaining RPA, being DNA-bound RPA,

d) Labelling a marker for DNA damage, such as DNA double strand breaks

e) Labelling the nucleus

f) Quantifying the DNA bound-RPA and the DNA damage marker (such as DSB marker) in the nuclei of the cells of the population by i. Acquiring images of the cell population,

ii. Processing the acquired images by software-based image analysis, iii. Quantifying the signal from the labelled DB-RPA from c ) and labelled DNA damage marker (such as DNA double strand break marker) from d),

iv. Determining the percentage of the total cell population that displays DNA damage marker (such as DNA double strand break marker) and that at the same time also displays maximum levels of DB-RPA, a state defined as replication catastrophe,

g) Assessing, based on the number of cells that are in Replication Catastrophe, that the therapy is effective if any cells are in

Replication Catastrophe, and where no cells are in Replication catastrophe that the therapy is not effective.

One embodiment relates to the method according to this aspect, wherein step f) /V or step e) iv comprises generating a scatter diagram of the two markers DB- RPA and DNA damage (c) and d), or b) and c)).

In an alternative embodiment, the assessment of step g) instead comprises comparing the percentage in iv. with a threshold level, and if the percentage in iv. is lower than said threshold, concluding that the therapy is ineffective; and if the percentage in iv. is equal to or higher than said threshold, concluding that the therapy is not effective.

In some embodiments, the cell population which is provided is from a patient who is not undergoing cancer therapy; and/or from a patient who has undergone a cancer therapy previously, but where the previous therapy is no longer detectable in patient samples.

Assessment

The assessment aims to determine whether the cells will be driven to replication catastrophe of the particular therapy or not. If any of the cells of the cell population are in RC state, the therapy is assessed to be effective, and where none of cells are in RC indicate that the therapy is ineffective. Recommendation

An embodiment of this method further comprises a step i) to be taken after the step of assessment, (in some embodiments step h), of making a recommendation for a cancer therapy based on the assessment.

Thus, if the therapy leads to replication catastrophe this therapy is considered effective and is recommended, while if a therapy does not lead to replication catastrophe the cancer therapy is considered ineffective, and is not

recommended.

However, even therapies which are effective, i.e. do induce RC, may benefit from being improved. Thus, it may be desirable to achieve higher levels of population in RC, for example 10%, 15%, 20%, 25%, 35%, 40%, 50% , 65%, 70%, 75%, 80%, 85 % of the cell population in RC. This may be achieved by combing cancer therapies. IT may then be determined from repeating a method according to the invention, if the new combination has an improved effectiveness.

Further embodiments

In one embodiment of the present aspect, the method is repeated with different cancer therapies, for example with different drugs and/or different concentrations of a drug.

In embodiment of the present aspect, the cell population provided may be a cell line.

4. Method of selecting a therapy and treating a cancer

The invention in a further aspect relates toa method of selecting a therapy and treating cancer in a patient comprising

a) Providing a cell population from the patient, wherein the cell population is a primary tumour cell population and incubating said cell population from the patient with a cancer therapy for a period of time

b) Labelling the DNA-bound RPA,

c) Labelling a marker for DNA damage, such as DSB marker

d) Labelling the nucleus e) Quantifying the DNA bound-RPA and the DNA damage marker in the nuclei of the cells of the population by

/. Acquiring images of the cell population

//. Processing the acquired images by software-based image analysis,

// ^' / ^' . Quantifying the signal from the labelled DB-RPA from b ) and labelled DNA damage marker from c),

iv. Determining the level of the total cell population that

displays DNA damage marker and that at the same time also displays maximum levels of DB-RPA, a state defined as replication catastrophe.

f) Assessing whether the therapy is effective by comparing the level of cells in Replication catastrophe in iv. with a threshold level, and if the level in iv. is lower than said threshold, concluding that the therapy is ineffective; and if the level in iv. is equal to or higher than said threshold, concluding that the therapy is effective.

g) Making a recommendation for a cancer therapy based on the assessment, where effective therapies are recommended h) Administering to a patient, the therapy recommended in step g). A further embodiment relates to a method of selecting a therapy and treating cancer in a patient comprising

a) Providing a cell population from the patient, wherein the cell population is a primary tumour cell population and incubating said cell population from the patient with a cancer therapy for a period of time

b) Removing free RPA from the cells in the cell population and fixing cells

c) Labelling the remaining RPA, being DNA-bound RPA,

d) Labelling a marker for DNA damage (such as DNA double strand

breaks)

e) Labelling the nucleus f) Quantifying the DNA bound-RPA and the DNA damage marker (such as DSB marker) in the nuclei of the cells of the population by

/. Acquiring images of the cell population

//. Processing the acquired images by software-based image analysis,

// ^' / ^' . Quantifying the signal from the labelled DB-RPA from c ) and labelled DNA damage marker (such as DNA double strand break marker) from d),

iv. Determining the percentage of the total cell population that displays DNA damage markers (such as DNA double strand break marker) and that at the same time also displays maximum levels of DB-RPA, a state defined as replication catastrophe.

g) Assessing, based on the number of cells that are in Replication Catastrophe, that the therapy is effective if any cells are in

Replication Catastrophe, and where no cells are in Replication catastrophe that the therapy is not effective

h) Making a recommendation for a cancer therapy based on the assessment, where effective therapies are recommended i) Administering to a patient, the therapy recommended in step h).

One embodiment relates to the method according to this aspect, wherein step f) /V or step e) iv comprises generating a scatter diagram of the two markers DB- RPA and DNA damage (c) and d), or b) and c)).

In an alternative embodiment, the assessment of step g) instead comprises comparing the percentage in iv. with a threshold level, and if the percentage in iv. is lower than said threshold, concluding that the therapy is ineffective; and if the percentage in iv. is equal to or higher than said threshold, concluding that the therapy is not effective. 5. Method of treatment

The invention in another aspect relates to a method for treating cancer in a patient comprising requesting a test providing the results of an analysis to determine the percentage of the total cell population that displays DNA damage marker, such as double strand break marker and that at the same time also displays maximum levels of DB-RPA, a state defined as replication catastrophe, and administering said cancer therapy to the patient if it leads to cells entering replication catastrophe.

6. Method for monitoring a cancer therapy

In the situation where a cancer therapy has been given to a patient, monitoring the response to the therapy is valuable. Thus, the invention in another aspect relates to a method for for monitoring the therapy of a cancer patient comprising a) Providing a cell population from the patient, wherein the patient is receiving cancer therapy

b) Labelling the DNA-bound RPA,

c) Labelling a marker for DNA damage, such as DSB

d) Labelling the nucleus

e) Quantifying the DNA bound-RPA and the DNA damage marker in the nuclei of the cells of the population by

/. Acquiring images of the cell population

//. Processing the acquired images by software-based image

analysis,

// ^' / ^' . Quantifying the signal from the labelled RPA from b) and

labelled DNA damage marker from c)

iv. Determining the level of the total cell population that displays DNA damage marker and that at the same time also displays maximum levels of DB-RPA, a state defined as replication catastrophe

f) Assessing whether the therapy is effective by comparing the level of cells in Replication catastrophe in iv. with a threshold level, and if the level in iv. is lower than said threshold, concluding that the therapy is ineffective; and if the level in iv. is equal to or higher than said threshold, concluding that the therapy is effective.

