The present invention relates to a method for determining whether a subject is resistant to certain types of chemotherapy, specifically chemotherapy in which DNA damaging agents are used.

Chemotherapy is a widely used treatment for many cancer types, but not all patients who would be expected to benefit from chemotherapy do so. Some patients exhibit the phenomenon of resistance to chemotherapy. When these patients are subjected to chemotherapeutic treatment, the expected effects of the chemotherapy (killing of cancerous cells) are not observed, or are observed to a lesser degree than would be expected. Whether or not a patient is resistant to chemotherapy is, however, generally only determined by carrying out the course of treatment and determining whether it has been effective. As chemotherapy is often associated with unpleasant side effects, it would be beneficial to be able to identify these patients before treatment is commenced. By doing so, it would be possible to avoid subjecting the patients to a treatment from which they will derive little or no benefit, and which may cause them to suffer side effects. The present invention addresses this need.

Resistance to chemotherapy may be inherent, or it may be acquired and much has been speculated about mechanisms by which resistance occurs or is acquired.

The present inventors postulated that resistance to chemotherapy using DNA damaging agents may be acquired by patients following exposure to irradiation, including the irradiation to which patients may be subjected during the course of their diagnosis with cancer. The inventors observed that irradiation of certain cells in vitro influences their subsequent response both to further challenge irradiation and to DNA damaging agents such as cisplatin. By investigating the mechanism underlying this effect and ways in which it can be modified or reversed, the inventors have identified a novel way of testing patients to determine whether they are resistant to chemotherapy in which DNA damaging agents are used, in other words whether chemo- resistance related to the mechanism removing HRS is activated.

Low dose hyper-radiosensitivity (HRS) is a phenomenon which has been observed in cells in culture in vitro and in vivo and is characterized by a high sensitivity per unit of low-linear energy transfer (LET) radiation dose for low doses of low LET radiation, for example doses below -0.5 Gy (Lambin et al., 1993; Marples and Joiner, 1993), although the precise dose will depend on the cell type, and the phenomenon may be observed at higher doses. In other words, HRS proficient cells, including cells that have the HRS phenotype, are more readily killed by low-LET radiation at these doses than would be expected. The HRS phenotype is thus characterised by an excess of cell killing at these doses of low LET radiation, relative to that which is predicted by the linear quadratic model (LQ-model). The LQ-model is described by the equation: S=exp(-ad- pd ² ) where S is the surviving fraction, d the dose and a and β the parameters describing the linear and quadratic component, respectively, of the intrinsic radiosensitivity (Lambin 1993, and Marples 1993). This is shown, for example in Figure 1A.

By low-LET radiation it is meant photon (gamma and X-ray), electron and proton radiation (Schettino, et al 2001). Any type of low LET radiation may be used.

HRS is in general observed at doses of low LET radiation of up to 0.6Gy, e.g. up to 0.5, 0.4, 0.3, 0.2, 0.1 , 0.05, 0.01 , or 0.005 Gy. A dose of low LET radiation that induces HRS can be any dose referred to above.

HRS may also be observed in response to high LET radiation if used under the appropriate conditions and doses (Xue et al 2009, Marples et al 1994). Some particles, which are normally considered to be high LET, will behave as low LET at very high energies on part of their track. These particles deposit small amounts of energy along their path until they reach what is called the Bragg peak where they stop up and deposit their rest energy within a short distance. This means that the LET is low along most of the track but very high in the Bragg peak. High LET radiation may thus also lead to HRS. Suitable types of high LET radiation that can lead to HRS in HRS proficient cells include all accelerated nuclei (this includes accelerated helium nuclei but not alpha-particles from decay processes) together with pi-mesons, electrons, gamma radiation, x-rays and protons (which can also be defined as accelerated nuclei).

Suitable doses of high LET radiation are doses of up to 0.6Gy, e.g. up to 0.5, 0.4, 0.3, 0.2, 0.1 , 0.05, 0.01 , or 0.005 Gy. A dose of high LET radiation that induces HRS can be any dose referred to above.

It can thus be seen that whilst the HRS phenotype has been largely characterised with respect to low LET radiation, high LET radiation can in some circumstances be used to induce the HRS phenotype in HRS proficient cells. As such, reference herein to radiation suitable to induce HRS or radiation suitable to induce the HRS phenotype, includes suitable doses of both low LET radiation and high LET radiation. As discussed elsewhere herein, the HRS phenotype has certain characteristics and it can readily be determined whether a certain dose and type of radiation gives rise to the HRS phenotype. Any dose and type of radiation that gives rise to the HRS phenotype in HRS proficient cells can be used in the various methods set out herein.

Preferred doses and types are set out above.

HRS is associated with a failure to induce the early (active 0-2 h post-irradiation) and transient (lasting -12 h) G2-phase checkpoint, which arrests cells irradiated in G2 before entering mitosis (Marples et al., 2003; Xu et al., 2002). HRS proficient cells that are exposed to an appropriate dose and type of radiation to induce HRS thus suffer damage to their DNA as a result of the exposure to radiation, but as the G2 phase checkpoint is not induced in these cells, they tend to leave G2 and to enter mitosis with unrepaired DNA damage. This leads to greater numbers of cells dying than would be predicted for these doses and types of radiation by the linear quadratic model (LQ-model), i.e. they die at levels higher than would be predicted by the linear quadratic model (LQ-model) for a given dose of the appropriate type. Alternatively stated, this leads to fewer cells surviving than would be predicted for these doses and types of radiation by the LQ-model, i.e. they survive at levels lower than would be predicted for these doses and types of radiation by the LQ-model.

HRS is thus associated with higher levels of cell killing at the appropriate dose and using an appropriate type of radiation than would be predicted by the LQ-model. Alternatively defined, HRS is thus associated with lower levels of cell survival at the appropriate dose and using an appropriate type of radiation than would be predicted by the LQ-model. This can be assessed to determine whether the HRS phenotype is present in a cell, i.e. whether a cell is HRS proficient. Exposing cells to one or more appropriate doses of radiation of the appropriate type, and determining the effect of this radiation on cell killing and/or cell survival, using standard techniques, thus allows one to determine whether cells are HRS proficient.

Furthermore, if a cell type is known to be HRS proficient, cells can be exposed to different test conditions and the level of killing or survival in response to an appropriate dose and type of radiation can be compared based on the different test conditions. This allows a determination of whether the test conditions have any influence on the HRS phenotype, e.g. whether the HRS phenotype is reduced. Figure 1A shows for example the response of cells to different low LET doses of radiation. The Figure compares cell survival for HRS proficient cells that have not been pretreated before exposure to the radiation, to those that have been pretreated with low dose rate (LDR) irradiation at 0.22Gy/h. It is clear from Figure 1 A that the pretreatment carried out in this experiment influences the HRS phenotype in these cells, as fewer of the pretreated cells are killed at the relevant low level LET doses of radiation than those which have not been pretreated. In fact in this case under the conditions used the HRS phenotype is removed since the pretreated cells behave in accordance with the prediction of the LQ model. This particular aspect of the HRS phenotype is referred to herein as the cell survival aspect of the HRS phenotype.

As noted above, HRS proficient cells that are exposed to doses and types of radiation that induce the HRS phenotype tend to enter mitosis even if they have damaged DNA.

Mitosis is the process during the cell cycle by which a eukaryotic cell separates the

chromosomes in its cell nucleus into two identical sets in two nuclei. During the cell cycle mitosis alternates with interphase (G1 , S and G2, where the cell prepares itself for cell division). As shown for example in Figure 1 B, cells with the HRS phenotype will tend to enter mitosis more readily following exposure to appropriate doses of low LET radiation than cells without the HRS phenotype. This Figure also shows an example of the characteristic profile of mitosis at increasing doses of low LET radiation that is seen in cells with the HRS phenotype. This Figure compares the mitotic ratio (irradiated/unirradiated) for HRS proficient cells that have not been pretreated before exposure to the radiation, and those that have been pretreated with low dose rate irradiation at 0.22Gy/h. The pretreatment, as discussed above, removes the HRS phenotype. It can be seen that the rate of mitosis in cells with the HRS phenotype is not reduced by the lowest levels of radiation, whereas in cells without the HRS phenotype the rate of mitosis decreases at the very lowest levels of radiation. Only once a threshold level of radiation is reached does the rate of mitosis start to decrease in the cells with the HRS phenotype. This is characteristic of HRS cells (Krueger 2007).

This particular aspect of the HRS phenotype is referred to herein as the mitotic ratio aspect of the HRS phenotype.

Thus in general terms, there is a phenotype characteristic of HRS. The HRS phenotype is observed after exposure of the HRS proficient cells to appropriate types and doses of radiation. The HRS phenotype can be expressed as a characteristic profile of rates of mitosis at different doses of radiation (e.g. at different doses of low LET radiation) (including the presence of a threshold dose above which a decrease in mitosis is seen and below which no decrease in mitosis is seen) and also in a greater than expected level of cell killing following exposure to an appropriate type of radiation at a given dose. These characteristic properties of the HRS phenotype can be assayed for using standard techniques. They provide information about whether a cell has the HRS phenotype and also, in a cell which is known to have the HRS phenotype, whether any test conditions to which the cell is subjected before the dose of radiation of the appropriate type have an effect on the HRS phenotype. Indeed, it is known that the HRS phenotype can be reduced or reversed and that any reduction or reversal in the HRS phenotype can be permanent or transient. The inventors have previously shown that the HRS phenotype can be reversed or reduced in vitro by exposing HRS-proficient cell cultures to a high dose-rate (HDR) pre-exposure of 0.2-0.3 Gy. This transiently abolishes the HRS-response to subsequent challenge irradiation (i.e. to irradiation of a dose and type suitable to induce the HRS phenotype in HRS proficient cells) ((Edin et al., 2007; Joiner et al., 1996; Marples and Joiner, 1995; Short et al., 2001 ; Wouters and Skarsgard, 1997). By reducing the dose-rate to 0.3 Gy/h (LDR) the effect on HRS in T-47D breast cancer cells of a preexposure of 0.3 Gy becomes permanent (Edin et al., 2007).

Transient removal of the HRS phenotype was achieved in cells which received conditioned medium from cells exposed to LDR irradiation (Edin et al., 2009b). LDR irradiation (0.3 Gy) induced the same factor in cell conditioned medium without cells present during irradiation (Edin et al., 2009a), as long as the cells were cultured in medium with foetal bovine serum (Edin et al., 2009a).

In view of the fact that the HRS phenotype could be removed in cells by applying conditioned medium, the inventors postulated that an active factor conferring this removal of HRS must be present in this medium. In work leading to the present invention, the inventors have identified the mechanism involved in the permanent removal of the HRS phenotype by LDR radiation and have shown that the active factor that is present in such conditioned medium and which removes the HRS phenotype is TGFP3. Indeed the effect of reversing or reducing the HRS phenotype can be achieved by exposing HRS proficient cells to TGFp3 (Figure 5). TGFp3 is present intracellular^ in a complex with the protein LAP3 by covalent bonds under normal conditions, but is present as free TGFp3 (i.e. not complexed with LAP3) in conditioned medium from cells which have been treated to remove or reduce the HRS phenotype, as set out above.

In addition to the above, the inventors have furthermore shown that serum from a mouse exposed in vivo to LDR radiation also has the ability to remove the HRS phenotype (see

Examples 9 and 13). This indicates that the same factor may be being induced in vivo in response to LDR radiation as in vitro. The data in Example 13 furthermore indicates that the serum from the irradiated mouse contains active TGFp3 and that this removed HRS in HRS proficient cells. The iNOS inhibitor furthermore prevented the effects of TGFp3 on the HRS proficient cells. Furthermore it has been shown by the inventors that pretreatment of HRS proficient cells with doses of radiation that remove the HRS phenotype in vitro will also cause the cells to have resistance to the DNA damaging chemotherapeutic agent cisplatin (see Example 10).

The inventors have thus studied the mechanism by which removal or reduction of HRS occurs in vitro and have identified an active factor that has the ability to remove or reduce the HRS phenotype when applied to HRS proficient cells in vitro. They have furthermore demonstrated that the factor is induced in vivo. Importantly they have also demonstrated that the pretreatment with radiation to cause removal of the HRS phenotype in cells in vitro (and which hence also causes the presence of the active factor in the medium) also causes the cells to be resistant to the DNA damaging chemotherapeutic agent cisplatin.

The fact that the conditions which cause removal of the HRS phenotype in cells in vitro (the phenotype as noted above being associated with an increase in cell killing at a given dose of low LET radiation) also cause resistance to cell killing by a DNA damaging agent such as cisplatin suggests that a common mechanism may be in action. This is consistent with the belief that individuals may obtain resistance to chemotherapy with DNA damaging agents as a result of exposure to radiation and the present invention thus seeks to identify these individuals. The common mechanism by which resistance to cell killing by DNA damaging agents and removal of HRS involves the induction of an active extracellular factor, the presence of which has the ability to reduce the HRS phenotype. Experiments set out in the Examples demonstrate that medium containing the active factor, either as a component of conditioned medium or when added in purified form to the growth medium can remove or reduce the HRS phenotype. It is thus considered to be possible to determine resistance to cell killing (e.g. by DNA damaging agents) by determining whether a sample from a subject has the ability to reduce the HRS phenotype.

Thus in a first embodiment of the invention, we provide a method of determining whether a subject is resistant to chemotherapy with a DNA damaging agent or DNA damaging agents, said method comprising determining whether a sample from said subject is capable of reducing the low dose hyper-radiosensitivity (HRS) phenotype in HRS proficient cells. If the sample from said subject is capable of reducing the low dose hyper-radiosensitivity (HRS) phenotype, this indicates that the subject is resistant to chemotherapy with a DNA damaging agent or DNA damaging agents.

In practice the step of determining whether a sample from the subject is capable of reducing the HRS phenotype in HRS proficient cells can be achieved in a number of different ways. ln one embodiment this is achieved by determining whether a sample from the subject is capable of reducing the HRS phenotype in HRS proficient cells, wherein the HRS phenotype is determined on the basis of the cell survival aspect of the phenotype.

In other words it is determined whether the sample from the subject has the ability to reduce cell killing or increase cell survival in HRS proficient cells in response to the appropriate radiation dose and type.

As discussed above, it is straightforward to determine the amount or level of cell killing or cell survival in HRS proficient cells in response to an appropriate dose and type of radiation. In the most basic strategies, the cells are exposed to an appropriate dose and type of radiation and after an appropriate period of time, viable remaining cells (or colonies of a predetermined size) are detected and enumerated (e.g. counted). Thus it is possible to simply determine the number of remaining viable cells or colonies of a predetermined size. The number of cells which were initially exposed to the an appropriate dose and type of radiation may be known or calculated and this enables the number of viable remaining cells to be expressed as a proportion of the number of cells which were initially exposed to the appropriate dose and type of radiation.

To determine the amount or level of cell killing or cell survival in response to the radiation, the total number of surviving cells may be enumerated, or else cell colonies of a certain size may be enumerated. The proportion of the cells which have been subjected to the appropriate dose of radiation of the appropriate type and which retain their capability to grow and divide is referred to as the "surviving fraction". The value for the surviving fraction is higher where more cells have survived, i.e. when cell killing is at lower levels. The value for the surviving fraction is lower where fewer cells have survived, i.e. when cell killing is at higher levels. The number of cells that are subjected to the appropriate dose and type of radiation is known, or can be calculated. The number of surviving cells or colonies can then be expressed as a proportion of this.

The surviving fraction can also be expressed relative to the proportion of cells surviving in a control group of cell which has not been exposed to radiation. The surviving fraction for each dose can be plotted graphically against the relevant radiation dose. If the surviving fraction is lower than would be predicted by the LQ-model for the relevant doses of radiation (e.g. doses of less than 0.6Gy (e.g. less than, 0.5, 0.4, 0.3, 0.2, 0.1Gy)), then this is indicative of the presence of the HRS phenotype. The term "surviving fraction" is understood in the art and thus provides a measure of the proportion of the cells which have been subjected to the relevant test conditions (in this case appropriate dose and type of radiation) which retain their capability to grow and divide. The "surviving fraction" of cells may be determined after an appropriate period of time (e.g. at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 , 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 , 28, 35 days, preferably 2-35, 4-28, 6-21 , 8-20, 9-18 or 10-14 days).

It will therefore be appreciated that the HRS phenotype can be assessed experimentally by subjecting cells to the appropriate type and dose of radiation, and observing the effect of this radiation on cell survival after an appropriate period of time. This can be carried out using radiation of the appropriate type and dose at a single dose or at multiple (at least 2, 3, 4, 5 or 6) doses.

The determination of the surviving fraction can be carried out by any appropriate method, and suitable methods are known in the art. By way of example for adherent cells, the surviving fraction can be determined by counting or otherwise determining the number of cells or colonies present (e.g. per flask or per appropriate area of culture vessel (cm ² ) (Pettersen et al., 1973)) at the appropriate time point following the treatment. To assist in this, it is usual although not necessary to fix the cells, e.g. using alcohol. The cells may also be stained to assist in determining the relevant numbers e.g. with methylene blue. Colonies containing more than a predetermined number of cells (e.g. at least 10, 20, 30, 40, 50, 60, 70, 80, 90, 100) are then scored as survivors. Increased mean multiplicity per colony-forming unit during the interval between seeding and irradiation can then corrected for according to known methods (see for example Gillespie et al., 1975).