A further embodiment relates to a method of monitoring the therapy of a cancer patient comprising

a) Providing a cell population from the patient, wherein the patient is receiving cancer therapy

b) Removing free RPA from the cells in the cell population and fixing cells

c) Labelling the remaining RPA, being DNA-bound RPA,

d) Labelling a marker for DNA damage (such as DNA double strand

breaks)

e) Labelling the nucleus

f) Quantifying the DNA bound-RPA and the DNA damage (such as DSB marker) in the nuclei of the cells of the population by

/. Acquiring images of the cell population

//. Processing the acquired images by software-based image

analysis,

// ^' / ^' . Quantifying the signal from the labelled RPA from c ) and

labelled DNA damage marker (such as DNA double strand break marker) from d)

iv. Determining the percentage of the total cell population that displays DNA damage marker (such as DNA double strand break marker) and that at the same time also displays maximum levels of DB-RPA, a state defined as replication catastrophe

g) Assessing, based on the number of cells that are in Replication Catastrophe, that the therapy is effective if any cells are in Replication Catastrophe, and where no cells are in Replication catastrophe that the therapy is not effective

In this method, the provided cells are from a patient sample wherein the patient is undergoing cancer therapy. The method aims to determine what effect the therapy that the patient is currently undergoing has on the cells of the cancer with respect to replication catastrophe.

One embodiment relates to the method according to this aspect, wherein step f) iv. or step e) iv. comprises generating a scatter diagram of the two markers DB- RPA and DNA damage (e.g. DSB) ( c) and d), or b) and c)).

In an alternative embodiment, the assessment of step g) instead comprises comparing the percentage in iv. with a threshold level, and if the percentage in iv. is lower than said threshold, concluding that the therapy is ineffective; and if the percentage in iv. is equal to or higher than said threshold, concluding that the therapy is not effective.

In embodiments of this method, the method is repeated at time intervals in order to evaluate the efficiency of the treatment over time. For example, the method may be performed once a day, once a week, twice a week, twice a month, once a month. The evaluation may extend over a period of time, and the method may be performed at a frequency according to the above, over a period of time for example at least one month, at least two months, at least three months, or for example for one to 6 months, one to 12 months or one to 18 months.

In one embodiment, if the results of the method shows that a number of cells that is over 20%, such as over 30%, over 35%, over 40%, over 45% or such as from 20 to 40% from 2o to 50% , or 20 to 55% of the total cell population is

undergoing replication catastrophe the cancer therapy may be considered particularly effective. If the method shows that for example 20% or less, such as 18% or less, 15% or less, 10% or less of the cells do not reach replication catastrophe, this is indicative that the therapy is less effective and should be adjusted. These considerations apply also to assessment performed in any of the methods of the invention.

The adjustment may mean to adjust doses of the current therapy, and/or to start treatment with other therapies. The method may include repeating testing of therapies using methods according to the invention.

Further embodiments of this method relate to the method further comprising a step h), Making a recommendation for a cancer therapy based on the assessment.

Yet further embodiments relate to the method further comprising a step i) after step h), Administering to a patient the therapy selected in h). 7. Method for monitoring a cancer therapy and treating a cancer

It is valuable to be able to monitor a cancer therapy so as to be able to adjust the therapy if there be a need. Thus, the invention in a further aspect relates to a method for monitoring the therapy and treating a cancer patient comprising

a) Providing a cell population from the patient, wherein the patient is receiving cancer therapy

b) Labelling the DNA-bound RPA,

c) Labelling a marker for DNA damage, such as DSB

d) Labelling the nucleus

e) Quantifying the DNA bound-RPA and the DNA damage marker in the nuclei of the cells of the population by

/. Acquiring images of the cell population

//. Processing the acquired images by software-based image

analysis,

// ^' / ^' . Quantifying the signal from the labelled RPA from b ) and

labelled DNA damage marker from c)

iv. Determining the level of the total cell population that displays DNA damage marker and that at the same time also displays maximum levels of DB-RPA, a state defined as replication catastrophe

f) Assessing whether the therapy is effective by comparing the level of cells in Replication catastrophe in iv. with a threshold level, and if the level in iv. is lower than said threshold, concluding that the therapy is ineffective; and if the level in iv. is equal to or higher than said threshold, concluding that the therapy is effective

A further embodiment relates to a method for monitoring a cancer therapy and treating a cancer, comprising a) Providing a cell population from the patient, wherein the patient is receiving cancer therapy

b) Removing free RPA from the cells in the cell population and fixing cells

c) Labelling the remaining RPA, being DNA-bound RPA,

d) Labelling a marker for DNA damage (such as DNA double strand break)

e) Labelling the nucleus

f) Quantifying the DNA bound-RPA and the DNA damage marker (such as DSB marker) in the nuclei of the cells of the population by

/. Acquiring images of the cell population

//. Processing the acquired images by software-based image analysis,

// ^' / ^' . Quantifying the signal from the labelled RPA and labelled DNA damage

iv. Determining the percentage of the total cell population that displays DNA damage marker (such as DNA double strand break marker) and that at the same time also displays maximum levels of DB-RPA, a state defined as replication catastrophe

g) Assessing, based on the number of cells that are in Replication

Catastrophe, that the therapy is effective if any cells are in

Replication Catastrophe, and where no cells are in Replication catastrophe that the therapy is not effective

h) Making a recommendation for a cancer therapy based on the

assessment

i) Administering to a patient, the therapy selected in step h).

One embodiment relates to the method according to this aspect, wherein step f) iv. or e) iv. comprises generating a scatter diagram of the two markers DB-RPA and DNA damage such as DSB, (c) and d) or b) och c)). In an alternative embodiment, the assessment of step g) instead comprises comparing the percentage in iv. with a threshold level, and if the percentage in iv. is lower than said threshold, concluding that the therapy is ineffective; and if the percentage in iv. is equal to or higher than said threshold, concluding that the therapy is not effective.

8. Method of treatment

The invention in a further aspect relates to a method of treatment of cancer. The considerations below apply to all aspects of the invention relating to a method of treatment.

It has been shown that when RPA pools are exhausted, DNA damage accumulates and the cells progress to RC. The point at which the RPA pool is exhausted depends on the total amount of RPA in the cell and the amount of ssDNA in need of protection. The methods of treatment of the present invention relate to affecting the amount of ssDNA in need of protection.

The amount of ssDNA in need of protection is a result of the level of replication stress (which will result in an amount of ssDNA per replication site) and the number of ongoing replication sites that are sensitive to that stress. Both parameters can be increased with different chemotherapeutic drugs, therefore achieving the highest level of ssDNA and RC when used in combination.

In order to elicit RC, cancer therapy may be selected in order to maximize the contribution from these two sources. Thus, methods of therapy according to the invention may comprise administering one or more drugs selected from the group comprising drugs which elicit replication stress in analogy with HU effects seen in experiments, and drugs which elicit uncontrolled origin firing in analogy with ATR- inhibitor effects seen in experiments, and combinations thereof.

In particular embodiments, the method of treatment relates to the administration of a combination of two or more drugs, wherein at least one drug elicits

replication stress in analogy with HU effects seen in experiments, and at least one further drug other elicits uncontrolled replication in analogy with ATR-inhibitor effects seen in experiments. Deoxynucleoside analogues are used to inhibit DNA synthesis and are examples of drugs which elicit replication stress in analogy with HU effects seen in the presented experiments. Cytarabine and gemcitabine are examples of

deoxynucleoside analogues.

They all belong to the group of antimetabolites, which can be nucleoside analogues and/or Ribonucleotide reductase inhibitors (like HU) or also

Dihydrofolate Reductase (DHFR) inhibitors

Thus, one embodiment of any of the methods of treatment according to the invention described herein, comprises administering cytarabine and/or

gemcitabine to a patient.