Thus in a preferred embodiment the invention provides a method of determining whether a subject is resistant to chemotherapy with a DNA damaging agent or DNA damaging agents, said method comprising determining the level of cell killing or cell survival in HRS proficient cells which have been exposed to a sample from said subject, and exposed to an appropriate type and dose of radiation to induce the HRS phenotype, determining the level of cell killing or cell survival in HRS proficient cells which have not been exposed to a sample from said subject, but which have been exposed to an appropriate type and dose of radiation to induce HRS, and comparing the values obtained. An increase in cell survival or reduction in cell killing following contact with the sample is indicative of resistance to chemotherapy with DNA damaging agents.

In a further embodiment, the step of determining whether a sample from the subject is capable of reducing the HRS phenotype can be achieved by determining whether a sample from the subject is capable of reducing the HRS phenotype in HRS proficient cells, wherein the HRS phenotype is determined on the basis of the mitotic ratio aspect of the phenotype.

In other words it is determined whether the sample from the subject has the ability to reduce the level of mitosis in HRS proficient cells observed when the cells are exposed to the appropriate dose and type of radiation. As discussed in more detail below, this is measured by measuring the "mitotic ratio" and/or the "mitotic ratio (irradiated/unirradiated)".

In any cell population, the number of cells that are undergoing mitosis can be determined and expressed as a proportion of the total number of cells present. This is referred to herein as the mitotic ratio and this can be determined for any given cell population. A mitotic ratio of 1 indicates that all cells are undergoing mitosis. A mitotic ratio of 0 indicates that no cells are undergoing mitosis.

Asynchronously growing cells are in general used in assays to determine whether a given test condition will influence mitosis. Such cells will tend to have a constant proportion of cells in mitosis, but to allow for any variation in this background level of mitosis during an experiment, the mitotic ratio for cells that are not subjected to test conditions can be determined (these are termed control cells). The mitotic ratio for the cells that have been subjected to test conditions can then be expressed as a proportion of the value obtained for the control cells. This is referred to herein as the "mitotic ratio (test cells/control cells)". As such, when determining the effect of radiation of the appropriate dose and type to induce HRS on cells (e.g. asynchronously growing cells) the mitotic ratio for non irradiated cells can be determined, as well as the mitotic ratio for irradiated cells. The mitotic ratio for the irradiated cells can then be expressed as a proportion of the value obtained for the non irradiated cells. This is referred to herein as the "mitotic ratio (irradiated cells/non irradiated cells)". Both the mitotic ratio for the irradiated cells and the mitotic ratio (irradiated cells/non irradiated cells) are informative as regards the HRS phenotype.

Whilst cells in general will show a trend of decreasing mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) as the dose of radiation (e.g. low LET radiation) to which they are exposed increases, it is recognised that HRS-proficient cells have a threshold dose above which the decrease in mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) starts. Below the threshold dose there is no or substantially no decrease in mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells). This is a reflection of the tendency of HRS-proficient cells to enter mitosis even with damaged DNA. In other words, in HRS-proficient cells, no decrease in mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) is observed at doses of radiation of the appropriate type to induce the HRS phenotype that are below the threshold level, while a decrease in mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) is seen even in response to the lowest doses of radiation in HRS-deficient cells (Krueger et al 2007).

Whilst this threshold value (the value of the dose of a radiation of the appropriate type to induce the HRS phenotype above which a decrease in mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) is observed) will be different for different cells, it can be determined experimentally.

The presence of a threshold dose of radiation of the appropriate type to induce the HRS phenotype, above which a decrease in mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) is observed, is thus indicative of the HRS phenotype. It can thus be determined whether a given cell or cell type is HRS proficient on the basis of the presence of threshold dose of radiation of the appropriate type to induce the HRS phenotype, above which a decrease in mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) is observed. Cells that exhibit HRS have a significant initial low-dose threshold before a decrease in mitotic ratio is seen with increasing doses of radiation of the appropriate type to induce the HRS phenotype. Cell lines without HRS show an immediate decrease in mitotic ratio with the increased dose of radiation of the appropriate type to induce the HRS phenotype.

The presence of a threshold value can be assessed to determine whether the HRS phenotype is present in a cell, i.e. whether a cell is HRS proficient. Exposing cells to appropriate doses and types of radiation, and determining the effect of the radiation on the mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) at several (e.g. 2 or more than 2, e.g. at least 3, 4, 5, 6) doses of low radiation of the appropriate type to induce the HRS phenotype, using standard techniques, thus allows one to determine if a threshold value is present and hence whether the cells are HRS proficient.

If it is known that a cell type is HRS proficient, the mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) in response to an appropriate dose and type of radiation can be determined following exposure of the cells to different test conditions to determine whether these test conditions have any influence on the HRS phenotype, i.e. whether the test conditions can reduce the HRS phenotype. It is demonstrated (see Figure 1 B) that conditions which cause a reduction of the HRS phenotype cause a decrease in the mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) in response to an appropriate dose of low LET radiation. As such, comparing the mitotic ratio of HRS proficient cells that have been exposed to a certain test condition before being subsequently subjected to an appropriate dose of radiation of the appropriate type to induce the HRS phenotype to the mitotic ratio of HRS proficient cells that were not exposed to a certain test condition before being subsequently subjected to an appropriate dose of radiation of the appropriate type to induce the HRS phenotype will allow a determination of whether that test condition influences the HRS phenotype. Similarly, comparing the mitotic ratio (irradiated cells/non irradiated cells) of HRS proficient cells that were exposed to a certain test condition before being subsequently subjected to an appropriate dose of radiation of the appropriate type to induce the HRS phenotype to the mitotic ratio (irradiated cells/non irradiated cells) of HRS proficient cells that were not exposed to a certain test condition before being subsequently subjected to an appropriate dose of radiation of the appropriate type to induce the HRS phenotype will allow a determination of whether that test condition influences the HRS phenotype

Methods for achieving this are discussed in more detail below.

The number of cells that are undergoing mitotis in any given population of cells provides can be determined. Various ways exist of detecting cells that are undergoing the process of mitotis, including the detection of markers of mitosis. Markers that are found in cells undergoing mitosis and not in cells at other phases of the cell cycle are thus referred to herein as mitosis markers. Markers that are found at higher levels (e.g. at least 1.1 , 1.25, 2, 2.5, 3, 5, 10, 20, 50, 100 or 1000 fold higher levels) in cells undergoing mitosis than in cells at other phases of the cell cycle are also considered to be mitosis markers.

These markers may be proteins, glycoproteins or polypeptides (which may be adhesion molecules, enzymes, receptors, signalling molecules, structural proteins) polysaccharides, nucleic acid molecules, lipids (e.g. phospholipids, glycolipids or other components of the plasma membrane). Such markers may be found at the surface of the relevant cell (e.g. within, attached to or associated with the cell membrane), or may be found within the relevant cell. Such markers can be detected, and if necessary quantitated using standard techniques that are well known in the art. The markers may also be specific forms of a protein, which are found only at the relevant stage of the cell cycle, e.g. phosphorylated forms of proteins. One particular example is histone 3 which is phosphorylated only during mitosis.

A cell undergoing mitosis can thus be detected by detecting a marker of mitosis. One such marker is phosphorylated histone 3. Histone 3 is specifically phosphorylated during mitosis and on exit of mitosis a global dephosphorylation of histone 3 takes place. As such the proportion of cells in a cell population in which histone 3 phosphorylation is detected provides an indication of the proportion of cells in that population which are undergoing mitosis (the mitotic ratio). A comparison of this value in irradiated and unirradiated cells provides the mitotic ratio (irradiated cells/non irradiated cells).

Markers of mitosis can be detected by any appropriate means, including assays based on binding of molecules to said markers and detection of these molecules. For example immunoassays based on antibody detection of the marker of interest can be used, and such techniques are well known. Immunoassays which allow the number of cells having the marker of mitosis to be determined are preferred. Phosphorylated histone 3 for example can be detected by the antibody referred to in the Examples (Upstate Catalogue number 06-570) or other antibodies having the binding properties of this antibody. Any antibody that binds to phosphorylated histone 3, and preferably that binds specifically to phosphorylated histone 3 may be used. Examples of techniques for determining the number of cells having a marker of mitosis include flow cytometry.

The number of cells undergoing mitosis or expressing or having a marker or mitosis is thus determined and compared with the total number of cells or the total number of viable cells in the assay. This can be carried out on the basis of knowing the number of cells that were used in the assay or by determining the number of cells (or the total number of viable cells) that are present. Any method of determining the total number of cells (or the total number of viable cells) in the assay can be used. Methods to determine cell numbers are widely known and include automated methods such as flow cytometry. The cells may be stained with a stain or dye indicative of viable cells or indicative of non-viable cells (e.g. with PI or Hoechst dye) to assist in determining cell number or viable cell number.

Thus in a preferred embodiment the invention provides a method of determining whether a subject is resistant to chemotherapy with DNA damaging agents, said method comprising determining the mitotic ratio and/or the mitotic ratio (irradiated/unirradiated) in HRS proficient cells which have been exposed to a sample from said subject, and exposed to an appropriate type and dose of radiation to induce the HRS phenotype, determining the mitotic ratio and/or the mitotic ratio (irradiated/unirradiated) in HRS proficient cells which have not been exposed to a sample from said subject, but which have been exposed to an appropriate type and dose of radiation to induce HRS or the HRS phenotype , and comparing the values obtained. A reduction in mitotic ratio and/or mitotic ratio (irradiated/unirradiated) is indicative of resistance to chemotherapy with DNA damaging agents. In other words, the ability of a sample from a subject to reduce the mitotic ratio and/or mitotic ratio (irradiated/unirradiated) is indicative of said subject being resistant to chemotherapy with a DNA damaging agent or DNA damaging agents. In a further embodiment the step of determining whether a sample from the subject is capable of reducing the HRS phenotype can be achieved by determining reduction of the HRS phenotype on the basis of the presence of free TGFp3 in said sample (e.g. the presence of free

extracellular TGFp3).

The active factor that removes HRS has now been identified as TGFP3. Indeed the effect of reversing or reducing HRS can be achieved by exposing HRS proficient cells to TGFp3 (Figure 5). TGFp3 is present intracellularly in a complex with the protein LAP3 by covalent bonds under normal conditions, but is present as free TGFp3 (i.e. not complexed with LAP3) in conditioned medium from cells which have been treated to remove HRS as set out above. Although free TGFp3 can be found in the cytoplasm (see Figure 6), the presence of free TGFp3 (preferably free extracellular TGFp3) is indicative of resistance to chemotherapy.

The proposed mechanism by which free TGFp3 appears in the conditioned medium, i.e. by which free extracellular TGFp3 appears is set out in Figure 7. Without wishing to be bound by theory, the proposed mechanism suggests that IL-13 binds to receptor IL-13a2, which leads to upregulation of the enzyme induced NO-synthase (iNOS) and conversion of pro-furin to active furin. Furin proactivates the TGFP3-LAP3 complex by proteolytic cleavage of the covalent binding leaving TGFP3 non-covalently associated with LAP3. iNOS and the pro-activated TGFP3-LAP3 complex are secreted by the cell. LDR irradiation results in sustained reactive oxygen species (ROS) production which is required for nitric oxide (NO) production by iNOS. ROS (not necessarily sustained) also activates TGFP3 by breaking the binding to LAP3. NO then scavenges LAP3, preventing reconstitution of the TGFp3-LAP3 complex and resulting in free activated TGFP3. Free extracellular TGFp3 is believed to bind to its receptor in the recipient cell, resulting in removal of HRS e.g. in the recipient cell. If this process takes place in cell conditioned medium, TGFP3 is believed to bind to a receptor in the recipient cell, resulting in removal of HRS.

The presence of free TGFp3 (e.g. the presence of free extracellular TGFp3) in a sample is thus also indicative that a sample is capable of reduction of the HRS phenotype in a HRS proficient cell. Detection of free TGFp3 can thus also be used to indicate reduction of HRS phenotype in a HRS proficient cell. As noted above, TGFp3 can be found complexed with LAP3. The two proteins may be covalently attached or complexed (whereby the covalent bond has been broken but the two proteins remain associated). Detection of TGFp3 when it is covalently attached or complexed with LAP3 is not an indication of reduction of HRS phenotype in a HRS proficient cell. Detection of free TGFp3 is, on the other hand, an indication of reduction of HRS phenotype in a HRS proficient cell. By free TGFp3 it is meant TGFp3 which is not bound to, complexed with or physically attached to or associated with the protein LAP3.

As discussed above, a sample containing free TGFp3 can reduce the low dose hyper- radiosensitivity (HRS) phenotype in HRS proficient cells. As such, an alternative way of determining whether a sample can reduce the low dose hyper-radiosensitivity (HRS) phenotype in HRS proficient cells is to determine whether free TGFp3 is present in said sample. Free TGFp3 protein is detected.

It is important to note that free TGFp3 and not the TGFp3-LAP3 complex is indicative of the ability to reduce the low dose hyper-radiosensitivity (HRS) phenotype in HRS proficient cells. Any assay in which free TGFp3 can be detected can be used. This includes assays in which a molecule that binds to TGFp3 is used, and detection is based on detection of that molecule, directly (e.g. that molecule is labelled) or indirectly (e.g. by using a labelled antibody which binds to the first molecule). Such a binding molecule may be an antibody and in general

immunoassays will be used, i.e. assays which depend on binding of an antibody to TGFp3 present in the sample and the subsequent detection of said binding. Free TGFp3 is preferably detected using immunoassays such as ELISAs, lateral flow assays and in particular the assay set out in US2005/0003519, which is incorporated herein by reference.

The sample may be fractionated using standard techniques known in the art such as electrophoresis, chromatography, (e.g. HPLC, thin-layer chromatography, FPLC, gel filtration, desalting etc.) and the resulting fractions or isolates may then serve as samples in this assay. This would be useful to enrich the sample for the presence of TGFp3 to assist in the detection thereof.

Assays include detection of TGFp3 following an appropriate separation (e.g. Western blotting), Where a separation technique is used prior to immunodetection and said technique separates components on the basis of their size (such as Western Blotting), free TGFp3 can be distinguished from TGFp3-LAP3 on the basis of the size of the complex that is detected. In this case, in the detection step any means of detecting TGFp3 can be used, since it can be determined whether the molecule that is detected is associated with LAP3 or not on the basis of the size of the complex that is detected. In all methods, free TGFp3 can be detected by detecting the presence of TGFp3 molecules in which regions of the TGFp3 molecules are exposed that are not exposed when TGFp3 is complexed with LAP3. This can be achieved for example by using binding partners, e.g.

antibodies which bind to this portion of TGFp3 (i.e. the portion of TGFp3 which interacts with LAP3 and hence is not available for binding when TGFp3 is present in the TGFp3-LAP3 complex). The portion of TGFp3 which interacts with LAP3 and hence is not available for antibody binding when TGFp3 is present in the TGFp3-LAP3 complex has been identified as the N-terminal region (Walton et al. (2010))

Commercially available antibodies exist which bind to the portion of TGFp3 which interacts with LAP3 (AF243-NA (R&D systems)). This antibody or another antibody which shares the binding properties of this antibody may be used.

Alternatively, antibodies to the portion of TGFp3 which interacts with LAP3 and hence is not available for antibody binding when TGFp3 is present in the TGFp3-LAP3 complex can be generated using this portion of the molecule as an immunogen. This can be achieved using standard techniques known in the art.

Any molecule or antibody that binds to the portion of TGFp3 which interacts with LAP3 and hence is not available for antibody binding when TGFp3 is present in the TGFp3-LAP3 complex can be used in an assay to detect free TGFp3. Preferably the molecule or antibody specifically binds to TGFp3 (e.g. free TGFp3). The antibody preferably binds specifically to TGFp3 (e.g. free TGFp3). By "binding specifically" is meant that the antibody is capable of binding to TGFp3 (e.g. free TGFp3) protein in a manner which distinguishes it from the binding to non-target molecules. Thus, the antibody either does not bind to non-target molecules or exhibits negligible or substantially reduced (as compared to TGFp3 (e.g. free TGFp3)) e.g. background, binding to non-target molecules. Thus the antibody specifically recognises TGFp3 (e.g. free TGFp3), in particular specifically recognises or binds to the N-terminus (the region of TGFp3 that interacts with LAP3). The antibody does not therefore bind or exhibits negligible binding to other proteins, and in preferred embodiments the antibody does not bind to or exhibits negligible binding to the TGFp3-LAP3 complex.

A method of determining the presence or amount TGFp3 in a sample, may thus comprise: contacting said sample with one or more binding partner, preferably carried on a solid phase, which recognises TGFp3, and detecting binding of TGFp3 to said binding partner. The method may optionally further include comparing said binding to control and/or reference samples, whereby to obtain a determination of the presence or amount of TGFp3. In the above method, control or reference samples may be appropriate negative or positive controls, e.g. blanks or spiked samples.