Particular embodiments relate to where the method of treatment comprises the administration of cytarabine and/or gemcitabine together with an ATR inhibitor and/or a CHK1 inhibitor.

ATR inhibitors and CHK1 inhibitors for clinical use are known, and are examples of drugs which elicit uncontrolled replication in analogy with ATR-inhibitor effects seen in experiments. Examples of ATR inhibitors are Roscovitine, HU, Gemcitabine and Cytarabine. Also CHK1 and Weel inhibitors may elicit uncontrolled origin firing in analogy with ATR-inhibitors.

In particular, the methods of treatment according to the invention may comprise an evaluation step where it is evaluated whether or not the cell population is in a state of RC.

Some embodiments relate to the method according to any of the aspects of the invention, wherein the provided cell population is incubated with one or more of ATR inhibitor, a Chkl inhibitor, a Weel inhibitor and/or an antimetabolite.

It should be noted that embodiments and features described in the context of one of the aspects of the present invention also apply to the other aspects of the invention.

All patent and non-patent references cited in the present application, are hereby incorporated by reference in their entirety. In particular, the paper "ATR Prohibits Replication Catastrophe by Preventing Global Exhaustion of RPA" Cell Volume 155, Issue 5, 21 November 2013, Pages 1088-1103; which is hereby incorporated by reference in its entirety. 9. Method for classifying a compound

The invention relates in a further aspect to a method for classifying a compound of interest, on basis of the relationship of the amounts of DNA-bound RPA to the amount of DNA damage, such as DSB in a cell nucleus induced by said compound. The inventors recognized that the cell response termed herein as Replication Catastrophe is characterized by specific dynamics of the distribution of DNA bound RPA and appearance of DNA damage such as double-stranded DNA breaks in the cell nucleus of a cell exposed to certain compounds. The inventors have described herein these dynamics, and methods by which they may be visualised.

The inventors have also surprisingly recognized that the response termed

Replication Catastrophe herein is one of several characteristic responses that may be elicited by exposure to a compound. Further, it was also recognized that several different compounds could elicit the same response, characterized by the same dynamics of distribution or amount of DNA bound RPA and amount of DNA damage (for example DSB) in the cell nucleus.

Thus it was recognized that incubation of cells with a compound of interest, such as a drug of interest, would elicit a response in the cell which is characteristic for the method of action of the compound, and corresponded to a specific relation between DB-RPA and DNA damage. Thus, the compounds may be classified based on the elicited response in cells, as measured by changes in amounts of DB-RPA and DNA damage. The response of the cell as measured in the amount of DB-RPA and amount of DNA damage (such as DSB) in the cell nucleus, can be visualized by scatter diagram, in the same manner as previously described herein.

Therefore, another aspect of the present invention relates to a method of claissifying a compound of interest, comprising quantifying the amount of DNA- bound RPA in a cell nucleus and correlating said amount to the amount of DNA damage, such as DSB, in the same cell nucleus, for the cells of a cell population that has been exposed to said compound of interest; and comparing the resulting correlated data for the compound of interest with correlated data from one or more reference compounds.

In some embodiments, correlating the amount of DNA-bound RPA in a cell nucleus to the amount of DNA damage in the same cell nucleus may be performed by presenting the amount of DNA-bound RPA present in a cell nucleus on one axis of a scatter plot, and the amount of DNA damage (such as DSB) in that cell nucleus on the second axis of a scatter plot. In particular embodiments, this is performed for a cell population.

In particular embodiments, the method according to this aspect of the invention may comprise the steps of :

a. Providing a cell population, and exposing the cell population to a compound of interest

b. Labelling the DNA-bound RPA (DB-RPA)

c. Labelling a marker for DNA damage, such as DSB marker d. Labelling the nucleus

e. Quantifying the DNA bound-RPA and the DNA damage marker (such as DSB marker) in the nuclei of the cells of the population by

i. Acquiring images of the cell population,

ii. Processing the acquired images by software-based image analysis,

iii. Quantifying the signal from the labelled DB-RPA from b) and labelled DNA damage marker (such as DSB marker) from c), iv. Correlating the labelled DB-RPA from b) and labelled DNA damage marker (such as DSB marker) from c) in the same cell nucleus, for the cells in a cell population that has been exposed to said compound of interest, and v. Comparing the correlated results obtained for the compound of interest with a library of results from reference compounds representing different classes of compounds, and deducing which class the compound of interest belongs to.

Further embodiments relate to methods according to this aspect of the invention, where the exposure to the compound of interest is analzyed at several time points.

The embodiments, features and considerations presented above in the context of other aspects and methods of the invention, and in particular the aspect of the invention relating to a method for measuring the number of cells in a population in Replication Catastrophe, also apply to the present aspect of the invention. In particular, the embodiments, features and considerations described in under the headings Cell population, Treated cell population, Cell culture, Removal of free RPA, Labelling of DB-RPA, RPA-binding, Labelling damaged DNA, Labelling the nucleus, Quantifiying labelled targets (and subheadings thereto), and Several time points, are also applicable to the present aspect of methods for classifying a compound.

In the same manner, the embodiments, features and considerations described in the context of the present aspect of the invention, are also applicable to the previously described aspects of the invention.

Briefly, the cells are exposed to one or more compounds of interest, after which the cells are assayed for the amount of DNA-bound RPA and the amount of DNA damage (sucah as double strand DNA breaks) in the cell nucleus.

The cells may be any cell of interest, but are typically mammalian cells. In specific embodiments, the cells are human cells, such as human primary cells, human immortalised cells (e.g a cell line, such as U20S cells) and/or combinations thereof.

The compound of interest may be any compound, such as a drug, and in particular a cancer therapy which is chemotherapy. Particular embodiments of chemotherapy comprises compounds which inhibit one or more of ATR and/or CHK1 signalling.

Typically the exposure of the cells to a compound of interest is done by incubating cells in cell culture medium containing the compound of interest. Incubation of cells in such manner is well known in the art, and may be done in any manner suitable. The cells are exposed to a compound of interest for a period of time. The period of time may be any length of time deemed to be of interest. For example, the cells may be exposed to a compound of interest for a period of time from 15 minutes to 800 mins, such as for example from 100 to 800, 100 to 600 or 100 to 400 minutes. Alternatively, the exposure of cells to a compound of interest may be done as described herein previously, e.g. cells from a patient treated with a chemotherapy.

In some embodiments, the compound of interest may be titrated from an appropriate starting concentration. Further, the compound of interest may be applied in combination with one or more other compounds to confirm/expose their properties. The more concentrations and combinations used, the richer and more informative the data the method will yield.

For each compound and experimental condition, the relationship of the amount of DB-RPA and the amount of DNA damage (such as DSB) in each cell nucleus of the cell population can be visualised by plotting the data in a scatter diagram, such as a QIBC scatter diagram as described herein previously, see in particular heading Quantifying labelled targets and Quantitative Image Based Cytometry (QIBC). The resulting patterns which emerge in these diagrams reflect the potential

effect/molecular target of the compound, which can be classified by comparing the QIBC diagrams of a given compound to a reference collection of QIBC scatter diagrams from reference compounds with known molecular target. The compound of interest may then be classified on basis on similarity of the effect elicited by it, to the effect elicited by a reference compound.