The binding of TGFp3 to said binding partner is then detected. The detection step, in terms of reading the signal, conveniently takes place in solution. However, an insoluble product or signal may be generated which is not read in solution. A readable signal may be generated for example depending on fluorescence, chemiluminescence, colorimetry or an enzyme reaction to produce the detectable signal.

Conveniently, an immunoassay may be used as the means of detection, and preferably an enzyme-linked immunosorbent assay (ELISA). However, test procedures other than ELISA are contemplated. Immunoassay, and particularly ELISA, techniques are well known in the art and described in the literature (see for example ELISA and other solid phase Immunoassays, Theoretical and Practical Aspects; 1988, ed. D.M. Kemeny & S.J. Challacombe, John Wiley & Sons).

Following the contacting of the sample, an enzyme-antibody conjugate may be added, for example in the ELISA detection method, which binds to the binding partner bound to TGFp3 on the solid phase. An enzyme substrate is then added in order to develop the detectable signal. In the present invention, a soluble substrate is conveniently used, yielding a signal detectable in solution. This is advantageous since it facilitates and simplifies the handling and processing of a large number of samples, and permits estimation of antibody production, although as mentioned above, absolute quantitation is not necessary, and if desired a qualitative or semiquantitative result may be obtained. For convenience the substrate may be selected to yield a spectrophotometncally detectable signal, which may simply be read by reading absorbance, e.g. using a standard ELISA plate reader. Indeed, standard ELISA reagents may be used, which has the advantage of rendering the assay of the invention compatible with existing methods and techniques routinely employed in clinical laboratories. However, other detection/signal generating systems may be used, yielding signals detectable by fluorescence,

chemiluminescence etc.

By antibodies it is meant monoclonal or polyclonal antibodies, and the term antibody extends also to antigen-binding fragments (e.g. F(ab)2, Fab and Fv fragments i.e. fragments of the "variable" region of the antibody, which comprises the antigen binding site). TGFp3 is not in general secreted or present as free TGFp3 extracellularly and as such the detection of any free TGFp3 in said sample is considered to be indicative of resistance to chemotherapy in which DNA damaging agents are used. The assay may simply determine whether TGFp3 is present or absent in the sample, or it may quantitate the amount of TGFp3 present in the sample. Free TGFp3 may exert its biological effects at extremely low

concentrations and as such detection of TGFp3 at concentrations of at least 0.001 , 0.01 , 0.1 , 0.5, 1 , 2, 5, 10, 15, 20, 50, 100 pg/ml or 1 , 2, 5, 10, 15, 20, 50, 100 ng/ml in said sample is considered to be indicative of resistance to chemotherapy in which DNA damaging agents are used.

The assay may involve the comparison of the amount of TGFp3 detected in the sample with appropriate controls, i.e. the amounts of TGFp3 detected in samples from subjects known to be resistant to chemotherapy in which DNA damaging agents are used or the amounts of TGFp3 detected in samples from subjects known not to be resistant to chemotherapy in which a DNA damaging agent or DNA damaging agents are used. A subject who is resistant to

chemotherapy in which a DNA damaging agent or DNA damaging agents are used may have at least 0.5, 0.6, 0.7, 0.8, 0.85, 0.9, 0.95, 0.99 x the amount of TGFp3 present in a sample compared to the amount that is present in a control subject known to be resistant to

chemotherapy in which a DNA damaging agent or DNA damaging agents are used.

A subject who is resistant to chemotherapy in which a DNA damaging agent or DNA damaging agents are used may have at least 1.1 , 1.25, 1.5, 2, 5, 10, 50, 100, 150, 200, 250, 500, 1000x the amount of TGFp3 present in a sample compared to the amount that is present in a subject known not to be resistant to chemotherapy in which DNA damaging agents are used. The presence of free TGFp in said sample indicates that the subject is resistant to chemotherapy in which a DNA damaging agent or DNA damaging agents are used.

Thus in a preferred embodiment the invention provides a method of determining whether a subject is resistant to chemotherapy with a DNA damaging agent or DNA damaging agents, said method comprising determining whether free TGFp3 is present in a sample from said subject. Preferably said method comprises determining the amount of free TGFp3 that is present in a sample from said subject. Even more preferably the method comprises the step of comparing the amount of free TGFp3 present in a sample from the subject to known amounts of free TGFp3 present in a sample from a subject known to be resistant to chemotherapy in which a DNA damaging agent or DNA damaging agents are used and/or a subject known not to be resistant to chemotherapy in which a DNA damaging agent or DNA damaging agents are used. The presence of free TGFp3 in said sample indicates that the subject is resistant to chemotherapy in which a DNA damaging agent or DNA damaging agents are used.

The invention uses HRS proficient cells. The HRS phenotype is discussed above and is characterised by a higher than predicted level of cell killing following exposure to appropriate types of radiation at certain low doses and the existence of a threshold dose of this radiation above which a decrease in the mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) is observed.

A cell is defined as being a "HRS proficient cell" if it responds in at least one of these specific ways to appropriate doses of radiation of the appropriate type. It can be determined whether a cell is a HRS proficient cell by carrying out assays to determine whether it responds in at least one of these specific ways to appropriate doses of radiation of the appropriate type. Assays to determine the presence of the HRS phenotype in a cell are discussed elsewhere in detail. As noted elsewhere herein the HRS phenotype can be modified.

Although the above descriptions refer to a "cell", it will be understood that multiple cells are in fact required to perform the above analyses and the assays of the invention. The term cell should thus be considered to include reference to a cell population. The cell(s) used in the assays will preferably be a cell population, especially preferably a clonally derived population, the individual cells of which are genetically identical. The cell population may alternatively be defined as being cells of the same cell line and may be referred to as a cell culture or cells of a certain or the same cell type.

Examples of HRS proficient cells are T98G, T47D, SNB19, LNCaP, DBTRG, RKO, MeWo, A172, U87MG, U138, Du145, A549, UMUC3, HT29, U1 , Be11 , A7, U1 18, MSU1 , L132, HGL121 , RT112, PC3, T98G (Joiner, M. C. et al 2001). Such cells are commercially available. In fact 80% of all known cell lines are HRS-proficient. In principle any of them could be used.

Examples of HRS deficient cells, i.e. cells which do not show the HRS phenotype when exposed to appropriate doses and types of radiation are HX142, SW48, U373, SiHa, NHIK3025, MCF7.

As set out herein, HRS proficient cells are cells that are capable of expressing the HRS phenotype, i.e. cells that behave in one or more of the above described ways when exposed to the appropriate doses and types of radiation. It should however be noted that the HRS phenotype can be reversed or reduced permanently or transiently (e.g. by exposing the cells to an appropriate dose of radiation), and this phenomenon is exploited in the invention.

In all cases reduction of the HRS phenotype in HRS proficient cells in response to contact with the sample as defined elsewhere herein indicates that the subject from whom the sample was obtained is resistant to chemotherapy with DNA damaging chemotherapeutic agents.

Reduction of the HRS phenotype in HRS proficient cells causes a decrease, preferably a statistically significant decrease in the severity or the degree of the HRS phenotype, e.g. as measured by the amount of cell killing or cell survival at an appropriate dose of radiation of an appropriate type. In other words, reduction of the HRS phenotype leads to a decrease, preferably a statistically significant decrease in the number of cells that are killed or an increase, preferably a statistically significant increase in the number of cells surviving at that radiation dose, e.g. as compared to cells without said reduction. This is referred to as an increase in cell survival or a decrease in cell killing. This will cause the cells to act in a way which is more like that predicted in the LQ-model in survival experiments, and will cause the cells to exhibit a mitotic ratio profile at different doses of radiation which is characteristic of cells without HRS.

Reduction of the HRS phenotype in HRS proficient cells also causes a decrease, preferably a statistically significant decrease in the mitotic ratio and/or mitotic ratio (irradiated/unirradiated) in the cells in response to the exposure to radiation at a certain dose of radiation of a type suitable to induce HRS. In other words, reduction of the HRS phenotype leads to a decrease, preferably a statistically significant decrease in the number of cells that are undergoing mitosis at that radiation dose, e.g. as compared to cells without said reduction.

Reduction of the HRS phenotype in HRS proficient cells may also mean that the threshold value of radiation above which a decrease in the mitotic ratio and/or mitotic ratio

(irradiated/unirradiated) is no longer observed. This will cause the cells to exhibit a mitotic ratio profile at different doses of radiation which is characteristic of cells without HRS. The absence of a threshold value of radiation above which a decrease in the mitotic ratio and/or mitotic ratio (irradiated/unirradiated) in HRS proficient cells which have contacted with a sample from the subject of interest may thus be indicative that said subject is resistant to chemotherapy.

In all cases, a positive control for the complete reduction (reversal) of the HRS phenotype is exposure of the HRS proficient cells being used in the assay, to conditions that are known to completely reduce or reverse the HRS phenotype (e.g. radiation with a dose rate of 0.3 Gy/h (LDR, 0.3 Gy total), or conditioned medium from cells exposed to that LDR irradiation. Alternative doses to achieve the same effect include 0.05-0.6, e.g. 0.1-0.5, 0.2-0.4 Gy/h, at the same low dose rate (0.05-0.6, e.g. 0.1-0.5, 0.2-0.4 Gy total).

As regards the cell survival phenotype, changes that render the cells in question less sensitive to the relevant doses of radiation are described as reducing the HRS phenotype in HRS proficient cells.

A reduction in the HRS phenotype in HRS proficient cells is observed if the levels of cell killing in HRS proficient cells are decreased (preferably statistically significantly), or if the levels of cell survival in HRS proficient cells are increased (preferably statistically significantly). Changes are described as leading to a cell population having a reduced HRS phenotype if the cells are more resistant to killing by radiation of the relevant dose and type i.e. if fewer cells are killed or more cells survive (e.g. if the surviving fraction is increased) following exposure to radiation of the relevant dose and type.

Comparison with the appropriate controls is required to determine whether certain conditions cause such a reduction as defined above. This can be achieved by comparing the levels of cell survival/cell killing in HRS proficient cells which have been contacted with the sample from the subject to the levels of cell survival/cell killing in HRS proficient cells which have not been contacted with the sample from the subject. A reduction is observed if the level of cell survival at a relevant radiation dose of the appropriate type is increased (preferably statistically significantly) or the level of cell killing at a relevant radiation dose of the appropriate type is decreased (preferably statistically significantly) as a result of contact with the sample from the subject.

Preferably, a decrease in the level of cell killing as a result of contacting HRS proficient cells with a sample from a subject results in a level of cell killing being less than 90, 80, 75, 70, 65, 60, 55, 50, 45, 40, 35, 30, 25, or 20 % of that which is observed when the cells are not contacted with the sample from the subject.

Alternatively or additionally, a decrease in the level of cell killing as a result of contacting HRS proficient cells with a sample from a subject results in a level of cell killing being at least 105, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200, 250, 300 % of that which is observed when the cells are exposed to conditions which are known to reverse the HRS phenotype as set out above. This can alternatively or additionally be achieved by comparing the levels of cell survival/cell killing in HRS proficient cells which have been contacted with the sample from the subject to the levels of cell survival/cell killing which is predicted for those cells by the linear quadratic model. A reduction in HRS phenotype is observed if the cells behave more in accordance with that which is predicted by the linear quadratic model than cells than in the absence of contact with the sample from the subject. This can be determined in isolation, or by comparing the levels of cell survival/cell killing to that of the same cell type which has not been contacted with the sample from the subject. Preferably a decrease in the level of cell killing as a result of contacting HRS proficient cells with a sample from a subject results in a level of cell killing being at least 105, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200, 250, 300 % of that which is predicted by the LQ model. Alternatively stated an increase in the level of cell survival as a result of contacting HRS proficient cells with a sample from a subject results in a level of cell survival being at least 90, 80, 75, 70, 65, 60, 55, 50, 45, 40, 35, 30, 25, 20 % of that which is predicted by the LQ model.

Preferably, an increase in the level of cell survival as a result of contacting HRS proficient cells with a sample from a subject results in a level of cell survival being at least 105, 1 10, 120, 130, 140, 150, 160, 170, 180, 190, 200, 250, 300 % of that which is observed when the cells are not contacted with the sample from the subject.

Alternatively or additionally an increase in the level of cell survival as a result of contacting HRS proficient cells with a sample from a subject results in a level of cell survival being at least 90, 80, 75, 70, 65, 60, 55, 50, 45, 40, 35, 30, 25, 20 % of that which is observed when the cells are exposed to conditions which are known to reverse the HRS phenotype, as set out above.

Complete reduction, i.e. where the HRS phenotype is reduced so that it cannot be detected by the above described methods (e.g. where cell killing/cell survival is substantially identical to that which is predicted by the linear quadratic model) is described as reversal or removal of the HRS phenotype, or abolishment of the HRS phenotype.

In all cases set out above, it is preferred that the same cell types are used in the comparisons. In other words the behaviour of a population of HRS proficient cells is assessed in the presence and absence of the sample from the subject, and/or the effect of the sample is compared to the effect of conditions known to reverse the HRS phenotype as set out above. As regards the effect on mitotic ratio, changes that decrease the mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) at a given dose of radiation of the appropriate type are described as reducing the HRS phenotype in HRS proficient cells.

A reduction in the HRS phenotype in HRS proficient cells is observed if the mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) in HRS proficient cells is decreased (preferably statistically significantly). Cells in a cell population will have a reduced HRS phenotype if they enter mitosis less frequently in response to the relevant dose of radiation of the appropriate type.

Comparison with the appropriate controls is required to determine whether certain conditions cause such a reduction as defined above. This can be achieved by determining and comparing the mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) in HRS proficient cells which have been contacted with the sample from the subject to the mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) in HRS proficient cells which have not been contacted with the sample from the subject, at a relevant dose of radiation of the appropriate type. A reduction in HRS phenotype is observed if the mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) in HRS proficient cells which have been contacted with the sample from the subject is decreased (preferably statistically significantly) compared to the mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) in HRS proficient cells which have not been contacted with the sample from the subject, when said cells are exposed to a relevant dose of radiation of the appropriate type.

Preferably, a decrease in the mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) as a result of contacting HRS proficient cells with a sample from a subject results in a mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) which is less than 90, 80, 75, 70, 65, 60, 55, 50, 45, 40, 35, 30, 25 or 20 % of that which is observed when the cells are not contacted with the sample from the subject.

Alternatively or additionally a decrease in the mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) as a result of contacting HRS proficient cells with a sample from a subject results in a mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) which is at least 105, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200, 250, 300 % of that which is observed when the cells are exposed to conditions which are known to reverse the HRS phenotype, as set out above. This can alternatively or additionally be achieved by observing the overall profile of the mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells) in HRS proficient cells which have been contacted with the sample from the subject at different doses of radiation of an appropriate type to induce HRS. A reduction in HRS phenotype is observed if the cells behave more in accordance the profile which is characteristic of a non HRS cell. This can be determined in isolation or by comparing the profile to that the same cell type which has not been contacted with the sample from the subject. A profile which is characteristic of a non HRS cell does not contain a threshold dose as defined above.

Complete reduction, i.e. where the HRS phenotype is reduced so that it cannot be detected by the above described methods (e.g. the mitotic ratio and/or mitotic ratio (irradiated cells/non irradiated cells shows no threshold value when assessed at different doses of radiation of the appropriate type) is described as reversal or removal of the HRS phenotype or abolishment of the HRS phenotype.

In all cases set out above, it is preferred that the same cell types are used in the comparisons. In other words the behaviour of a population of HRS proficient cells is assessed in the presence and absence of the sample from the subject, and/or the effect of the sample is compared to the effect of conditions known to reverse the HRS phenotype in said cell type as set out above.

Any reduction or reversal in the HRS phenotype can be permanent or transient. The inventors have shown that the HRS phenotype can be reversed or reduced in vitro by exposing HRS- proficient cell cultures to a high dose-rate (HDR) pre-exposure of 0.2-0.3 Gy. This transiently abolishes the HRS-response to subsequent challenge irradiation ((Edin et al., 2007; Joiner et al., 1996; Marples and Joiner, 1995; Short et al., 2001 ; Wouters and Skarsgard, 1997). By reducing the dose-rate to 0.3 Gy/h (LDR) the effect on HRS in T-47D breast cancer cells of a pre-exposure of 0.3 Gy becomes permanent (Edin et al., 2007).

Transient removal of HRS was achieved in cells which received conditioned medium from cells exposed to LDR (Edin et al., 2009b). LDR irradiation (0.3 Gy) induced the same factor in cell conditioned medium without cells present during irradiation (Edin et al., 2009a), as long as the cells were cultured in medium with foetal bovine serum (Edin et al., 2009a).

It will be appreciated that there are several ways in which it is possible to determine whether a sample is capable of reducing the low dose hyper-radiosensitivity (HRS) phenotype in HRS proficient cells. HRS proficient cells may be contacted with or exposed to the sample. Following an appropriate period of time it is determined whether the low dose hyper-radiosensitivity (HRS) phenotype has been reduced in response to the contact with or exposure to the sample. This can be achieved by subjecting the cells to a challenge irradiation dose (which is radiation of a type and at a dose at which a HRS phenotype would be expected, as defined above) and, after a further appropriate period of time, assessing the HRS phenotype by any means. Determining whether the HRS phenotype has been reduced as a result of exposure to said sample can be carried out.