For example, the classification as being in the same class as for example an ATR inhibitor, will give important information regarding the potential target of the compound, and potential effects of the compound. See also examokes 13 a nd figures 11 and 12. The invention will now be described in further details in the following non-limiting examples.

Examples

10. Experimental Procedures

Plasmids and RNA interference

cDNAs for human RPA subunits were provided by Marc S. Wold (University of Iowa, USA). P2A sequences were obtained as primers (Invitrogen). Cloning into pAc-GFP-Cl (Clontech) is described in the Supplementary information. SiRNA duplexes were from Ambion (Silencer Select) : ATR (s536), Cdc45 (sl5829), Chkl (s503), RPA1 (sl2127). Plasmid transfections were performed with Lipofectamine LTX and Plus Reagent (Invitrogen). siRNA transfections were performed with Lipofectamine RNAiMAX (Invitrogen). Unless specified, siRNAs were used at 25nM. In siRNA titration assays, total siRNA concentration was kept constant with the addition of control siRNA. Cell Culture

Human U-2-OS osteosarcoma, DLDl-ATR-WildType and DLDl-ATR-Seckel colorectal cancer cells (Hurley et al., 2006) were grown in Dulbecco's modified Eagle's medium with 10% fetal bovine serum (GIBCO). For live imaging cells were grown in C02-independent medium (phenol red and riboflavin free, GIBCO). U-2- OS cell lines stably expressing approximately 2-fold excess of RPA2-EGFP were generated by standard procedures; RPA2-GFP avidly accumulated in replication factories regardless of replication stress. The AcGFP-RPAl-P2A-RPA2-P2A-RPA3 cell lines are characterized in Figs. 4 and S4. Drugs and other cell culture supplements are described in Supplemental information.

Immunochemical and biochemical methods

Antibodies used for immunolabeling techniques are specified in the Supplemental information. Whole cell extracts (WCE) were obtained by lysis in RIPA buffer (50 mM Tris-HCI pH 8.0, 150 mM NaCI, 1.0% Igepal CA-630, 0.1% SDS, 0.1% Na- deoxycholic acid, supplemented with protease and phosphatase inhibitors) containing MgCI2 (2mM) and Benzonase (Novagen), and analyzed by SDS-PAGE following standard procedures. For pre-extraction cells were washed once with PBS and incubated with ice-cold PBS containing Triton X-100 (0.2%) for 1 minute on ice prior to fixation. Immunostaining procedure is described in the

Supplemental information. EdU and TUNEL detection was performed by Click-it® assay following manufacturers instructions (Life Technologies).

Microscopy

Images used in QIBC were obtained with a motorized Olympus IX-81 wide-field microscope equipped with fast-switching filter wheels for excitation and emission of DAPI, FITC, Cy3 and Cy5 fluorescent dyes, an MT20 Illumination system, and a digital monochrome Hamamatsu C9100 CCD camera. Olympus UPLSAPO lOx/0.4 NA and 40x/0.9 NA objectives were used. Automated unbiased image acquisition was carried out by the propriety Scan® acquisition software. For full description of QIBC see Supplemental Information. Standard wide-field microscopy was performed on a Zeiss Axio Imager A2 equipped with EC Plan-Neofluar lOx/0.3, 20x/0.5, 40x/0.75 dry objectives, and an AxioCam MRm camera. Wide-field fluorescence time-lapse imaging was carried out on a Zeiss AxioObserver.Zl microscope with a Zeiss Plan-APO 63x/1.4 oil immersion objective and a Coolsnap HQ CCD camera (Roper Scientific), and on an ImageXpress Micro XL Wide-field Automated microscope with a Plan Fluor 40x/0.75 Nikon objective and a 4.66 megapixel scientific CMOS camera .

Cloning

AcGFP-RPA3-P2Al-RPAl-P2A2-RPA2 fusion construct was obtained as follows. ORFs for the different RPA subunits were amplified by standard PCR procedures using the following primers.

RPA1 :

5 '-TATATAG AGCTCATGGTCG GCCA ACTG AGCG - 3 ' [SEQ ID NO: l]

3'-TATATAGTCGACCATCAATGCACTTCTCCTGATGC-5'. [SEQ ID NO: 2]

RPA2 :

5'-TATATAGGTACCATGTGGAACAGTGGATTCGAAAGC-3' [SEQ ID NO: 3]

3 '-T ATAT AG G ATCCTTCTG C ATCTGTG G ATTT AA A ATG GTC - 5 ' . [SEQ ID NO:4]

RPA3 :

5'- TATATATCCGGAATGGTGGACATGATGGACTTGC-3' [SEQ ID NO: 5]

3'- TATATAAGATCTATCATGTTGCACAATCCCTAAAG-5' [SEQ ID NO: 6]

P2A sequences were introduced with the following primers:

P2A1 :

5'- GATCTGGAAGCGGAGCTACTAACTTCAGCCTGCTGAAGCAGGCTGG- AGACGTGGAGGAGAACCCTGGACCTGAGCT-3' [SEQ ID NO: 7]

3'-CAGGTCCAGGGTTCTCCTCCACGTCTCCAGCCTGCTTCAGCAGGCTG- AAGTTAGTAGCTCCGCTTCCA-5' [SEQ ID NO: 8]

P2A2 :

5'-TCGACGGAAGCGGAGCTACTAACTTCAGCCTGCTGAAGCAGGCTGGA- GACGTGGAGGAGAACCCTGGACCTGGTAC-3' [SEQ ID NO: 9]

5'-CAGGTCCAGGGTTCTCCTCCACGTCTCCAGCCTGCTTCAGCAGGCTG- A AGTT AGT AG CTCCG CTTCCG - 3 ' . [SEQ ID NO: 10] The different fragments were digested and ligated into AcGFP-Cl (Clontech). EGFP-RPA2 fusion construct was obtained by introducing RPA2 ORF into pEGFP-Cl (Clontech).

Drugs and cell culture supplements

HU (Sigma-Aldrich), aphidicolin (Sigma-Aldrich), gemcitabine (Sigma-Aldrich), and cytarabine (Sigma-Aldrich) were used as indicated in Figure legends.

Inhibitors for ATM (Ku55933, Ku60019, Selleckchem), CDC7 (PHA-767491, Selleckchem), multiple CDKs (Roscovitine/Seliciclib, Selleckchem), CDK1 (RO- 8803, Calbiochem), CHK1 (UCN01, Sigma-Aldrich), WEE1 (MK-1775,

Selleckchem) and ATR (donated by Oskar Fernandez-Capetillo, CNIO, Spain) were applied as indicated. dNTP analogues BrdU, CldU, IdU (Sigma-Aldrich), and EdU (Life Technologies) were used as indicated. For native BrdU labeling, cells were grown with BrdU (lOmicroM) for 48 h.