The HRS phenotype observed after contact with or exposure to the sample can be compared to the HRS phenotype observed in cells which have not been contacted with or exposed to the sample. If such a comparison is required, the method of the invention may include the steps of subjecting HRS proficient cells which have not been exposed to the sample to a challenge irradiation dose, as defined above and, after an appropriate period of time, assessing said cells with respect to the HRS phenotype as defined herein e.g. assessing cell survival or mitotic ratio.

The HRS phenotype observed after contact with or exposure to the sample can be compared to the HRS phenotype observed in cells which have been treated with conditions known to reverse the HRS phenotype as set out above. If such a comparison is required, the method of the invention may include the steps of subjecting HRS proficient cells which have not been exposed to the sample to conditions known to reverse the HRS phenotype as set out above, and subjecting said cells to a challenge irradiation dose, as defined above and, after an appropriate period of time, assessing said cells with respect to the HRS phenotype as defined herein e.g. assessing cell survival or mitotic ratio.

The methods of the invention thus include the step of determining the HRS phenotype as defined herein in cells after contact with or exposure to the sample and optionally determining the HRS phenotype as defined herein in cells which have not been exposed to the sample, and/or determining the HRS phenotype as defined herein in cells which have not been exposed to the sample but which have been treated with conditions known to reverse the HRS phenotype. Methods may also include the step of comparing the HRS phenotypes obtained to determine whether any reduction in HRS phenotype has occurred as a result of the contact with said sample and hence whether the subject from whom the sample was obtained is resistant to chemotherapy.

Contacted as used herein refers to providing suitable contact between the sample and the cell (e.g. cell population) so as to allow the active component(s) in said sample (including but not limited to TGFp3) to interact with the cells and to have an effect on the properties of the cell. Thus, the sample may simply be brought into contact with the cell(s) for example by adding it to the cells or to a medium containing the cell(s) e.g. a culture medium or culture of the cells.

Conveniently, the cells may be grown (or cultured or maintained) in a liquid medium to which the inhibitor/test sample is introduced or added. Alternatively, the cells may be contained in or on a solid medium (e.g. a culture dish or vessel or plate) to which the inhibitor/test sample is introduced or added

The period of contact is any time which allows the active component(s) in said sample (including but not limited to TGFp3) to interact with the cells and to have an effect on the properties of the cell. This may be 1-96 hours e.g. 2-72, 4-60, 6-48, 12-36, 18-30, 20-24 hours or at least 6, 12, 18, 24, 36, 48, 60, 72 or 96 hours.

During the period of contact with the test sample the cells are preferably maintained under normal growth conditions. Such conditions are well known in the art and can be determined for any cell type.

After the challenge irradiation dose (the dose and type of radiation to induce the HRS

phenotype) the cells are preferably maintained under normal growth conditions. Such conditions are well known in the art and can be determined for any cell type. The period of time that is allowed to elapse after the challenge irradiation dose and before carrying out assays to determine whether the HRS phenotype has been reduced will depend on the nature of the assay that is being used. It can be at least or up to 20, 30, 40, 50, 60, 70, 80, 90, 100, 1 10, 120 minutes or at least or up to 3, 4, 5, 6, 7, 8, 9, 10, 11 , 12, 24, 36, 48, 60, 72 hours or at least or up to 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 , 12, 13, 15, 16, 17, 18, 19, 20 days e.g. 1-18, 2-16, 10-14 days.

Methods for assessing the HRS phenotype can be assessed by determining cell survival or cell killing in response to an appropriate challenge irradiation dose, which may be any appropriate dose, but is preferably a dose of low LET radiation at which the HRS phenotype is observed, as set out above. Multiple challenge doses (e.g. at least 2, 3, 4, or 5 doses) may be used.

The sample may be any body fluid or tissue sample but is preferably a body fluid sample.

The active factor which mediates the resistance to chemotherapy is believed to be present in its active form extracellularly. As such any sample which contains extracellular material can be used. Cells or tissue may be present in the sample but it is the presence of extracellular material which is required. Examples of body fluid samples include blood, aqueous humour and vitreous humour, breast milk, cerebrospinal fluid, endolymph and perilymph, mucus (including nasal drainage and phlegm), peritoneal fluid, pleural fluid, saliva, sebum (skin oil) semen, sweat, tears, vaginal secretion, vomit and urine.

Surgical intervention techniques have been developed which allow such material to be taken from a subject and such procedures may be used where appropriate to obtain samples for use in the method of the invention. In some embodiments therefore the invention further comprises the step of obtaining a sample from a subject.

Blood samples including preparations obtained from blood samples are preferred. Blood serum or plasma samples are particularly preferred. Standard techniques are known for preparing blood serum and plasma samples from whole blood samples. In some embodiments therefore the invention further comprises the step of purifying blood serum or blood plasma from said whole blood sample.

In all cases described above, the subject as referred to herein is preferably a mammal, reptile, bird, insect or fish (e.g. salmon or cod). Preferably the subject is a mammal, particularly a primate, domestic animal, livestock or laboratory animal. Thus preferred subjects include mice, rats, rabbits, guinea pigs, cats, dogs, monkeys, pigs, cows, goats, sheep and horses.

Especially preferably the subject or the patient is a human. It will be understood from the description that the subject or patient preferably has cancer and that preferably said

chemotherapy is to treat cancer.

Chemotherapy with DNA damaging agents means chemotherapy using an agent which causes direct or indirect damage to DNA. This includes DNA alkylating agents such as

Cyclophosphamide, Isophosphamide, Chlorambucil, Melphalan, Mitomycin, Busulfan,

Estramustin, Lomustine (CCNU), Streptozocine, Dacarbazin (DTIC), Temozolomide,

Altretamine and DNA intercalating agents such as Cisplatin, Carboplatin, Oxaliplatin. Such agents and their mechanisms of action are well known in the art.

By resistant to chemotherapy with DNA damaging agents it is meant that when a subject is treated with one or more of these agents (for example to treat cancer), the treatment does not achieve cell killing (i.e. effective treatment) at the level which would be expected by this treatment. This is assessed clinically using conventional means. As discussed above, it is believed that the mechanism underlying resistance to or reduction in the HRS phenotype in cells e.g. HRS proficient cells in vitro may also give rise to resistance to chemotherapy with DNA damaging agents in vitro (see Example 10), and also in vivo. As such, detecting the above parameters related to the inventor's proposed mechanism allows a determination of whether a patient is resistant to chemotherapy with a DNA damaging agent or DNA damaging-agents.

The assays described herein thus preferably detect patients who are resistant to chemotherapy using a DNA damaging agent or DNA damaging agents, wherein the resistance is conferred by the above described pathway (which may for example be initiated by exposure to radiation). It will therefore be appreciated that not every patient who is resistant to chemotherapy with a DNA damaging agent or DNA damaging agents will have acquired resistance through the above- described pathway, and hence not every patient who is resistant to chemotherapy with DNA damaging agents will be identified using the above assays. On the other hand the subjects who are positively identified in the present assay will have resistance to chemotherapy. In particular this will be resistance to chemotherapy with a DNA damaging agent or DNA damaging agents which has been acquired through the above pathway. Thus in a patient who is known to be resistance to chemotherapy, the above described assays can be used to determine whether the resistance to chemotherapy was acquired by the above described pathway. The assays can thus be carried out for example on a patient who has cancer, or a patient who is known to have resistance to chemotherapy with a DNA damaging agent. All assays described herein can be used to this end.

The methods and assays as set out herein thus determine whether a patient or subject (e.g. with cancer) is resistant to chemotherapy using a DNA damaging agent or DNA damaging agents. The methods can also be used to determine whether a patient or subject who is known to be resistant to chemotherapy using a DNA damaging agent or DNA damaging agents has acquired said resistance as a result of the pathway as set out herein, i.e. as a result of the activation of the pathway that causes a reduction in the HRS phenotype in HRS proficient cells.

As such, a patient who is positively identified according to the methods described herein as being resistant to chemotherapy with a DNA damaging agent or DNA damaging agents has preferably acquired said resistance as a result of exposure to radiation, e.g. of an appropriate type and dose to cause activation of the pathway that causes a reduction in the HRS phenotype in HRS proficient cells as discussed elsewhere herein. As discussed above, the mechanism by which radiation induces the pathway involves induction of ROS and ROS is required for NO production by iNOS. Alternatively stated, the patient who is positively identified according to the methods described herein has preferably acquired resistance to chemotherapy with a DNA damaging agent or DNA damaging agents as a result of the activation of the pathway that causes a reduction in the HRS phenotype in HRS proficient cells. This pathway is defined in detail elsewhere herein, and for example includes the induction of ROS, NO production by iNOS and the consequent release of free extracellular TGFp3. These steps leading to free extracellular TGFp3 are set out in detail above. TGFp3 is then believed to then act via its receptor.

Cells that are used in the above assays of the invention are maintained under standard growth conditions. The precise conditions under which the cells are maintained or cultured will depend on the nature of the cells, but in general standard growth conditions for the particular cell types that are used or conditions that allow cells to maintain viability should be used. In most cases standard conditions to maintain viability will be known. For example mammalian cells are generally maintained in liquid growth medium at 35-39°C, preferably 36-37°C for mammalian cells. The components of the liquid growth medium may be determined by the person skilled in the art depending on the particular cell type that is used. If standard conditions are not known, appropriate conditions can readily be determined by the person skilled in the art. The cells that are used are preferably maintained in exponential growth. HRS proficient cells that are used are preferably maintained in asynchronous division.

All of the assays referred to above are preferably carried out in vitro.

The invention will now be described with reference to the following non-limiting Examples in which:

Figure 1 shows HRS in T98G cells. (A) Unprimed T98G cells (■) or T98G cells given a LDR (0.3 Gy/h) priming dose of 0.3 Gy (T98G-P) (A ) were irradiated with a single HDR dose of ⁶⁰ Co γ-radiation. The curves represent model-fits to the data from unprimed T98G cells by the IR- model (solid line) and the LQ-model (dashed line), respectively. The parameters from the fit to data from unprimed T98G cells by the IR-model are presented in Table I. (B) The ratio

(irradiated/unirradiated) of mitotic cells (staining positive for phosphorylated histone H3) as a function of radiation dose given 1 h before cell harvest. (■): unprimed T98G cells, (A): LDR primed T98G (T98G-P). The bars represent standard errors of the mean (SEM) for 3 individual experiments. (C) Surviving fractions after 0.2 or 0.3 Gy HDR challenge doses to T98G cells, 0.06 Gy/h for 1 h (■) or 0.19 Gy/h for 15 minutes (A ). Data from T98G controls(D) and T98G cell primed with 0.22 Gy/h for 1 h (T98G-P) (o) are shown as references. p<0.01 for data points (■) compared to T98G controls.

Figure 2 shows the effect of pretreatments using interleukins in serum-free medium.

Surviving fractions after 0.2 or 0.3 Gy challenge doses (A) T98G cells, (B) T47D cells. Pretreaments: None (controls) (black), LDR irradiation of cells with serum-free medium with 10 ng/ml IL-4 (red), serum-free medium with 10 ng/ml IL-13 and no irradiation (blue), LDR irradition of cells with normal medium (purple), LDR irradiation of cells with serum-free medium with 10 ng/ml IL-13 (green), medium transfer of LDR irradiated cell conditioned normal medium, 48h before challenge irradiation (dark blue), medium transfer of LDR irradiated cell conditioned serum-free medium with 10 ng/ml IL-13, 48h before challenge irradiation (violet), medium transfer of LDR irradiated cell conditioned serum-free medium with 10 ng/ml IL-13, 2 weeks before challenge irradiation (brown). The bars represent standard errors of the mean (SEM) for 3 individual experiments.

Figure 3: shows (A)The combination of a NO-donor and HDR irradiation could imitate LDR irradiation: Surviving fractions of T98G cells after 0.2 or 0.3 Gy challenge doses. Controls (black), cells were exposed to 0.1 mM DEANO for 24h before plated in fresh medium the day before challenge irradiation (red), 0.1 mM DEANO (for 24h) and 0.3 Gy HDR irradiation, 48h before challenge irradiation (blue), 0.1 mM DEANO (for 24h) and 0.3 Gy HDR irradiation, 2 weeks before challenge irradiation (purple), medium transfer of 0.3 Gy HDRly irradiated cell conditioned medium with 0.1 mM DEANO (medium transfer) (green). The bars represent standard errors of the mean (SEM) for 3 individual experiments.

(B) The combination of a NO-donor and reoxygenation after hypoxia could imitate LDR irradiation: Surviving fractions after 0.2 or 0.3 Gy challenge doses. T98G cells grown with 20% 0 ₂ (controls) (black), T98G cells grown for 6 weeks with 4% 0 ₂ in the gas phase (red), T98G cells grown for 6 weeks with 4% 0 ₂ in the gas phase followed by 2 weeks in 20% 0 ₂ (blue), T98G cells grown for 6 weeks with 4% 0 ₂ in the gas phase followed by 2 weeks in 20% 0 ₂ with 0.1 mM DEANO added 2 h prior to reoxygenation (purple). The bars represent standard errors of the mean (SEM) for 3 individual experiments.

Figure 4 shows the effect of iNOS inhibitor 1400W: Surviving fractions after 0.2 or 0.3 Gy challenge doses. (A) and (C) T98G-P cells, (B) and (D) T47D-P cells. T98G-P or T47D-P cells (LDR primed 6 months previously) were exposed to medium containing 10 μΜ 1400W on 3 consecutive days. This treatment restored HRS when tested the next day (blue/β) or after 3 weeks with 2 weekly reseedings (purple/*). LDR irradiated cell conditioned medium (LCCM) removes HRS in recipient cells (green/Δ). When 10 μΜ 1400W was added to cell conditioned medium before LDR irradiation of the medium, the medium did not remove HRS in recipient cells (dark blue/A). In (C) and (D) p<0.01 for data points (■), (·) and ( A) compared to LDR primed cells. The curves represent model-fits to the data from unprimed cells by the IR-model (solid lines) and the LQ-model (dashed lines), respectively. (E) and (F): The effect of iNOS inhibitor 1400W in HRS-negative cell line NHIK 3025. (E) Doses up to 8 Gy. (F) Doses up to 3 Gy. The curves represent model-fits to the data from untreated NHIK 3025 cells by the LQ- model (solid lines) and to NHIK 3025 treated with 1400W by the IR-model (dashed lines), respectively. p<0.02 for data points of NHIK 3025 cells with and without 1400W. The bars represent standard errors of the mean (SEM) for 3-5 individual experiments.

Figure 5: shows (A)The effect of TGFp neutralizers: Surviving fractions after 0.2 or 0.3 Gy challenge doses. T98G controls (black), cells receiving medium transfer of LDR irradiated cell conditioned medium (LCCM) (red), medium transfer of LCCM with 100 μg/ml pan-specific TGF- β neutralizer (red), 5 μg/ml TGF- β1 neutralizer (purple), 1 μg/ml TGF- β2 neutralizer (green), 2 μg/ml TGFP3 neutralizer μg/ml (dark blue). In all experiments the recipient cells were exposed to LCCM for 24 before being plated in fresh medium the day before challenge irradiation. The bars represent standard errors of the mean (SEM) for 3 individual experiments.

(B) & (C)The effect of TGFP3 on HRS: Surviving fractions after 0.2 or 0.3 Gy challenge doses. The cells were exposed to medium (not cell conditioned) containing 1 (purple), 0.01 (green) or 0.001 (dark blue) ng/ml recombinant TGFP3 for 24h before being plated in fresh medium 19h prior to challenge irradiation. T98G cells were cultured for 2 weeks after 24 h exposure to 0.01 ng/ml TGF β3 (violet). (B) T98G cells, (C) T47D cells. The bars represent standard errors of the mean (SEM) for 3 individual experiments

(D) The effect of TGFp neutralisers : T98G cells receiving medium transfer of LDR irradiated cell conditioned medium (LCCM) with 100 μg/ml pan-specific TGF- β neutralizer (■), 5 μg/ml TGF- β1 neutralizer (·), 1 μg/vΓ^\ TGF- β2 neutralizer (A), 2 g/ml TGFP3 neutralizer (*). In all experiments the recipient cells were exposed to LCCM for 24 h before being plated in fresh medium the day before HDR challenge irradiation. ** p<0.01 for data points 0.2 and 0.3 Gy (■) and (*) compared to T98G cells exposed to transfer of LCCM.

(E) and (F) The effect of TGFp on HRS The cells were exposed to medium (not cell conditioned) containing 1 (■), 0.01 (·) or 0.001 (A) ng/ml recombinant TGFP3 for 24h before being plated in fresh medium 19h prior to HDR challenge irradiation. T98G cells were cultured for 2 weeks after 24 h exposure to 0.01 ng/ml TGF β3 (*). (C) T98G cells, (D) T47D cells. p<0.01 for data points (■), (·) and (A) compared to controls. In each of 5(D) to (F), data from controls(ci) , LDR primed cells (o), and cells receiving medium transfer of LDR irradiated cell conditioned medium containing serum, 20-48h before HDR challenge irradiation (Δ) are shown as references. These curves represent model-fits to the data from unprimed T98G cells by the IR-model (solid lines) and the LQ-model (dashed lines), respectively. The bars represent standard errors of the mean (SEM) for 3 individual experiments.