Immunostaining and immunoblotting

For immunostaining, cells growing on 12mm coverslips were fixed in

formaldehyde 4% (VWR) for 15 minutes at room temperature (wherever specified in Figure legends, pre-extraction was carried out before fixation as described in experimental procedure. When Click-it® reactions were combined (EdU and TUNEL), these were performed prior to incubation with the primary antibodies. Primary antibodies were diluted in filtered DMEM containing 10% FBS and 0.05% Sodium Azide. Incubations with the primary antobodies were performed at room temperature for 1 to 3 hours. Coverslips were washed twice with PBS-Tween20 (0.01%) and incubated in DMEM/FBS/SA containing secondary fluorescently labelled antibodies (Alexa fluorophores, Life Technologies) for 45 minutes. PBS-T containing 4',6-Diamidino-2-Phenylindole Dihydrochloride (DAPI, 0.5 microg/ml) was applied for 5 minutes at room temperature to stain DNA. After two more washes in PBS-T, coverslips were dipped in distilled water, placed on 3MM paper to dry, and mounted on 10 microl Mowiol-based mounting media : Mowiol 4.88 (Calbiochem)/Glycerol/TRIS. For immunoblotting, primary antibodies were incubated over night at 4 degrees Celsius in PBS-T containing 5% powder milk. Secondary Peroxidase-coupled antibodies (Vector labs) were incubated at room temperature for 1 hour. ECL-based chemiluminescence was detected on

Hyperfilms (GE) and with an Odyssey-Fc system. Primary antibodies were used at the indicated dilutions (applies to both immunostaining and immunoblotting) : ATM (Cell Signaling, 2873, 1/500), ATM-pS1981 (Rockland, 200-301-400, 1/1000), ATR (Cell Signaling, 2790, 1/500), BrdU (GE, RPN202, 1/1000), Cdc45 (Santa Cruz, sc-20685, 1/500), Chkl (Santa Cruz, sc-8408, 1/1000), Chkl-pS345 (Cell signalling, 2348, 1/250), Chk2-pT68 (Cell signalling, 2661, 1/250), H2AX-pS139 (Biolegend, 613401, 1/1000), Kapl (Bethyl, A300-274A, 1/1000), Kapl-pS824 (Abeam, ab70369, 1/250), RPA1/RPA70 (Abeam, ab79398, 1/1000), RPA2/RPA32 (Abeam, abl6850, 1/1000), RPA32-pS33 (Bethyl, A300-246A, 1/1000), RPA32- pS4/8 (Bethyl, A300-245A, 1/1000), RPA32-pT21 (Abeam, ab61065, 1/1000), RPA3/RPA14 (Abeam, ab6432, 1/500).

Quantitative Image Based Cytometry (QIBC)

Images were acquired in an unbiased fashion with the Scan® acquisition software controlling a motorized Olympus IX-81 wide-field microscope. Acquisition times for the different channels were adjusted to obtain images in not saturating conditions (12-bit dynamic range) for all the treatments analyzed within the experiment. The majority of the data was obtained with the Olympus UPLSAPO lOx/0.4 NA objective. Depending on cell confluence, 20 to 30 images were acquired, containing in total 5000 to 10000 cells per condition. For the foci analysis in Fig. 4 the images were generated with the UPLSAPO 40x/0.9 NA objective. A total of 256 images were acquired, containing 5400 cells. More than 600000 individual foci were analyzed. After acquisition, the images were

processed for image analysis with the Scan® image analysis software. A dynamic background correction was first applied to the images. DAPI signal was used for segmentation of the nuclei according to intensity threshold, generating a mask that identified each individual nucleus as an individual object. This mask was then applied to quantify pixel intensities in the different channels for each individual cell/object. In Fig. 4, a second mask was generated by segmentation of RPA2 images in individual spots (using the spot detection module included in the software), and used for quantification of pixel intensities. After segmentation and pixel quantification, the desired quantified values for each cell/foci (Mean and total intensities, area, number of foci) were extracted and exported to the propriety Spotfire™ software. Spotfire™ was used us to quantify percentages and average values in cell populations and to generate all color-coded scatter diagrams in a flow-cytometry-like fashion. This approach allowed us to visualize and relate key features of replication stress and DNA damage signaling for thousands of cells in different stages of the cell cycle, integrating their temporal dynamics. The key features of the QIBC assay can be described as follows:

Single cell analysis by high content microscopy not only provides the spatial resolution of fluorescence imaging, but also greatly exceeds flow- cytometry and immunoblotting in resolution and quantitative power. An example of this capability is illustrated in Fig 5E, where slight reductions in RPA levels are acurately resolved by quantitative imaging.

The examination of asynchronous populations is key to explore and resolve the dynamic complexity of the cell cycle (whose implications in the interpretation of cell biology data are commonly overseen). Moreover, it provides outstanding quality and robustness to our datasets, as the intrinsic phenotypical heterogeneity among different cell cycle stages (characterized by different markers) serves as internal normalization control in each experiment.

The integration of thousands of measurements into multidimensional scatter diagrams allows a new 'supra level' of data visualization, which has been essential for quality controlling, understanding, interpreting, and further developing our experimental results.

Processing of time lapse images

The spatial visualization of dynamic intra-nuclear structures (like RPA foci) through time is challenged for the observer by the movement of the cell across the field of view. In order to enhance it, we immobilized individual cells η silico' throughout the time-lapse experiments. We processed the images as stacks with Fiji (ImageJ, http://fiji.se/Fiji) and applied the Stackreg/Turboreg plugin

(http://bigwww.epfl.ch/thevenaz/stackreg/). This plugin recognizes similar features in consecutive images and fixes their coordinates, generating a new stack in which the cells appear to be immobile (Supplemental Videos 1 to 5). The multicolor coding in Fig. 4A was generated by changing the LUT (Look Up Table) of the stack containing the images in Fiji.

Recover and repair assay

In experiments described in Fig. 5, we considered that cells whose forks were not irreversibly damaged by the HU + ATRi treatment would double every 24 hours. Based on this assumption, we develop the following model : As an example, a starting proportion of 50 / (50 + 50) (Damaged / (Damaged + Not damaged)) would change to (50-R) / ((50-R) + (50*2+R*2)) in 24 hours, R being the number of recovered cells that resumed cell proliferation. The same calculation was applied for every 24-hour cycle. Thus, if no cells recovered the expected proportions would be 1 : 0,33 : 0,2 : 0,11 (0 to 72h). When comparing these values to the experimentally measured ones, we obtained the ratios depicted in Fig. 5C. When we included recovery rates higher than 0 in the model (down to 0.05), the predicted proportions of damaged cells were lower than the ones obtained experimentally. This, together with the rest of supporting evidence presented in Fig. 6, led us to conclude that no cells recovered from the replication damage.

Metaphase Spreads

U-2-OS cells were incubated with EdU (10 microM) for 10 minutes just prior to the different treatments (HU + ATRi or HU + WEEli). The inhibitors were replaced with fresh medium after 80 to 120 minutes (the only criteria here was to choose a time that would yield enough cells with DSB to analyze), and 24 to 48 hours later, they were incubated on the plate with Calyculin ΙΟΟηΜ for 45 minutes. Finally cells were processed to obtain metaphase spreads following standard protocols. Briefly, cells were spun down and resuspended in pre-warmed KCI 0.075 M. After 15 minutes at 37 degrees, cells were fixed in freshly prepared Metanol: Acetic acid (3: 1) for 25 minutes. Cells were dropped on 50x25mm coverslips, chromosomes stained first with a 594 azide (Click-it®) for 30 minutes and then with DAPI for 1 minute. Finally, the coverslips were mounted on slides with mowiol and imaged in a Zeiss LSM-780 confocal microscope with a Zeiss 63x/1.2 water immersion objective.

ll. Example 1

Stalled forks are transiently resistant to breakage independent ofATR Signaling

Stalled forks are transiently resistant to breakage independent of ATR signaling

To investigate mechanisms involved in replication fork stability, we induced replication stress by treating U-2-OS cells with hydroxyurea (HU), which causes rapid depletion of dNTPs (Eklund et al., 2001). To manipulate ATR activity, we applied a specific ATR inhibitor (ATRi), which faithfully recapitulates phenotypes associated with genetic ablation of ATR (Toledo et al., 2011). In addition to inducing gradual phosphorylation of ATR targets such as H2AX (pS139), CHK1 (pS345) and RPA2 (pS33) , combination of HU with ATRi resulted in a delayed phosphorylation of ATM targets including ATM (pS1981), KAP1 (pS824), CHK2 (pT68) and RPA2 (pT21, pS4/8). This was accompanied by generation of DSBs detected by the TUNEL assay specifically at the RPA-decorated stalled forks with hallmarks of ATM activity . Whereas fork breakage in ATR-deficient cells was expected, it was surprising to see that this occurred with a substantial delay (Fig. 1) despite the kinase is inhibited within seconds to minutes (Toledo et al., 2011). This raised the possibility that for a limited period of time the fork protection machinery operates autonomously and independently of ATR.

To investigate this possibility, we established a Quantitative Image-Based

Cytometry (QIBC) method, which allowed us to monitor the complete dynamics of replication stress responses with unprecedented detail and in a fully automated and high content fashion (see Supplemental methods). The key feature of this assay is its ability to separate with great accuracy and reproducibility the fraction of cells with intact stalled forks from those where the combination of HU and ATRi treatment triggered fork breakage accompanied by hallmarks of ATM activity such as H2AX hyper-phosphorylation (Fig. D). Interestingly, also at this analytical level, the conversion of stalled forks to DSBs lagged behind acute ATR inhibition and we therefore set out to investigate the mechanistic underpinnings of fork surveillance between ATR inhibition and DSB generation.

12. Example 2:

Breakage of stalled forks correlates with increased levels of chromatin -loaded RPA

We noticed that before DSBs became detectable, the amount of RPA loaded on chromatin accumulated to levels that were several fold higher than in cells exposed to HU alone (Fig. 1C). Increasing the HU concentration did not further increase RPA loading, confirming that replication was fully stalled . These data allowed for two predictions: i) that ATR may have a specific role in limiting RPA chromatin loading, and ii) that the excessive RPA loading in the absence of ATR might be coupled to DSB generation. Consistent with these predictions, the accumulation of RPA on chromatin reached its peak well before DSB generation, and the conversion of stalled forks to DSBs was confined to cells with the highest degree of chromatin-loaded RPA both in U-2-OS cells, used as a model system throughout this study (Figs. 2A, 2B, ), and in primary, non-immortalized human fibroblasts (Fig. 2C). Furthermore, we reproduced these results by knocking down ATR with siRNA , and in DLDl-ATR-Seckel cells carrying a hypomorphic ATR mutation (Hurley et al., 2006) . Because prolonged dNTP deprivation was reported to trigger DSBs even in cells with intact ATR (Petermann et al., 2010) we applied QIBC on cells treated with HU alone for up to 24h. Indeed, we saw progressive accumulation of cells with broken forks in the later time points and DSB generation was again restricted to cells with the highest levels of chromatin- loaded RPA . Thus, the excessive accumulation of RPA at stalled forks precedes DNA breakage and ATR delays this pathological outcome of replication stress.

13. Example 3

Nuclear pool of RPA is rate limiting for ssDNA protection at stalled forks

Because RPA avidly interacts with ssDNA (Fanning et al., 2006), we considered the possibility that ATR inhibition might increase ssDNA generation up to the point where it would deplete all available RPA. Initially, generation of ssDNA detected by BrdU incorporation under non-denaturing conditions and accumulation of chromatin-loaded RPA followed an expected linear trend. However, when RPA loading reached its limit, ssDNA continued to accumulate . This was reflected by a deviation from linearity of the ssDNA/RPA ratio, suggesting that cells reached a stage when RPA became limited for binding newly generated ssDNA. Performing the same assay in a cell line stably expressing RPA2-EGFP allowed us to include a DSB marker (ATM-phosphorylated RPA2-pT21) and thus directly correlate ssDNA formation, RPA loading, and DNA breakage. Remarkably, DSBs were strictly confined to cells that had sequestered all RPA and generated an excess of uncoated ssDNA . Thus, in the absence of ATR, ssDNA at stalled forks

progressively depletes the nuclear pool of RPA, which is accompanied by a conversion of stalled forks to DSBs.

14. Example 4

Excessive origin firing depletes RPA and triggers simultaneous fork breakage in active replication factories

ATR activity is propagated by the CHKl kinase, which diffuses from stalled forks to restrain origin firing. Although this 'global' spreading of the ATR pathway seems to have limited bearings on 'local' protection of stalled forks, our data indicated a mechanistic connection. First, the increase in DNA-bound RPA after ATR inhibition could be explained by hyper-accumulation of stalled forks following unscheduled firing of dormant origins (Ge et al., 2007; Ibarra et al., 2008). Second, the rate- limiting nature of RPA would only manifest after excessive origin activity had time to deplete all available RPA, explaining the temporal gap between ATR inhibition and DSB appearance.

To test these predictions, we exploited the paradigm that origin firing requires CDK2 (Gillespie and Blow, 2010). Consistently, application of roscovitine, a broad CDK inhibitor widely used to suppress origin activity, abolished both excessive RPA loading and fork breakage in cells treated with HU and ATRi (Figs. 3A, 3B, . These results were reproduced by inhibiting other regulators of origin firing such as CDC7 or CDC45 (Figs. 3A, 3B, )..The relatively high concentration of HU in all these experiments allowed only minimal fork progression, indicating that the observed effects of roscovitine, CDC7 inhibition, or CDC45 depletion are related to new origin firing. Importantly, substitution of roscovitine by a specific inhibitor of mitotic CDK1 did not prevent fork breakage in HU and ATRi-treated cells , ruling out the possibility of premature mobilization of structure specific nucleases, which are known to be hyperactivated in mitosis (Matos et al., 2013).

The above results prompted us to ask why some cells tend to exhaust RPA earlier than others. Given the tight correlation between the number of active origins and sequestration of RPA to stalled forks, we hypothesized that cells with the highest replication activity would be the first ones to reach the RPA threshold. Indeed, monitoring of H2AX phosphorylation during the cell cycle revealed that fork breakage after HU and ATRi treatment occurred first in cells with the highest replication rate . Together, these data suggest that in the absence of ATR, the coordinated fork breakage occurs during S phase after ssDNA generated by unscheduled origin firing exceeds the nuclear pool of RPA.

15. Example 5

Increased origin activity triggers fork breakage in ATR- proficient cells

In the above experiments, fork breakage was assayed under conditions when HU was combined with ATR inhibition, leaving the possibility that an unknown ATR effector might exert its protective function directly at stalled forks. To address this, we asked whether ATR could have prevented breakage of stalled forks in cells primed to increase the number of active forks above the physiological threshold. We forced otherwise unstressed cells to fire extra origins by transiently inhibiting ATR but without any additional replication stress. Indeed, DNA fiber analysis showed that the number of active forks increased after incubation with ATRi and the unscheduled replication was also visible by QIBC as higher levels of chromatin-loaded RPA. Importantly, the time and dose of the ATRi under these conditions did not cause DNA breakage . We then washed away the ATRi and incubated, or not, cells with HU. Strikingly, whereas in the absence of replication stress cells continued to replicate normally, addition of HU rapidly exhausted the residual RPA pool and caused DNA breakage , . ATR activity recovered back to normal after washing out the inhibitor and roscovitine no longer prevented DSB generation, because the majority of available origins had already fired during the priming step . Thus, the unscheduled origin firing in ATR-deficient cells appears to be the major source of RPA depletion and the ensuing fork breakage.