Figure 6 shows the density of TGFP3 in the cytoplasm: Post embedding immunogold electron microscopy was used to calculate the density of ΤΘΡβ3 in the cytoplasm. (A) Low magnification electron micrograph of T98G-P cells. Red lines show the nuclear membranes. Scale bar 2 mm. (B) High magnification of green square in (A). Scale bar 0.2 mm. Αηίί-ΤΘΡβ3 labeling with gold particles are marked with red arrows. (C) The density of ΤΘΡβ3 in the cytoplasm of T98G and LDR irradiated T98G (T98G-P) cells. (D) The density of TGF$3 in the cytoplasm of T-47D and LDR irradiated T-47D (T47D-P) cells. The bars represent standard errors of the mean (SEM) for 30 analyzed micrographs.

Figure 7 shows the hypothesis for the mechanism of permanent elimination of HRS by LDR radiation. (A) IL-13 binds to receptor IL-13a2, which leads to upregulation of iNOS and conversion of pro-furin to active furin. Furin proactivates the ΤΘΡβ3-ίΑΡ3 complex by proteolytic cleavage of the covalent binding leaving ΤΘΡβ3 non-covalently associated with LAP3. iNOS and the pro-activated ΤΘΡβ3-ίΑΡ3 complex are secreted by the cell. (B) LDR irradiation results in sustained ROS production which is required for NO production by iNOS. ROS (not necessarily sustained) also activates ΤΘΡβ3 by breaking the binding to LAP3. NO then scavenges LAP3 preventing reconstitution of the ΤΘΡβ3-ίΑΡ3 complex and resulting in free activated ΤΘΡβ3.Ιί this process takes place in cell conditioned medium, ΤΘΡβ3 will bind to a receptor in the recipients resulting in removal of HRS. (C) If the cells are exposed to LDR radiation, activated ΤΘΡβ3 will be present intracellular^. This starts a self-sustained process resulting in continued neutralization of LAP3. Intracellular ΤΘΡβ3 uncouples iNOS resulting in production of superoxide and peroxynitrite. LAP3 is scavenged by peroxynitrite resulting in permanent presence of activated ΤΘΡβ3, which is secreted.

Figure 8 shows testing of mouse serum in a cell model. Cells received medium with 4% mouse serum and 6% foetal calf serum. 24 h later cells were seeded for colony formation in fresh medium and 16-20 h after seeding, the cells were challenge irradiated. Left panel shows experiments done with T-47D cells , right panel shows experiments with T98G cells. Surviving fraction is shown for each condition used. Figure 9 shows low dose priming protects the cell against cisplatin. T-47D cells and low dose-rate (LDR) primed T-47D cells (T-47D-P) were exposed to different doses of cisplatin and the surviving fraction determined.

Figure 10

EPR measurements of superoxide levels. Superoxide measurements were done using the spin trap CMH with or without added NOS inhibitor L-NAME (LN) show that iNOS was not uncoupled in LDR primed cells . The bars represent standard errors of the mean (SEM) for 3 individual experiments.

Figure 11

Surviving fraction in response to small challenge doses of T-47D cells exposed to serum from mice sacrificed 12 (female mice) or 15 months (male mice) after irradiation with 0.3 Gy/h for 1 h. Serum from unirradiated animals at the same age and gender was used as control.

Figure 12

Surviving fraction in response to challenge doses of T-47D cells exposed to serum from mice irradiated with 0.3 Gy/h for 1 h with added TGFp3-neutralizer.

Figure 13

Surviving fraction in response to challenge doses of T-47D cells exposed to serum from mice irradiated with 0.3 Gy/h for 1 h ( A ) and from mice irradiated with 0.3 Gy/h for 1 h and

subsequently treated with 1400W (*)

Figure 14

Surviving fraction in response to small challenge doses of T-47D cells exposed to serum from mice irradiated with 0.03 Gy/h for 1 h. Serum from unirradiated animals at the same age and gender was used as control.

EXAMPLES

Experimental procedures

Cell culture

Human T-47D breast cancer cells and human glioblastoma T98G cells (purchased from ATCC, LGC Standards AB, SE-501 17 Boras, Sweden) were grown as monolayer cultures in RPMI (Roswell Park Memorial Institute) 1640 medium (JRH Biosciences, Lenexa, KS, USA),

supplemented with 10% fetal calf serum (Gibco, Paisley, UK), 2mM L-glutamine (SIGMA, St Louis, MO, USA), 200 units I ^"1 insulin (SIGMA), and 1 % penicillin/streptomycin (Gibco) at 37°C in air containing 5% C0 ₂ . T-47D cells contain mutations in p53 gene (Bartek et al., 1990; Nigro et al., 1989) and phosphoinositide-3-kinase (PIK3CA) gene (Bachman et al., 2004). T98G cells contain mutations in exon 7 of the p53 gene (Matsumoto et al., 1994). The human cervical carcinoma in situ cell line NHIK 3025 (Norbye & Oftebro 1969; Oftebro & Nordbye 1969) was grown in MEM plus 15% fetal calf serum. The cells were kept in exponential growth by reculturing of stock cultures two times a week. The cells were tested negative for the presence of mycoplasma.

In experiments with interleukins 4 or 13 (IL4-10H and IL13-22H, Creative Biomart, NY 1 1967, USA), the cells were washed 3 times with serum-free medium before serum-free medium with 10 ng/ml interleukin was added for cell conditioning and LDR priming. 10% foetal calf serum was added 1 h after end of irradiation. NO-donor Diethylamine NONOate sodium salt hydrate (DEANO) and iNOS inhibitor 1400W were purchased from SIGMA (D184, SIGMA, St Louis, MO, USA). Pan-specific TGF neutralizer, ΤΰΡβΙ neutralizer, ΤΰΡβ2 neutralizer, and ΤΘΡβ3 neutralizer were purchased from R&D (AB-100NA, MAB240, AB112NA, AF243-NA, R&D systems, Minneapolis, MN, USA)

Hypoxic culture

The cells were cultured with two weekly reseedings in an IN VIV0 ₂ 400 glove box hypoxia workstation operated to contain 4 % 0 ₂ and 5% C0 ₂ in the gas phase. The cells grown under these conditions are exposed to cycling hypoxia with oxygen levels decreasing from 4% to below 0.1 % between reseedings.

Irradiation procedures

The cells were irradiated in T25 flasks (Nunc, Roskilde, Denmark) from below with a ⁶⁰ Co source (Theratron 780-C, MDS Nordion, Ottawa, ON, Canada). The irradiation field was 40 ^χ 40 cm ² and the source-to-flask distance was 80 cm giving a HDR (HDR) of ~ 32 Gy/h. The cell flasks were placed on a hollow water-filled perspex plate, which was heated to maintain 37°C in the medium of the flasks by circulating water from a water bath (Grant Instruments, Cambridge, England).

The low dose-rate (LDR) was obtained by shielding the source by 10 cm Roos metal (Sn 25%, Pb 25%, Bi 50%, melting point 96°C, specific weight 9.85 g/cm ³ ). Because of decay, the dose- rate used in the present experiments was -0.22 Gy/h. The radiation time was still 1 h, so the total dose in all LDR radiations was -0.22 Gy. Dose and dose-rate measurements were

performed using thermoluminescence dosimetry. Assessment of Anti-phospho-histone H3 (ser28) Staining

The method of assessment of anti-phospho-histone H3 staining was adapted from those of Juan et al. (Juan et al., 1998) and Xu et al. (Xu et al., 2002) as described previously (ref Edin. In short, asynchronously growing cells were irradiated at 37°C and incubated for 60 minutes before trypsinization fixation in freezer-cold (-20°C) 70% ethanol. The samples were stored overnight at 4°C before 1 h incubation with the primary antibody (anti-phospho-histone H3 (ser 28) rabbit polyclonal IgG mitosis marker, Upstate Catalogue number 06-570, Charlottesville VA, USA). After incubation with fluoroscein (FITC)-conjugated secondary antibody (Imgenex, Catalogue number 20302, San Diego CA, USA), the cell pellet was resuspended in 500 μΙ PBS containing 35 μg propidium iodide (Sigma) and 100 μg/ml ribonuclease (RNase) (Sigma,). The samples were analyzed on an Accuri C6 flow cytometer (Accuri Cytometers, Inc., Ann Arbor, Ml USA).

Transfer of medium irradiated without cells present

The medium used for these experiments was harvested from 60-80% confluent flasks with unirradiated cells (denoted cell conditioned medium CCM). The CCM was filtered before irradiation through a 0.22 μηι filter used to sterilize solutions in order to exclude cellular debris. The medium was transferred to T-25 flasks with unirradiated cells 1 h after irradiation with 0.3 Gy at 0.3 Gy/h. After incubation with the transferred medium for 24 h, the cells were plated for colony formation (200 cells per flask) in fresh medium and the challenge doses were given at HDR (40 Gy/h) after another 20 h or 16-20h of incubation.

Cell Survival

After challenge irradiation, the 25 cm ² flasks were placed with open lids in an incubator with 5 % C0 ₂ in air of high humidity for 10-14 days. The surviving fraction was determined by counting the number of colonies per flask after fixation in absolute alcohol and staining with methylene blue (Pettersen et al., 1973). Colonies containing more than 50 cells were scored as survivors and increased mean multiplicity per colony-forming unit during the interval between seeding and irradiation was corrected for according to a formula previously published (Gillespie et al., 1975).

Statistical analysis

All experiments were repeated at least 3 times using five flasks for each dose and ten for controls. Within each experiment, the arithmetic means were calculated weighing the errors.

In figure 1 the curves represent fits to data from T-47D cells irradiated with a single HDR dose by either the induced repair-model (IR-model) (solid line) or the linear quadratic-model (LQ- model) (dashed line) using the method of least-squares and weighing the errors. For the fit by the LQ-model, only data above 1 Gy were used.

The LQ-model is described by the equation:

S = e -ad - pd ² ) (equation 1)

Where S is the surviving fraction, d the dose and a and β the parameters describing the linear and quadratic component, respectively, of the intrinsic radiosensitivity.

In the I R-model a is replaced by: J J (equation 2) where d is dose, a _r is the value of a extrapolated from the high dose LQ response (equation 2), and a _s is the actual value of a derived from the initial part of the curve (i.e. at very low doses). d _c is the dose where the change from a _s to a _r is 63% complete.

Two-tailed Student's t-test was used to compare the surviving fractions in response to challenge irradiation of pre-treated cells compared to controls or LDR irradiated cells.

Post-embedding immunogold electron microscopy.

The procedure for post embedding immunogold electron microscopy was adapted from

Bergersen and colleagues (Bergersen et al. , 2008). Small rectangular blocks from cell boluses (T98G-P, T98G, T-47D-P and T-47D cells) (typically 0.5 mmx0.5 mmx l mm) were

cryoprotected by immersion in graded concentrations of glycerol (10%, 20% and 30%) in PBS. The samples were then plunged into liquid propane cooled to -170°C by liquid nitrogen in a Universal Cryofixation System KF80 (Reichert-Jung). The cell blocks were handled using a precooled forceps. For freeze-substitution, the cell samples were immersed in a solution of anhydrous methanol and 0.5% uranyl acetate overnight at -90°C. The temperature was then raised stepwise in 4°C increments per hour from -90 to -45°C, where it was kept for the subsequent steps. The cell samples were then washed several times with anhydrous methanol to remove residual water and uranyl acetate. Samples were infiltrated stepwise with Lowicryl resin HM20, starting from a 1 :2 solution of Lowicryl: methanol (1 hour) and progressing through 1 : 1 and 2: 1 (1 hour each) solutions to pure Lowicryl (overnight). For polymerization, the samples were placed in a precooled embedding mall, and polymerization was catalyzed by exposure to ultraviolet light at a wavelength of 360 nm for 2 days at -45°C, followed by 1 day at room temperature. Ultrathin sections (90 nm) were cut on a Reichert-Jung ultra microtome by a diamond knife and mounted on nickel grids using an adhesive pen (Electron Microscopy Sciences, USA). The grids containing ultrathin sections were processed at room temperature in solutions containing Tris-buffered saline (TBS) and 0.1 % Triton X-100 (TBST), with additions as noted. The sections were first washed in TBS containing H ₂ 0 ₂ followed by TBST containing 2% human serum albumin (HSA). They were then incubated overnight with primary antibodies against TGFbeta3 (AF243-NA, R&D systems, Minneapolis, MN, USA). The concentration of the primary antibody was 0.667μg/ml. Bound antibody was visualized by incubating for 2 hours with secondary (rabbit anti goat) immunoglobulin conjugated with 10 nm diameter colloidal gold (British Biocell International, UK). The secondary antibody was diluted 1 :20 in TBST containing 2% HSA and centrifuged at 1000 rpm for 10 minutes before use to sediment aggregated gold particles. The ultrathin sections were rinsed in ultra pure water and dried before applying contrast agents uranyl acetate (1 %) and lead citrate (0.3%). Sections were observed using a Tecnai electron microscope. Pictures were taken randomly from the cytoplasm at primary magnifications of *43,000.

Quantitative analysis of electron microscopic immunogold micrographs were taken randomly from the sample. For each of the four kinds of cells (T98G-P, T98G 20% oxygen, T-47D-P and T-47D 20% oxygen cells), 30 micrographs were analyzed. For each micrograph, cytoplasm area was marked by software ImageJ with Point density plug-in. The area was calculated and gold particles situated in the region were marked by the software. The density of TGFbeta3 in the cytoplasm was calculated as the number of gold particles divided by the area.

EPR measurements of superoxide

EPR measurements were carried out in a Bruker escan EPR spectrometer (Noxygen; Elzach, Germany). For superoxide measurements Noxygen Temperature Controller was used to keep temperature at 37°C. Measurements were normalized to protein content.

Cells were grown to 70% confluency, washed and scraped in Krebs Hepes buffer containing 25 μΜ desferroxamine and 5 μΜ diethyldithiocarbamate (DETC). Cells were measured in the presence of 100 μΜ 1-hydroxy-3-methoxycarbonyl-2,2,5,5-tetramethylpyrrolidine (CMH) in a 50 μΙ precision micropipette in a synthetic quartz glass capillary holder (Noxygen) for 10 min at 37°C. Cells were incubated with 300 μΜ N-Nitro-L-Arginine Methyl Ester (L-NAME) ten minutes prior to measurement to inhibit NOS activity.

EPR measurement was carried out with settings as following: microwave power=23.89 mW, center field=3455G, modulation frequency=86 kHz and modulation amplitude of 2.93 G for detection of CMH signal. Mouse model and in vivo irradiation

DBA/2 mice (Anver & Haines 2004) were used. After sex segregation at weaning they were kept 8 a box and fed mouse pellets and water ad libitum. The animal room, the animal care and the experimental use of the animals were in accordance the Slovak Ethical rules. The animals entered experiments when 3 months old. Animals were placed in a circular pen with wedge shaped individual rooms, giving each animal an identical exposure to the radiation. The dose rates used were 0.3 and 0.03 Gy/h and the duration of irradiation one hour. A [ ⁶⁰ Co]-source (Theraton Elite 100, Best Theatronics, Canada ) was used. The received dose was measured with total cerrobend filtration in solid PMMA phantom by optimal SSD (-217 cm) for a dose rate of the 300 mGy/h with field size 35x35 cm ² . The time for application of 300m Gy was in the interval 5.96-60.3 min. The treatment parameters were corrected at each application to ensure the mice received the demanded dose rate. The applied dose was controlled by a ionization chamber (FC-65G, Wellhofer, Germany) calibrated by Standard laboratory IAEA Siebersdorf.

Eight males and eight females were irradiated and a similar number of age matched non- irradiated mice were used as controls in each experiment. The 0.3 Gy irradiated mice were bled at different time points from one day to 15 months after irradiation later, the 0.03 Gy irradiated mice were bled after one month. Furthermore eight males and eight females were irradiated and a week later given the selective iNOS inhibitor 1400W ((N-(3)aminobenzyl)benzyl)acetamine), dihydrochloride, SigmA-Aldrich, Germany), 10mg/kg intraperitoneally in 0.1 ml 0.9 per cent saline every 6th hour for 5 consecutive days and the mice bled 24 h after the last inoculation.

Reporter cell culture for mouse studies

Human breast cancer cells of the line T-47D and human glioblastoma cells of the line T98G were grown as described above and mouse serum (4%) was added to RPMI with 6% fetal calf serum and the medium was filtered (0.22 μηι) before transferred to the reporter cells. The reporter cells were exposed to the mouse serum for 24h before being plated for colony formation in fresh medium (10% calf serum). 2 μg/ml TGF$3 neutralizer (AF243-NA, R&D systems, Minneapolis, MN, USA) was added to the medium containing mouse serum before transfer to reporter cells.

Example 1 T98G glioblastoma cells

The sustained effect of LDR irradiation on HRS, previously found in T-47D cells, was also observed in another HRS-proficient cell line, T98G glioblastoma cells. Surviving fractions from unprimed and LDR primed T98G (in the following named T98G-P) cells are shown in figure 1A. The parameters from the fit by the IR-model to the data from the unprimed T98G cells are given in Table 1. The mitotic marker histone H3 phosphorylation was used to measure the mitotic ratio in T98G and T98G-P cells 1 h after doses between 0.1 and 1 Gy. Coincident with the absence of HRS (figure 1A (▲)), the T98G-P cells showed a trend of decreasing mitotic ratio with increasing radiation dose (figure 1 B ( A )) without a lower threshold dose as was seen for T98G cells (figure 1 (■)). These data correlate with our previous data for T-47D cells (Edin et al., 2007; Edin et al., 2009a; Edin et al., 2009b), and suggest that the early G ₂ - checkpoint is activated at doses below the HRS-threshold in LDR primed HRS-proficient cells.

TABLE 1

Parameters of the fit by the IR-model to the data points from unprimed T98G cells in figure 1A a ± SEM (Gy ¹ ) a _s ± SEM (Gy ¹ ) d _c ± SEM (Gy) β ± SEM (Gy ² ) χ ² /ν*

0.33 ± 0.03 3.39 ± 1.1 1 0.184 ± 0.05 0.020 ± 0006 1.1 1

* Chi-squared divided by the number of degrees of freedom associated with chi-squared

S = ^~ cd - d ² ) (equation 1)

describes the linear quadratic model. S is the surviving fraction, d the dose and a and β the parameters describing the linear and quadratic component, respectively, of the intrinsic radiosensitivity.

In the IR-model a is replaced by: J (equation 2) where d is dose, a _r is the value of a extrapolated from the high dose LQ response (equation 2), and a _s is the actual value of a derived from the initial part of the curve (i.e. at very low doses). d _c is the dose where the change from a _s to a _r is 63% complete.

T98G cells primed with 0.22 Gy/h for 1 h lost HRS (Figure 1A). The response to 2 doses in the HRS range was investigated for T98G cells irradiated with 0.06 Gy/h for 1 h and 0.19 Gy/h (because of decay not 0.22 Gy/h) for 15 min. The lowest dose-rate of 0.06 Gy/h had the same priming effect as 0.22 Gy/h for 1 h and the response of these cells to HDR challenge doses of 0.2 and 0.3 Gy was significantly more resistant than control unprimed cells. However, as shown in Figure 1 C 15 minutes priming irradiation with 0.19 Gy/h did not affect the response to the HDR challenge doses, indicating that 15 minutes was too short for the effect to be induced. Example 2 IL-13 can replace foetal calf serum for the effects of LDR radiation

The fact that LDR irradiation of medium conditioned on cells could induce the active factor removing HRS only if serum was present during cell conditioning suggested that there was a factor in the serum that cultured cells do not produce. We guessed that this could be related to the immune system and literature studies pointed towards a possible candidate: lnterleukin-13 (I L-13). 11-13 can bind to two receptors IL-13Ra1 and IL-13Ra2 (Wills-Karp and Finkelman, 2008), one of which (IL-13Ra1) also binds to IL-4. Both interleukins were tested on both T98G cells (figure 2A) and T-47D cells (figure 2B). IL-13 in serum-free medium did not remove HRS itself but was sufficient to make LDR irradiation induce the active factor that removes HRS both in cells and in cell conditioned medium. However, LDR irradiation of cells with serum-free medium with IL-4 did not remove HRS indicating that the effect is induced by IL-13 binding to receptor IL-13Ra2. The cells receiving LDR irradiated cell conditioned serum-free medium with added IL-13 regained HRS within 2 weeks as was the case with cells receiving serum- containing medium from LDR irradiated cells (Edin et al., 2009a).

Example 3 Chemical NO-donor in combination with ROS induction mimics LDR radiation

IL-13Ra2 has been known to activate furin from pro-furin (Liu et al., 2009) as well as induce iNOS (Authier et al., 2008; Suresh et al., 2007) and TGF^ (Fichtner-Feigl et al., 2006;

Shimamura et al., 2010) . This lead us to hypothesize that the active factor removing HRS could be ΤΘΡβ. Furins pro-activate the ΤΘΡβ-ίΑΡ complex making it accessible for ROS activation by liberating TGFfi from LAP (Leitlein et al., 2001). NO produced by iNOS scavenges LAP

(Vodovotz et al., 1999) which inhibits the re-association of LAP to ΤΘΡβ. When this takes place intracellular^, TGF is secreted in the active form (Shao et al., 2008). Intracellular presence of activated TGF has been shown to lead to sustained ROS production (Sturrock et al., 2006) which activates TGF as well as upregulates TGF expression (Kim et al., 1989).

First, the hypothesis of ROS liberating TGF from LAP and of NO preventing re-association was tested.

The chemical NO-donor diethylamine nitric oxide (DEANO) (present for 24h before the cells were plated in fresh medium 1 day before challenge irradiation) did not affect HRS in T98G cells at the concentrations used (0.1 mM (figure 3A) and 1 mM (data not shown)). However, a HDR radiation dose of 0.3 Gy in combination with 0.1 mM DEANO mimicked the effect of LDR irradiation and removed HRS for at least 2 weeks. Also, a HDR radiation dose of 0.3 Gy in combination with 0.1 mM DEANO given to cell conditioned medium induced the putative factor. We have previously found that culturing T-47D or T98G cells with 4% oxygen in the gas phase (i.e. exposing the cells to cycling hypoxia) removed HRS. However, HRS was restored within 2 weeks after the cells were transferred to a normal C0 ₂ -incubator. Reoxygenation after hypoxia results in a burst of ROS (McCord, 1985). We therefore wanted to test whether this burst of ROS could have the same effect as HDR irradiation in combination with NO. DEANO was administered 2 h before the cells were brought from the hypoxia workstation to a normal incubator with 5% C0 ₂ in air. After 2 weeks the cells were tested for HRS and it was found that the cells exposed to DEANO during reoxygenation had not regained HRS (figure 3B).

Thus, the effect of LDR irradiation could be mimicked by a combination of a NO-donor and HDR radiation or hypoxia/reoxygenation-induced ROS production.

Example 4

iNOS inhibitor 1400W reversed the permanent elimination of HRS in LDR primed cells and induced HRS in a cell line without HRS

Our hypothesis was that the presence of intracellular TGF in cells with IL-13-induced iNOS would start a self-sustained process in which the production of NO by iNOS played a central role in maintaining the presence of activated ΤΘΡβ. If this was true, the cells in which HRS was permanently removed by LDR priming (T-47D-P and T98G-P) should recover HRS if we inhibited iNOS. The cells were exposed to medium with 10 μΜ iNOS inhibitor1400W on 3 consecutive days and were tested 24h and 3 (T98G) or 5 (T-47D) weeks (with 2 weekly reseeding) after the last treatment (figure 4A to D). At all time points the cells had regained wild type HRS.

NHIK 3025 human cervix cancer cells do not express HRS (Figure 5C). However, treatment with 10 μΜ 1400W added on three consecutive days induced the HRS response even in these cells (Figure 4E and F).

Example 5

iNOS inhibitor 1400W prevented the effect of LDR irradiation of cell conditioned medium

The fact that LDR irradiation of cell conditioned medium could induce the factor that removed HRS and that this mechanism could be mimicked by HDR irradiation in combination with NO (figure 3A) suggested the presence of iNOS in the cell conditioned medium. To verify this, 10μΜ 1400W was added to the cell conditioned medium prior to LDR irradiation and we found that transfer of this medium to unirradiated cells did not remove HRS (figure 4A and C). We next tested the hypothesis of TGFfi as the active factor. Example 6

Neutralization of TGFfi3 inhibits the effect of LDR irradiation

100 μΜ pan-specific ΤΘΡβ antibody, which neutralizes the biological activity of ΤΘΡβΙ , ΤΘΡβ2, ΤΘΡβ3, ΤΘΡβ1.2 and Γ8ΤΘΡβ5, was added to cell conditioned medium prior to LDR irradiation. Unirradiated T98G cells were exposed to this medium for 24 h before being plated for colony formation in fresh medium and challenge irradiated. These cells retained HRS indicating that TGF is indeed involved in the elimination of HRS by LDR irradiated cell conditioned medium (figure 5A and 5D).

We next tested neutralizers for specific ΤΘΡβε and found no effect of ΤΘΡβΙ or ΤΘΡβ2 neutralizers, but a full effect of ΤΘΡβ3 neutralizer.

Example 7

TGFfi3 removes HRS

Cells were then exposed to medium containing 1 ng/ml recombinant ΤΘΡβ3 for 24 h before being plated for colony formation in fresh medium 18 h before challenge irradiation (the same procedure as used in all medium transfer experiments). In both cell lines HRS was removed by ΤΘΡβ3 (figures 5B, C, E and F). Lower concentrations of ΤΘΡβ3 were tested on T98G cells and both 0.01 and 0.001 ng/ml ΤΘΡβ3 were found to remove HRS (figures 5B and E). However, HRS returned within 2 weeks after exposing T98G cells to 0.01 ng/ml ΤΘΡβ3 (Figures 5B and 5E).

Example 8

TGFfi3 levels are increased in the cytoplasm of LDR irradiated cells

Post-embedding immunogold electron microscopic analysis using the same ΤΘΡβ3 antibody that was used for neutralization (figure 5A) showed a significant increase in active ΤΘΡβ3 in the cytoplasm of LDR irradiated T-47D and T98G cells (P<10 ^"11 , Student's t-test) (figure 6).

Example 9

Testing of mouse serum in a cell model

Cells received medium with 4% mouse serum and 6% foetal calf serum. 24 h later cells were seeded for colony formation in fresh medium and 16-20 h after seeding, the cells were challenge irradiated. Left panel shows experiments done with T-47D cells , right panel shows experiments with T98G cells. Surviving fraction is shown for each condition used in Figure 8. The data show that adding 4% serum from unirradiated mice to the medium did not remove HRS (low dose hyper-radiosensitivity). However, serum from mice exposed to 0.3 Gy/h for 1 h removed HRS in recipient cells, even when the serum was harvested 4 weeks after irradiation. These data indicate that the same factor that we have seen in cell models after low dose irradiation is also induced in vivo (Figure 8).

Example 10

Low dose priming protects the cell against cisplatin

T-47D cells and low dose-rate (LDR) primed T-47D cells (T-47D-P) were exposed to different doses of cisplatin.

The data show that LDR irradiation not only protects the cells against subsequent irradiation but also against chemo therapeutics, which mimic radiation effects (Figure 9).

Example 11

Microarray data comparing mRNA levels in LDR irradiated T-47D cells (T-47D-P) with control T- 47D cells and HDR irradiated T-47D cells (harvested 24h after irradiation, a time where HRS had returned).

The microarrays were preprocessed in R and the "loess" method was used for normalization. The three microarrays "HDR vs T-47D" was compared to the three "HDR vs T-47D-P" microarrays using limma package. The differentially expressed genes between the two groups were adjusted with FDR (using topTable method in limma).