16. Example 6

Exhaustions of RPA triggers fork breakage in all active replication factories

Replication of eukaryotic genomes takes place in replication factories manifesting as nuclear foci that contain both active replicons and dormant origins (Gillespie and Blow, 2010). Given this spatial organization and the emerging evidence that RPA is rate-limiting for protecting stalled forks anywhere in the nucleus, our current model predicted that after all RPA becomes sequestered, unprotected ssDNA should break in all active replication factories. Indeed, this was what we observed. Firstly, in a time-lapse analysis of RPA2-EGFP cells treated with HU and ATRi we found that RPA progressively accumulated in the same nuclear foci, indicating that the unscheduled origin firing is largely confined to replication factories that were active already at the beginning of the replication stress . This was confirmed by comparing the spatial distribution of replication activity before and at the end of replication stress using endogenous RPA . Secondly, the hallmarks of fork breakage such as RPA phosphorylation on S4/8 or T21 occurred simultaneously in all replication factories and regardless of the total extent of damage. Thus, breakage of stalled forks in a given nucleus after global RPA exhaustion occurs in all replication factories active at the time of replication stress. 17. Example 7

Partial knockdown of RPA accelerates breakage of stalled forks

To validate these conclusions, we reasoned that lowering RPA levels should render cells more sensitive to excessive origin firing. We partially knocked down RPAl to the degree that did not impair DNA replication, preserved normal ATR activity, and did not cause DNA damage . Strikingly however, after applying replication stress by combined HU and ATRi treatment, reduction of RPA levels progressively increased the fraction of cells with DSBs. We verified that the dynamics of DSB generation closely correlated with accumulation of unprotected ssDNA, consistent with the notion that DNA damage is triggered after ssDNA exceeds RPA buffering capacity. Of note, the accelerated fork breakage occurred in cells with inhibited ATR, indicating that the ability of RPA to shield stalled forks is an autonomous mechanism that counteracts DNA breakage.

18. Example 8

Stoichiometric increase of the RPA subunits delays fork breakage

A reverse prediction was that a surplus of RPA should extend the dynamic range within which ATR-deficient cells escape fork breakage. To this end, we generated 'SuperRPA' cell lines stably expressing 2- to 3-fold excess of all three RPA subunits, expressed from the same transcript in a stoichiometric fashion,. Such degree of RPA overexpression did not alter DNA replication or ATR activation. Strikingly however, these cells dramatically extended the dynamic range of RPA loading on chromatin and became remarkably resilient to dNTP depletion and ATR inhibition at time points where cells with normal RPA levels underwent massive fork breakage. Furthermore, the acquired resilience of the SuperRPA cells was associated with reduced formation of unprotected ssDNA . Since all these experiments were performed in cells with inhibited ATR, we conclude that as long as cells contain sufficient levels of RPA complex, it can be deployed to ssDNA and replication forks remain stable irrespective of ATR signaling. Of note, single overexpression of RPAl subunits, or other ssDNA-binding proteins such as hSSBl or Rad51 did not prevent fork breakage after RPA depletion (our unpublished observation), indicating that the main cellular activity that can effectively shield replication intermediates against breakage is the fully assembled RPA complex, likely due to its exquisitely high affinity for ssDNA. 19. Example 9

Exhaustion of the RPA pool generates irreversible damage to replication factories

Previous work indicated that forks exposed to replication stress eventually lose the ability to restart (Jossen and Bermejo, 2013). To test whether RPA exhaustion can explain this, we monitored the DSB dynamics in cells that were exposed to HU and ATRi for 2.5 h (when the majority of S-phase cells exhausted RPA and underwent DNA breakage) and then released by removing both drugs . Strikingly, 72 h after the release, a fraction of cells still contained an unusually high number of RPA foci (100 foci per cell in average) indicating the presence of unrepaired forks . After excluding contribution of undamaged cells, we could see that the fraction of cells with aberrant forks was very similar to that detected at the time of release from the HU/ATRi treatment (Figs. S5A, S5B), suggesting that all such cells suffered a permanent proliferation blockade linked to irreparable fork breakage. Indeed, although cells with hallmarks of unrepaired forks could resume progression through S phase, likely due to replicons that were activated only after releasing from replication stress , they all arrested at the G2-M boundary and eventually presented morphological features of senescence such as dramatic increase in size of cell nuclei. Thus, RPA exhaustion marks a 'point of no return' for cell proliferation.

To test whether such 'point of no return' is reached exactly at the time of RPA exhaustion, we incubated cells with HU and ATRi for 40 minutes when RPA loading approaches its threshold yet still without DNA breakage (Fig. 2A), and then removed the drugs to stop any further RPA depletion. Under such conditions, RPA was rapidly unloaded from the chromatin whereas, when we extended the

HU/ATRi treatment beyond the point of fork breakage (Fig. 2B), an increasing fraction of cells failed to dissolve the RPA foci and arrested. Significantly, elevation of the RPA complex in the SuperRPA cell lines delayed this form of irreversible fork breakage and allowed recovery from the extended stress despite the overall ssDNA accumulation was 2- to 4-fold higher than in naive cells. These results further support the model that RPA exhaustion, and not just an accumulation of supra-physiological levels of ssDNA, marks the 'point of no return'. Furthermore, because ATR was inhibited in these experiments, the excess of RPA appears necessary and sufficient to shield supernumerary forks against breakage. We reasoned that the accumulation of damaged cells in G2 is due to persistent checkpoint signaling from DNA lesions generated on stalled forks deprived from the RPA protection. Indicative of the presence of chromosome breaks, we observed cases of massive nuclear fragmentation in damaged cells that attempted to enter mitosis. To validate that these lesions are bona fide DSBs, we pulse- labeled active forks by EdU, and then applied calyculin A, a phosphatase inhibitor that has been used as a tool to overcome cell cycle checkpoints and induce premature chromosome condensation (El Achkar, et al., 2005). Remarkably, virtually all metaphase spreads obtained after forcing the G2-arrested cells to mitosis were composed of shattered chromosomes with multiple breaks.

Furthermore, the EdU signal was localized predominantly at sites with hallmarks of chromosome breaks such as terminal parts of shattered chromosomes or boundaries between fused chromosome fragments, supporting the conclusion that RPA exhaustion leads to the simultaneous breakage of active forks. Because of its fatal consequences, we henceforth refer to this event as 'replication catastrophe' (RC).

20. Example 10

RPA exhaustion is a common denominator of replication catastrophe regardless of the source of replication stress or unscheduled origin firing

A potential caveat of the previous results is that these were achieved under conditions where replication stress imposed by HU was combined with

unscheduled origin firing induced by ATRi. To validate that RPA shields stalled forks in a broader biological context, we revisited all major predictions by testing either drug alone and by using independent replication inhibitors and origin regulators. We first reasoned that if the model is generally applicable, RPA should be rate-limiting for fork breakage in cells that are treated with HU or ATRi alone, because both treatments lead to steady generation of ssDNA (by helicase and polymerase uncoupling, and unscheduled origin firing, respectively). Indeed, the increasing reduction of nuclear RPA levels lead to progressive fork breakage under these conditions (Fig. 6A, 6B,). Thus, 'single hits' of replication perturbation can eventually deplete the RPA pool and combining them merely accelerates RPA depletion and the ensuing fork breakage.