T-47D-P vs HDR T-47D-P vs C T-47D-P vs T-47D-P vs

GeneName logFC logFC HDR B C B

CXCR4 2.199288558 2.29805806 4.807731251 4.80873067

ADM 1.702869794 2.474792485 2.026679266 8.255104407

VEGFA 1.402087461 2.209253147 2.49475675 11.29442296

DDIT3 1.345815947 2.498001981 3.0311 14301 6.1913271 18

JUN 1.27068357 2.239465177 0.190549426 7.746249955

C1orf63 1.266891306 1.143792553 3.075104565 4.833123038

HSPA1A 1.264408259 1.669151687 1.150064613 7.064804941

HIST2H4A 1.177556868 0.749827187 10.81567179 3.372606263

VEGFA 1.158860205 1.555097364 2.49475675 11.29442296 ARRDC3 1.128668485 1.417137478 0.474958107 4.112305299

ZFP36 1.108109248 0.977928563 6.371835281 3.786111305

ID2 1.093771646 0.938636815 1.907674655 3.515309766

ZMAT4 1.065894098 0.535179903 3.374191522 1.678813061

ID2 1.035034199 0.845108838 1.907674655 3.515309766

ATF3 1.023695131 1.52713035 1.608571519 4.795116116

HIST1 H4I 1.022151099 0.60162202 6.993858357 3.133394336

LOX 1.012677897 0.805178828 0.788421343 2.601861205

TMEM45A 1.005076317 1.254409097 1.689595989 1.743285851

SCD 0.999777841 0.964151468 3.906249157 3.144998878

ATF3 0.990475652 0.860354046 1.608571519 4.795116116

ACTG1 0.974599377 0.805348506 1.170397036 2.05477811

HIST1 H3D 0.958637801 0.66113432 2.411017665 5.023599731

PER1 0.955512695 0.85378601 8.908649525 2.99675417

DUSP10 0.947423162 0.619411564 3.056153767 2.392360294

LOC 148709 0.946460037 0.731893428 2.411156109 0.154163401

HIST1 H4F 0.942848128 0.491770757 9.389662978 4.317508693

PFKFB4 0.930807701 1.477525819 0.995479469 3.02611754

TUFT1 0.915068763 0.733095268 0.389867946 1.433782843

ELF3 0.904758712 2.187650736 2.22696022 6.530944557

CBX4 0.884723336 1.557810603 1.913811388 3.431205519

PGK1 0.879620324 0.702898775 1.640763957 1.420529845

LOC751071 0.855421788 0.992828071 8.292700799 5.676354454

CEBPD 0.85071337 0.727300808 1.084708924 2.986355469

TNFAIP3 0.849332863 0.510266329 3.413961627 1.468948493

C1orf107 0.84900893 0.7996017 5.009409821 4.373417254

Z25424 0.843403081 1.273496716 1.735535147 3.745691442

ZNF395 0.832280453 1.077156114 0.3427376 7.222893251

ZMYND8 0.823425167 0.592119003 1.776424838 0.330983564

IRF1 0.797966802 0.720461833 5.019440663 3.623549497

BNIP3 0.795216196 1.338187144 2.304152658 4.096907665

AL122093 0.794906258 0.827752364 0.482880545 2.969450332

MKNK2 0.790094618 1.311925404 1.794605351 4.307698956

FUT11 0.78864326 0.835865119 1.763428505 2.291455183

RSRC2 0.782121885 0.332244985 4.43883025 0.446763962 HSPA6 0.781379064 1.648764008 2.606175535 4.120687414

MAFF 0.775852168 0.700443661 5.201754473 2.637519277

TMTC1 0.773082989 0.88902404 6.335131006 5.067187155

AP2B1 0.7545309 0.826252864 3.853556502 1.824882579

NR1 D1 0.751575783 0.586710159 0.002968178 0.825806634

GADD45B 0.75089866 1.176017261 4.704979897 6.32195534

ARID5B 0.749860828 0.648906778 3.978696241 4.027823295

CDKN1 B 0.748501253 0.6211112 0.24885203 5.115589574

ZBTB2 0.743300846 0.505702024 0.857561442 3.278496076

KLF11 0.739570607 1.327654954 1.651562602 5.220729984

PCDH1 0.724539634 1.237757051 1.836495392 2.93270881

PPP1 R15A 0.724343056 1.057715311 1.01601547 2.745096094

FOSL2 0.721217316 0.699770647 1.617585835 2.692488414

OXR1 0.70796662 0.423983083 5.516629038 3.524202731

ER01 L 0.70759898 0.806013543 0.293763023 2.613767297

LOC51233 0.67612725 0.631282725 3.06143744 1.665072768

TRIB3 0.665618107 1.290107372 0.132903907 4.739266679

KLF10 0.663252941 0.954106807 1.551002753 6.36171332

RHOH 0.63867241 0.377214902 0.558245137 0.197917385

DBP 0.632904178 0.855517292 1.680710684 1.076736978

ALDH3B2 0.625882945 0.327632416 2.362041631 1.265806802

TRIB1 0.620758155 0.373491427 0.757357536 0.109966729

OVGP1 0.618190856 0.331809955 5.201829792 0.444661667

KIAA1754 0.613977738 0.55003091 2.539085052 5.922653982

FLCN 0.6037313 1.148375625 2.187987431 4.018937576

ALB 0.601623603 1.014361453 1.482467444 4.993717554

ARID5A 0.588474342 0.905127064 1.796896162 6.829831192

BG993059 0.58578394 0.996916717 1.500024142 4.812619818

PDK1 0.582008484 1.646827081 0.521892517 8.045113343

ABCA12 0.574999343 0.929462996 1.023597513 2.046589372

GPRC5A 0.574852938 0.730518995 2.060061067 3.872429447

RORC 0.568020621 1.048287078 0.81919116 2.935284578

FBX028 0.565204888 0.388345595 3.153315851 0.205627608

HSPA1A 0.564461281 1.76725501 1.150064613 7.064804941

NDRG1 0.563350464 1.526556023 0.043804585 6.101955323 CLK3 0.563336176 0.679404793 1.528543975 2.25769382

MYO10 0.55794972 1.161980027 1.64347641 6.27676754

DCP2 0.555497324 0.587782259 3.287898009 4.200572814

BANP 0.554698857 0.887215762 0.791252035 5.240938084

CXCR7 0.553018963 0.843371983 0.794729325 8.431179879

FBX032 0.547312071 1.756281446 1.030612462 7.574872868

RUNDC2C 0.545182837 0.932153791 0.3776218 3.492023075

SERTAD3 0.53764413 0.422848968 1.470007574 1.150728698

SP1 0.531557276 0.373208137 0.157647503 2.900634682

NFX1 0.520724378 0.438469931 4.055037146 0.901926557

KLHDC7B 0.511916356 2.465655677 0.625839488 10.92232837

JMJD1A 0.501452792 1.650616994 1.388533397 9.591470647

ZHX2 0.499038237 0.432246711 1.364559978 0.526727774

C10orf12 0.493535828 0.55637735 1.144926783 3.262242851

IRS2 0.489153445 0.77285796 1.835410886 2.937443228

ZFY 0.488438909 0.314249913 3.568006998 1.396106049

BNIP3L 0.487781292 1.171321965 0.607063093 5.873772906

ZBTB10 0.483240912 0.862605856 1.346128148 5.54196018

BHLHB2 0.479959639 0.909119685 0.158291095 8.097958192

MTHFD1 L 0.478559866 0.513272557 0.194963116 2.506862251

CNNM2 0.475808856 0.432752804 0.374737536 0.068991942

WARS2 0.473299577 0.270843486 0.289140883 1.839046555

RNMT 0.469477275 0.548147353 1.921259259 1.275954695

DCTN4 0.466480117 0.425414767 5.054002454 0.477147414

GPR146 0.465632276 0.596668505 3.517919001 1.118937529

TNF 0.464835883 0.237872093 0.53104715 0.368764299

SLC25A38 0.464731013 0.528908011 3.44753896 2.38900293

MCC 0.46075217 0.311732539 6.584450477 0.7665667

MKI67IP 0.45780791 0.695189539 4.153056546 3.990161972

BCL2L11 0.457596632 0.392909962 1.904210068 1.047253121

DNAJC12 0.456423442 0.407755567 2.335558868 0.817652877

FNDC3B 0.455248447 0.581489843 1.395360955 4.141298393

CBLB 0.451842964 0.933001517 2.509482627 8.712246434

BANP 0.448042059 1.036951927 0.791252035 5.240938084

RNF111 0.443279763 0.372435569 1.054781496 0.017642498 STOM 0.436387523 1.102251643 0.023838615 4.504117454

TLK2 0.428441378 0.397182186 0.283044021 0.53523947

BRAF 0.419314284 0.283858675 0.282459738 0.970821939

LONP2 0.417813654 0.499368829 1.087020412 1.344914027

TRIO 0.409164931 0.63400961 4.080783366 5.792226518

EFNA4 0.407678325 0.320970222 1.734853407 2.588568017

ICAM1 0.405913738 0.550514793 0.017483423 2.656782972

ZNF323 0.403725312 0.407349183 0.838557797 1.110525826

TCP11 L2 0.399942131 0.493503276 0.22837312 0.113782749

SGMS2 0.390740707 0.510668324 0.536384523 1.158855178

AF086205 0.388989041 0.599777996 2.740874527 4.561457279

Bl 910665 0.386925168 1.60664602 1.598987553 10.98409676

M87790 0.374293185 0.547806716 1.789703909 6.022469694

RABGGTB 0.373040056 0.578731253 0.602920799 1.198409665

MTHFD1 L 0.37265139 0.492364294 0.194963116 2.506862251

RIPK2 0.364772372 0.458575359 1.253189553 0.872475397

NR1 D2 0.364321898 0.747434302 0.092650013 5.507862136

IL1 R1 0.358690741 0.396704685 1.925214084 0.555327919

AV701505 0.346953648 0.350416232 0.997460197 0.962623599

ZNF295 0.342736068 0.406599017 2.655464118 4.742612379

DIP2B 0.336301892 0.347678983 2.596077384 1.222285762

ZNF395 0.321837843 0.881116074 0.3427376 7.222893251

GPRC5A 0.316925054 0.307648948 2.060061067 3.872429447

SEC31A 0.312681223 0.645278542 0.008110993 4.557392091

POLR3D 0.306585238 0.517159269 1.428376967 3.443826136

ACBD3 0.289836333 0.849289663 2.612691304 7.901847449

EXDL2 0.276401856 0.467295449 0.211381228 3.524362622

SEC31A 0.26997838 0.749327117 0.008110993 4.557392091

IGF1 R 0.250463073 0.313255212 0.122711827 2.444378525

ZFP36L1 0.247651413 0.607524132 0.203525709 4.361307918

KIAA1345 0.208647732 0.277659179 1.805196411 0.892417607

THC2677499 -0.186201293 -0.406633446 0.013827633 3.402164647

C8orf51 -0.253152036 -0.532410228 2.868861007 5.350660054

C1QTNF6 -0.262369171 -0.337753017 0.878261167 1.194297422

ENST00000378820 -0.267809244 -0.471456469 0.547627547 2.332199537 TMEM51 -0.272922688 -0.430680933 0.377055909 1.239342887

THC2541607 -0.290553695 -0.337332549 0.731269754 1.52872708

GMIP -0.300455552 -0.462600529 2.301249279 1.829505069

AGPAT3 -0.301018794 -0.531679387 0.871598317 3.254138015

AK124576 -0.306664087 -0.543751402 0.161542303 4.169324175

MGC23284 -0.323435173 -0.332858469 0.349154162 0.84548191

TPM3 -0.325353163 -0.332322721 0.445689005 4.617101864

PPARG -0.328021606 -0.526578068 0.414520612 3.44175924

PPARG -0.346520666 -0.526578068 0.414520612 3.44175924

NEK3 -0.347094199 -0.573447077 1.664796594 3.931971018

C9orf89 -0.351967601 -0.455126117 0.718840094 1.178605306

TNNI3 -0.357376914 -0.663976605 0.929738532 0.624454912

NXF5 -0.383817658 -0.717182167 0.938016375 4.251990964

LYNX1 -0.383828587 -0.735974359 1.571501876 3.228619162

TSPAN3 -0.40335319 -0.598674409 1.802485765 4.847181518

MTERFD3 -0.413194504 -0.883111687 1.230099691 5.093685735

CNOT6 -0.416632562 -0.516535086 0.676240534 2.204335255

BFSP2 -0.417054773 -0.556574219 0.455305206 2.632951332

OLFM1 -0.426337849 -1.41420545 0.676052923 8.532439227

PELI3 -0.429427698 -0.596524303 2.284650868 1.5454809

OLFM1 -0.43359529 -1.104496268 0.676052923 8.532439227

AK055981 -0.436532315 -0.569822957 0.313901993 0.648785351

S100A13 -0.439511316 -0.710744171 0.305139292 2.021529447

AK057088 -0.44860656 -0.387374945 1.529863586 2.404042523

THC2655527 -0.451317123 -0.581293525 2.281597398 4.255230291

ZNF780B -0.45596554 -1.000271162 0.112357916 3.910049013

NXF2 -0.456147881 -0.615323691 3.00016015 2.451597203

GATA4 -0.477662486 -0.603229868 1.257050986 3.923516168

CHCHD7 -0.481684823 -0.956412917 0.831325235 5.628053899

GNB1 L -0.484148224 -0.568420281 1.452476539 0.596828729

DGKZ -0.501288837 -0.423980257 1.9670611 2.740742514

BLMH -0.503663985 -0.417444994 2.110396073 1.920436217

SNN -0.508272803 -1.360682021 5.587504099 6.436256104

NFE2 -0.534626888 -0.510560307 2.741826915 1.01630997

PHTF1 -0.539608965 -0.549275175 2.960340403 1.094113045 CHST8 -0.573043855 -0.6376462 4.878918181 4.081984004

CAPG -0.586333392 -0.541406178 0.971259144 4.066762667

TP53AP1 -0.597351844 -0.442509748 0.334555348 6.160570973

ITPKA -0.604037246 -1.128687791 1.958436855 3.578381687

ENST00000322831 -0.605087223 -0.468873066 2.758164615 2.425453956

ASRGL1 -0.606004942 -1.160130957 0.26018443 7.666658585

GGA2 -0.614782784 -0.692870432 2.831905758 5.067894644

ZNF219 -0.617129982 -0.807333166 0.883358893 2.247661851

B4GALT5 -0.625980777 -1.03511045 0.688954066 3.615643597

ASRGL1 -0.628707612 -1.324276063 0.26018443 7.666658585

SLC30A3 -0.634205636 -0.834408361 1.412363081 1.256750347

CHCHD6 -0.636692364 -0.439949692 3.695257363 1.506435435

ZNF219 -0.650107087 -0.565980806 0.883358893 2.247661851

PPP1 R10 -0.65054381 -1.052307419 4.206394583 5.495655097

C1orf38 -0.656137741 -0.759334617 1.885343197 3.508409262

THC2650668 -0.694646539 -0.801833657 0.236204049 0.666857366

MUM1 -0.732779588 -0.819443372 4.303417573 3.27978562

C9orf58 -0.754554097 -0.405559973 4.408839764 2.166688778

KLK8 -0.762896027 -0.99910682 1.073334932 0.704050517

SHANK2 -0.872974743 -0.882995357 5.148535024 1.865456153

THC2721785 -1.075482066 -1.076903956 3.769435585 3.494797718

Example 12

iNOS was not uncoupled to produce superoxide in LDR primed cells

In order to test whether uncoupling of iNOS to produce superoxide was involved in the sustained elimination of HRS, EPR measurements of superoxide were done in LDR primed T98G and T-47D cells as well as in respective unprimed control cells. 100 μΜ 1-hydroxy-3- methoxycarbonyl-2,2,5,5-tetramethylpyrrolidine (CMH) was used as spin trap. 300 μΜ L-NAME was added to half the samples to inhibit NOS activity during spin trapping. Superoxide levels were higher in all cell types in the presence of L-NAME ruling out iNOS uncoupling (Figure 10).

Example 13 in vivo model

It was investigated if the same mechanism induced by LDR irradiation of cells in culture could be induced in vivo. DBA/2 mice with normal microbial flora were given whole-body irradiation for 1 h at low dose-rates of 0.3 or 0.03 Gy/h. T-47D and T98G cells were used as reporter system and it was found that serum from irradiated mice, but not from unirradiated controls, contained active TGFp3 which removed HRS in the reporter cells. Inhibiting iNOS reversed the mechanism resulting in active TGFp3 in the serum.

Example 9 shows that serum collected from whole body irradiated mice (0.3 Gy/h for 1 h) removed HRS in reporter cells (T-47D and T98G). No change was observed in the response to challenge irradiation in cells exposed to serum from unirradiated mice, In Figures 1 1 and 14 only the challenge doses to which the responses of the reporter cells were found to be significantly different were examined (i.e. 0.2 and 0.3 Gy). In order to see if the effect of the low dose-rate irradiation was persistent in mice, mice were allowed to live for 12 (females) or 15 (males) months after irradiation. Even so long after irradiation the mice still produced the factor that removed HRS in the reporter cells (figure 1 1) (p=0.0003 and 0.00005 (0.2 and 0.3 Gy data pooled)). Serum from unirradiated control mice of the same age did not remove HRS, although cells exposed to serum from unirradiated male mice appeared to have slightly less pronounced HRS that those exposed to serum from unirradiated female mice.

When TGFp3 neutralizer was added to the medium containing serum from irradiated mice before transfer to the reporter cells the effect of the mouse serum on HRS in the reporter cells was inhibited (Figure 12) (p=0.0006 (0.2 and 0.3 Gy data pooled)) . Irradiated mice (0.3 Gy/h for 1 h 1 week before) were then treated with 1400W intraperitonally every 6th h for 5 days. The serum harvested from these mice did not affect HRS in the reporter cells (Figure 13). It should be noted that the treatment with 1400W appeared to make the mice aggressive.

Cell studies have shown that the low dose-rate could be reduced by a factor 5 with the same exposure time and still have the same effect. In the mouse model the dose-rate (and dose) was reduced by a factor 10 to 0.03 Gy/h for 1 h. Still serum from these mice was able to remove HRS in the reporter cells (Figure 14) (p=0.0003 (0.2 and 0.3 Gy data pooled)).

This shows that the long-lasting activation of TGFp3 after exposure to 1 h of LDR irradiation, was also induced in an in vivo mouse model. Serum collected from unirradiated mice did not remove HRS in reporter cells regardless of the age of the mice. However, serum from mice exposed for 1 h to whole-body irradiation at a dose-rate of 0.3 Gy/h removed HRS in the reporter cells even 15 months after the exposure. The response of the reporter cells to doses above the HRS-range was not affected in concordance with experiments with LDR irradiation of cells (Edin et al. 2007). Also serum harvested 3 months after the mice had been irradiated with 1/10 of the dose-rate (0.03 Gy/h) for 1 h removed HRS. Increased cytoplasmic level of active TGFp3 was demonstrated in LDR primed cells. The present experiments verify the proposed mechanism in an in vivo model. The inhibitory effect of adding a neutralizing TGFp3-antibody on the reporter cell response is a strong indication that mice exposed to 1 h of LDR whole body irradiation have active TGFp3 in the blood. This response to LDR irradiation appears to last at least 15 months after the exposure but can be reversed by injections of iNOS inhibitor 1400W.

Discussion

The present study provides strong evidence that TGF 3 is the factor removing HRS after activation by LDR irradiation and also give some indications of the regulatory cascade activating TGF 3. Microarray analyses of LDR primed T-47D cells compared to unprimed or HDR primed (harvested 24h after irradiation when HRS had returned) T-47D cells did not show change in expression of the genes that code for the proteins that were shown to be involved (TGF 3 or iNOS) in the elimination of HRS or others which might explain the observed effects. Also, we have previously found that inhibiting protein synthesis during irradiation of T-47D cells did not inhibit the effect of LDR irradiation (Edin et al., 2009a). Thus, the mechanisms induced by LDR irradiation most likely involve protein activation or modifications.

The possible regulatory pathway

Figure 7 illustrates a model for the molecular cascade involved in the protective effect based on our previously published studies (Edin et al., 2011 ; Edin et al., 2007; Edin et al., 2009a; Edin et al., 2009b) and tested in the present study.

1. The first step is IL-13 binding to receptor IL-13a2 (figure 7A). Without IL-13 in the serum- free medium during cell conditioning, LDR irradiation did not induce active TGF 3 even when serum was added before the irradiation. However, addition of IL-13 to serum-free medium restored all the effects of LDR irradiation seen when using medium with foetal bovine serum (figure 2). Addition of IL-4 to serum-free medium did not have that effect indicating that the mechanisms involved are induced by IL-13 binding to receptor IL-13a2 which is specific for IL- 13 (Kawakami et al., 2001).

IL-13a2 has been shown to be involved in the conversion of pro-furin to active furin (Liu et al.,

2009) and in upregulation of iNOS (Authier et al., 2008; Suresh et al., 2007). IL-13a2 has also been found to induce TGF i promoter activity (Fichtner-Feigl et al., 2006; Shimamura et al.,

2010) but it is not known whether this also applies to TGF 3.

We propose that IL-13 is required because it induces iNOS and proactivation of the TGF 3 - LAP3 (latency-associated protein) complex trough activation of furin (Dubois et al., 1995). The proactivation of a ΤΘΡβ-Ι_ΑΡ complex by furin consists of proteolytic cleavage of the ΤΘΡβ-Ι_ΑΡ complex leaving ΤΘΡβ non-covalently associated with LAP (Annes et al., 2003).

2. The cells secrete iNOS and the non-covalently associated TGFP3 - LAP3 complex (figure 7A). The fact that LDR irradiation of cell conditioned medium without cells present induce active ΤΘΡβ3, indicate that the non-covalently associated ΤΘΡβ3 - LAP3 complex is present in the cell conditioned medium. The effect of LDR radiation on cell conditioned medium could be mimicked by HDR irradiation or reoxygenation after hypoxia in combination with NO (from DEANO) (figures 3 and 4). However, LDR irradiation did not induce the active factor in the cell conditioned medium in the presence of iNOS inhibitor 1400W (figure 4). Thus iNOS must be present in the cell conditioned medium.

3. Low dose-rate irradiation produces ROS continuously over time (the whole irradiation period). ROS activate TGFP3, and NO, which scavenges LAP3 (figure 7B). The

experiments shown in figures 3 and 4 give an indication of the mechanisms induced by LDR irradiation. NO can modify ΤΘΡβ activity by nitrosylation of LAP which interferes with the ability of LAP to neutralize ΤΘΡβ (Vodovotz et al., 1999). HDR irradiation, using the same dose of 0.3 Gy, did not activate ΤΘΡβ3 in sufficient amounts to remove HRS (Edin et al., 2009a; Edin et al., 2009b) except when a chemical NO donor was added before irradiation. We suggest that even though HDR irradiation induces ROS that can activate ΤΘΡβ3 by breaking the non-covalent binding to LAP3, LAP3 will scavenge ΤΘΡβ3 if NO is absent. NO production by iNOS is a slow process (Stuehr et al., 2004). We propose that iNOS is only activated after a certain period of sustained ROS production. Using a G-value of 2.8 for hydroxyl radicals induced by Co-60 γ- radiation in water (Yamamoto, 1982), 0.3 Gy/h is equivalent to -5.2 x 10 ¹³ radicals ml ^"1 h ^"1 or, in a cell with diameter 15 μηι, 0.3 Gy/h is equivalent to -9.3 x 10 ⁴ (2.1 x 10 ⁵ for 20 μηι) radicals ceirV or about 26 (61 for 20 μηι) radicals cell ^' V ¹ .