The next prediction was that RPA should be rate limiting for active fork protection regardless of the sources of replication stress and/or unscheduled origin firing. Also this was confirmed by experiments showing that independent inducers of fork stalling (aphidicolin, gemcitabine, citarabine, UV light)

phenocopied HU by synergizing with ATRi in progressively depleting RPA up to the point of RC (Fig. 6C, . Likewise, the ATRi-induced RPA exhaustion was reproduced not only by inhibiting or depleting CHKl (the downstream component of the ATR pathway), but also by inhibiting WEE1, another suppressor of unscheduled origin firing . As predicted by the model, RC after CHKl or WEE1 inhibition was mitigated by simultaneous suppression of origin firing, and it showed strong correlation with the RPA levels when applied alone, as well as when combined with distinct forms of replications stress. Finally, the EdU-flanked chromosome breaks on calyculin A-induced metaphase spreads confirmed that the alternative sources of replication stress caused DSBs at active forks. Thus, exhaustion of nuclear RPA represents a biological disaster for replicating genomes regardless of the sources of replication stress or unscheduled origin firing.

21. Example 11- Flow cytometry.

1 to 2 million cells are provided per condition.

Samples are prepared as follows:

Cells are spun down at lOOOg for 2 min.

Cells are washed PBS and again spun down at lOOOg 2min

The supernatant is discarded and the cell pellet is re-suspended in ice-cold PBS containing 0.2% (vol/vol) Triton X-100 for 1 minute. The cells are spun down again for 2 minutes at 2000g

The pellet is re-suspended first in 1 volume of PBS and 3 volumes of 4%

formaldehyde are added for fixation, and incubated for 12 minutes at room temperature.

The cells are spun down at 2000 g for 2 minutes.

The cell pellet is washed twice in PBS. At this point the sample may be stored at 4°C, or may continue directly with staining.

Staining

The cells are spun down at 2000 g for 2 minutes. Cell pellet is re-suspended in blocking media (filtered DMEM + 10% FBS) containing the primary antibody diluted and incubated for 1 hour at room temperature.

The cells are washed twice with PBS, and the cell pellet is re-suspended in blocking media (filtered DMEM + 10% FBS) containing the secondary antibody diluted and DAPI at a final concentration of 1 microgram/ml, incubated for 1 hour at room temperature.

Acquisition of sample data is done by flow cytometry.

The sample data acquired is processed and the number of cells in each staining quantified.

A scatter image of the cell population identifying which cells in the cell population comprise labelled DB-RPA, and simultaneously identifying the amount of labelled DB-RPA in the cell by intensity may be generated.

22. Example 12- Cells in suspension

Samples are prepared as for Example 11. After staining the cells are re-suspended in sterile filtered water and a drop of cells placed on a glass slide. Water is left to dry out and 10 microliters of Mounting Media (Mowiol, Vectashield, others) are added and a coverslip placed on top.

Image Acquisition is performed as described for QBIC method, which method comprises the steps

i. Acquiring images of the cell population

ii. Processing the acquired images by software-based image analysis iii. Quantifying the signal from the labelled compounds, specifically RPA,

nucleus and damaged DNA

iv. Optionally exporting the data of the quantified signal to software to

generate visualization

In a further alternative, the cells are re-suspended in 96 well plate and either spun down or allowed to settle, and image acquisition is performed from this plate. 23. Example 13- Method of classifying a compound of interest

The protocol below is performed in 96- well plates (well volume is max 300 microl). In each well the volume of medium used is usually 100 to 200 microl). However, the protocol is scalable to 384 or even 1536 well plates.

All steps are preferably performed in sterile conditions. In the case where the equipment cannot be used in sterile conditions, drug solution preparation and transfer can be performed in non-sterile conditions (e.g. aseptic conditions). This introduces no risk of biological contamination of the assay, as incubations are carried out only for a maximum of few hours and cells are fixed immediately after. Day 0

Cells are seeded in imaging (black) 96 well plates. The appropriate number of cells corresponds to a number where cells will be in exponential growth at the time of the experiment and will depend on the cell line used. For U20S (human bone osteosarcoma) cells, 20000 cells are seeded per well in 100 ul medium. Day 1

Appropriate dilutions of the compounds are prepared first separately in cell culture medium.

Stock solutions of the compound are lOmiliM or ImiliM, so a 1/100 dilution (transferring 1 microliter to 100 microliters) will give 100 or 10 microM.

Preparing the dilutions separately allows for proper homogenization of the solutions, avoiding unnecessary manipulation and stress for the cells. Once prepared, the media is added to the cells and incubated right away.

Dilutions are prepared in the same 96 well format, in normal plastic sterile culture plates.

For each plate, positive and negative controls are included. Negative control is fresh medium + drug solvent (usually DMSO). Positive controls are for example 2 to 4 reference compounds. The reference compounds are known to induce a consistent response (and the corresponding data by QIBC). The number of total positive and negative controls per plate will depend on the number of compounds in the plate to be tested. Optimal distribution of drugs in plates should allow for at least 4 wells left for controls, ideally 8. This leaves space for 90 to 94 test drugs per plate. This can be scaled accordingly to 384 and 1536. Multichannel (8, 12 or 96) pipettes are recommended to transfer and operate

Different concentrations are tested, prepare a sufficient surplus of medium with the main concentration and do subsequent dilutions in fresh medium to achieve the lower concentrations wished.

If combinations of two drugs are to be tested, prepare solutions at 2x final concentration, which after mixing will give lx. Beware of the maximum volume admitted per well (300 microliters in 96well plate). Ideally use 100 + 100 microliters.

The appropriately diluted compounds are incubated in cell culture incubator for a time sufficient to warm the medium and equilibrate the pH, typically about 20 minutes.

The medium is removed from the cells, and the compound dilutes are added instead.

The cells are incubated with the compound solutions for the designated times. At the end of the incubation, the medium is removed, and the samples prepared as previously described, i.e. free nucleoplasms RPA is removed, and the cells are fixed.

The remaining steps can be done subsequently in the same day.

If labeling and data acquisition cannot be performed in the same day, cells can be stored :

After fixation in PBS at 4-8 degrees for a maximum of 2 weeks.

After labeling in PBS at 4-8 degrees for a maximum of 1 week.

In any case, cells should be preferably stored after fixation (A), waiting for the labeling to be carried out as closest as possible to the time of data acquisition. Labelling of DB-RPA and DNA damage (such as DSB), quantification and data acquisition is performed as described previously. The data for DB-RPA is correlated to the data for DNA damage by plotting the DB-RPA data for a single cell on the x-axis and the corresponding DNA damage data for said cell on the y- axis. This generates a scatter diagram which visualizes the relationship between the amount of DB-RPA and the amount of DNA damage in the cells of the population. See for example Figure 11 A, for a scatter diagram indicative of replication catastrophe in a cell population.

The resulting scatter diagram is compared to scatter diagrams generated for Reference compounds, and it is deduced based on the similarity of the scatter diagram for the compounds of interest to the scatter diagrams for Reference compounds, which class of compounds the compound of interest belongs to.

See Figure 11 for examples of scatter diagrams generated for Reference compounds, or in the case of 11 A, a mixture of compounds.

It can be seen that the Reference compounds display distinct scatter plot patterns, and this is indicative of the molecular target and/or pathways in the cell involved in the response. Thus, any compound which when tested displays a scatter plot the same as or similar to the reference compound, can be classified as similar to e.g. Reference A, and inferences made as to the mode of action of the compound of interest.

In Figure 12, scatter diagrams for compounds X and Y ( figures 12 B and C respectively), display the same pattern as for the Reference A, shown in figure 12 A and 11 B.