4. Intracellular active TGFP3 and NO starts a self-sustained process resulting in continued neutralization of LAP3 after uncoupling of iNOS (figure 7C). Cells that received medium containing ΤΘΡβ3 (from LDR irradiated cells or cell conditioned medium), recovered HRS within 2 weeks while the cells that were exposed to LDR irradiation permanently lost HRS. We propose that the permanent effect is induced by the intracellular presence of active ΤΘΡβ3 in combination with NO. The levels of active ΤΘΡβ3 in the cytoplasm were measured by post- embedding immunogold electron microscopic analysis using the same ΤΘΡβ3 antibody that was used for neutralization in figure 5A. The cytoplasmic ΤΘΡβ3 levels were significantly higher in the LDR irradiated cells, but there appeared to be a certain amount in the control cells as well. It is possible that this implies a threshold level of ΤΘΡβ3 for inducing the permanent response and for secreting enough to remove HRS. However, the specificity of the ΤΘΡβ3 antibody is given by the producer to be less than 5% cross reactivity with ΤΘΡβΙ and ΤΘΡβ2. In the experiments in figure 5A, where the antibodies were used as neutralizers, this did not affect the results because the specificity of the ΤΘΡβΙ and ΤΘΡβ2 antibodies were 100%. Thus, cross reactivity could account for some of the measured ΤΘΡβ3 levels in the control cells.

The first idea was that LAP3 was continuously scavenged by NO produced by iNOS. However, EPR measurements showed no difference in NO levels in LDR irradiated cells compared to controls (figure 4C). iNOS is coupled with NO generation in the presence of tetrahydrobiopterin (BH ₄ ) (Mayer and Hemmens, 1997). However, depletion of BH ₄ leads to uncoupling of iNOS, which then generates more superoxide than NO (Moens and Kass, 2006). Depletion of BH ₄ can be a result of oxidative stress but has also been found in response to ΤΘΡβΙ (Schoedon et al., 1993). The effect of ΤΘΡβ3 on BH ₄ has not been investigated. We propose that the presence of intracellular ΤΘΡβ3 results in uncoupling of iNOS through depletion of BH ₄ resulting in generation of superoxide. Superoxide reacts with NO producing peroxynitrite (ONOO ^" )- Peroxynitrite has been shown to nitrate the aromatic amino acids tyrosine, tryptophan and phenylalanine (Alvarez et al., 1996; van der Vliet et al., 1994) and may scavenge and inhibit LAP3 in a way similar to NO. In LAP1 , 2 and 3, a cysteine is present close to the binding site for ΤΘΡβ, which could be the target for NO inhibition of LAP activity. In LAP3, but not LAP1 or 2, a phenylalanine is located close to this, and could be a specific target for LAP3 inhibition. Hence, ΤΘΡβ3 remains active even after cell division and is secreted in this form. The active ΤΘΡβ3 binds to a receptor on the cell membrane which leads to increased ability to activate the early G ₂ -checkpoint for doses normally in the HRS-range. The exact mechanism for this remains to be investigated and other ΤΘΡβ3 induced pathways can not be ruled out.

In order to test the role of iNOS in maintaining the intracellular production of ΤΘΡβ3, we added iNOS inhibitor 1400W to LDR primed cells that had been grown for 5 month without recovering HRS. Consistent with our theory, HRS was recovered in these cells (figure 4).

The regulatory pathway is not induced by DNA damage

The suggested mechanism is not associated with DNA-damage activated pathways, which dominate radiation responses following large doses or acute irradiation. This is contrary to the general view of how radiation triggers cellular responses. Even with so-called non-targeted effects such as genomic instability, bystander effects, adaptive responses and low dose hyper- radiosensitivity, the common comprehension has been that DNA-damage is the factor triggering the process although the effects are clearly transmitted through other mechanisms. Our previous data showing that a factor removing HRS was induced by low dose-rate irradiation of cell conditioned medium (Edin et al., 2009a), however, suggested that DNA is not the target for the HRS-modifying mechanisms. In concordance with this, Maeda et al. (2008) previously reported that irradiation of nuclei using X-ray microbeams resulted in more pronounced HRS than whole-cell irradiation of V79 Chinese hamster cells and proposed that cytoplasmic irradiation either suppresses HRS or enhances IRR. To our knowledge, the present study is the first to propose a mechanism for radiation induced responses that are independent of DNA- damage and nucleus reactions and is selectively induced by low dose-rate irradiation, not by acute irradiation.

Hypoxia

A question that arises in connection with the data concerning hypoxia, is why HRS was recovered after reoxygenation of the long term hypoxic cells (figure 3B). TGFp3 has been shown to be upregulated and activated by HIF-1 in hypoxic placentae (Caniggia et al., 2000). An explanation could be that the stabilization of HIF-1a after each reoxygenation induce TGF 3. During chronic hypoxia HIF-1 levels decrease, but in the experiments the effect of adding TGF 3 to medium was measured up to 24 h after the medium with TGF 3 was replaced, which indicates that HIF induced TGF 3 would be present at all times between reseedings and medium changes. This is also consistent with the observation that HRS was not recovered 48 h after reoxygenation (figure 3B). However, when DEANO was added before reoxygenation, an effect similar to that of giving HDR irradiation in combination with NO was induced, which was independent of the HIF-1 induced activation of TGF 3. In support of this, inhibiting iNOS in LDR primed cells recovered HRS (figure 4) but exposing the LDR primed cells to cycling hypoxia did not.

Conclusion

In conclusion, we have found that TGF 3 removes HRS in T98G and T-47D cells and ΤΘΡβ3 levels are elevated in LDR irradiated cells . HRS is removed by LDR irradiation of cells, medium transfer from LDR irradiated cells or transfer of LDR irradiated cell conditioned medium. In all three cases, LDR irradiation could be mimicked by HDR irradiation in combination with NO from a chemical NO-donor. Inhibiting iNOS during LDR irradiation of cell conditioned medium prevented the effect of the radiation. LDR irradiation of T98G or T-47D cells removes HRS permanently, but the elimination of HRS could be reversed by adding the iNOS inhibitor 1400W. The effect of LDR radiation depended on cell conditioning in the presence of IL-13.

This application not only provides mechanisms to turn low dose resistance on (by TGFp3) or off (by iNOS inhibition) in cells, but also contributes to the understanding of why it is so important to distinguish between acute radiation and low dose-rate radiation when discussing radiation protection issues. It is evident that the collective dose concept is not valid for low dose-rates.

References

Alvarez, B., Rubbo, H., Kirk, M., Barnes, S., Freeman, B.A., and Radi, R. (1996). Peroxynitrite- dependent tryptophan nitration. Chem Res Toxicol 9, 390-396.

Annes, J. P., Munger, J.S., and Rifkin, D.B. (2003). Making sense of latent TGFfi activation. Journal of Cell Science 116, 217-224.

Anver MR, Haines DC (2004). Chapter 21 - Gerontology. In: Professor Hans JH, Professor Gillian B ed. The Laboratory Mouse. London, Academic Press. Pp. 327-343.

Authier, H., Cassaing, S., Coste, A., Balard, P., Gales, A., Berry, A., Bans, V., Bessieres, M.H., and Pipy, B. (2008). lnterleukin-13 primes iNO synthase expression induced by LPS in mouse peritoneal macrophages. Molecular Immunology 45, 235-243.

Bachman, K.E., Argani, P., Samuels, Y., Silliman, N., Ptak, J., Szabo, S., Konishi, H., Karakas, B., Blair, B.G., Lin, C, et al. (2004). The PIK3CA gene is mutated with high frequency in human breast cancers. Cancer Biology & Therapy 3, 772-775.

Bakkenist, C.J., and Kastan, M.B. (2003). DNA damage activates ATM through intermolecular autophosphorylation and dimer dissociation. Nature 421 , 499-506.

Bartek, J., Iggo, R., Gannon, J., and Lane, D.P. (1990). Genetic and immunochemical analysis of mutant p53 in human breast cancer cell lines. Oncogene 5, 893-899.

Dubois, CM., Laprise, M.H., Blanchette, F., Gentry, L.E., and Leduc, R. (1995). Processing of transforming growth factor beta 1 precursor by human furin convertase. J Biol Chem 270,

10618-10624.

Edin, N.F.J., Olsen, D.R., Sandvik, J.A., Malinen, E., and Pettersen, E.O. (2011). Low dose hyper-radiosensitivity in T-47D cells is eliminated by chronic moderate hypoxia but returns after reoxygenation. International Journal of Radiation Oncology, Biology, Physics.

Edin, N.J., Olsen, D.R., Stokke, T., and Pettersen, E.O. (2007). Recovery of low-dose hyper- radiosensitivity following a small priming dose depends on priming dose-rate. International journal of low radiation 4, 69-86.

Edin, N.J., Olsen, D.R., Stokke, T., Sandvik, J.A., Ebbesen, P., and Pettersen, E.O. (2009a). Mechanisms of the elimination of low dose hyper-radiosensitivity in T-47D cells by low dose-rate priming. International journal of radiation biology 85, 1157-1 165.

Edin, N.J., Sandvik, J.A., Olsen, D.R., and Pettersen, E.O. (2009b). The Elimination of Low- Dose Hyper-radiosensitivity by Transfer of Irradiated-Cell Conditioned Medium Depends on Dose Rate. Radiation Research 171 , 22-32. Fichtner-Feigl, S., Strober, W., Kawakami, K., Puri, R.K., and Kitani, A. (2006). IL-13 signaling through the IL-13a2 receptor is involved in induction of TGF-βΙ production and fibrosis. Nature Medicine 12, 99-106.

Gillespie, C.J., Chapman, J.D., Reuvers, A. P., and Dugle, D.L. (1975). The inactivation of Chinese hamster cells by x rays: synchronized and exponential cell populations. Radiation Research 64, 353-364.

Joiner, M.C., Lambin, P., Malaise, E.P., Robson, T., Arrand, J.E., Skov, K.A., and Marples, B.

(1996). Hypersensitivity to very-low single radiation doses: its relationship to the adaptive response and induced radioresistance. Mutat Res 358, 171-183.

Joiner, M.C., Marples, B., Lambin, P., Short, S.C., Turesson, I., (2001). Low dose

hypersensitivity: current status and possible mechanisms. Int J Radiant. Oncol. Biol. Phys, 49,

379-89.

Juan, G., Traganos, F., James, W.M., Ray, J.M., Roberge, M., Sauve, D.M., Anderson, H., and Darzynkiewicz, Z. (1998). Histone H3 phosphorylation and expression of cyclins A and B1 measured in individual cells during their progression through G2 and mitosis. Cytometry 32, 71- 77.

Kawakami, K., Takeshita, F., and Puri, R.K. (2001). Identification of distinct roles for a dileucine and a tyrosine internalization motif in the interleukin (IL)-13 binding component IL-13 receptor alpha 2 chain. J Biol Chem 276, 251 14-25120.

Kim, S.J., Jeang, K.T., Glick, A.B., Sporn, M.B., and Roberts, A.B. (1989). Promoter sequences of the human transforming growth factor-βΐ gene responsive to transforming growth factor-βΐ autoinduction. Journal of Biological Chemistry 264, 7041-7045.

Krueger, S.A., Collis, S.J., Joiner, M.C., Wilson, G.D., and Marples, B. (2007) Transition in survival from low-dose hyper-raadiosensitivity to increased radioresistance is independent of activation of ATM Ser1981 activity. International Journal of Radiation Oncology Biology Physics 69(4): 1262-1271.

Lambin, P., Marples, B., Fertil, B., Malaise, E.P., and Joiner, M.C. (1993). Hypersensitivity of a human tumour cell line to very low radiation doses. International journal of radiation biology 63, 639-650.

Leitlein, J., Aulwurm, S., Waltereit, R., Naumann, U., Wagenknecht, B., Garten, W., Weller, M., and Platten, M. (2001). Processing of immunosuppressive pro-TGF^1 ,2 by human

glioblastoma cells involves cytoplasmic and secreted furin-like proteases. Journal of

Immunology 166, 7238-7243.

Liu, T.F. , Cai, J., Gibo, D.M., and Debinski, W. (2009). Reoxygenation of hypoxic glioblastoma multiforme cells potentiates the killing effect of an interleukin-13-based cytotoxin. Clinical Cancer Research 15, 160-168. Marples, B., and Joiner, M.C. (1993). The response of Chinese hamster V79 cells to low radiation doses: evidence of enhanced sensitivity of the whole cell population. Radiation Research 133, 41-51.

Marples, B.; Lam, G. K.; Zhou, H.; Skov, K. A.: The response of Chinese hamster V79-379A cells exposed to negative pi-mesons: evidence that increased radioresistance is dependent on linear energy transfer. Radiation Research 1994, 138, S81-4

Marples, B., and Joiner, M.C. (1995). The elimination of low-dose hypersensitivity in Chinese hamster V79-379A cells by pretreatment with X rays or hydrogen peroxide. Radiat Res 141 , 160-169.

Marples, B., Wouters, B.G., and Joiner, M.C. (2003). An association between the radiation- induced arrest of G2-phase cells and low-dose hyper-radiosensitivity: a plausible underlying mechanism? Radiation Research 160, 38-45.

Matsumoto, H., Shimura, M., Omatsu, T., Okaichi, K. , Majima, H., and Ohnishi, T. (1994). p53 proteins accumulated by heat stress associate with heat shock proteins HSP72/HSC73 in human glioblastoma cell lines. Cancer Letters 87, 39-46.

Mayer, B., and Hemmens, B. (1997). Biosynthesis and action of nitric oxide in mammalian cells. Trends Biochem Sci 22, 477-481.

McCord, J.M. (1985). Oxygen-derived free radicals in postischemic tissue injury. N Engl J Med 312, 159-163.

Moens, A.L., and Kass, D.A. (2006). Tetrahydrobiopterin and cardiovascular disease.

Arterioscler Thromb Vase Biol 26, 2439-2444.

Nigro, J.M., Baker, S.J., Preisinger, A.C., Jessup, J.M., Hostetter, R. , Cleary, K., Bigner, S.H., Davidson, N., Baylin, S., Devilee, P., et al. (1989). Mutations in the p53 gene occur in diverse human tumour types. Nature 342, 705-708.

Norbye K, Oftebro R (1969). Establishment of four new cell strains from human uterine cervix I. Experimental Cell Research 58: 458.

Oftebro R, Nordbye K (1969). Establishment of four new cell strains from human uterine cervix II. Experimental Cell Research 58: 459-460.

Pettersen, E.O., Oftebro, R., and Brustad, T. (1973). X-ray inactivation of human cells in tissue culture under aerobic and extremely hypoxic conditions in the presence and absence of TMPN. International Journal of Radiation Biology and Related Studies in Physics, Chemistry, and Medicine 24, 285-296.

Schettino, G.; Folkard, M.; Prise, K. M.; Vojnovic, B.; Bowey, A. G.; Michael, B. D.: Low-dose hypersensitivity in Chinese hamster V79 cells targeted with counted protons using a charged- particle microbeam. Radiation Research 2001 , 156, 526-34.

Schoedon, G., Schneemann, M., Blau, N., Edgell, C.J., and Schaffner, A. (1993). Modulation of human endothelial cell tetrahydrobiopterin synthesis by activating and deactivating cytokines: new perspectives on endothelium-derived relaxing factor. Biochem Biophys Res Commun 196, 1343-1348.

Shao, C, Prise, K.M., and Folkard, M. (2008). Signaling factors for irradiated glioma cells induced bystander responses in fibroblasts. Mutation Research - Fundamental and Molecular Mechanisms of Mutagenesis 638, 139-145.

Shimamura, T., Fujisawa, T., Husain, S.R., Joshi, B., and Puri, R.K. (2010). Interleukin 13 mediates signal transduction through interleukin 13 receptor a2 in pancreatic ductal

adenocarcinoma: Role of I L- 13 Pseudomonas exotoxin in pancreatic cancer therapy. Clinical Cancer Research 16, 577-586.

Short, S.C., Kelly, J., Mayes, C.R., Woodcock, M., and Joiner, M.C. (2001). Low-dose hypersensitivity after fractionated low-dose irradiation in vitro. Int J Radiat Biol 77, 655-664. Stuehr, D.J., Santolini, J., Wang, Z.Q., Wei, C.C., and Adak, S. (2004). Update on mechanism and catalytic regulation in the NO synthases. Journal of Biological Chemistry 279, 36167-36170. Sturrock, A., Cahill, B., Norman, K., Huecksteadt, T.P., Hill, K., Sanders, K., Karwande, S.V., Stringham, J.C., Bull, D.A., Gleich, M., et al. (2006). Transforming growth factor-βΐ induces Nox4 NAD(P)H oxidase and reactive oxygen species-dependent proliferation in human pulmonary artery smooth muscle cells. American Journal of Physiology - Lung Cellular and Molecular Physiology 290.

Suresh, V., Mih, J.D., and George, S.C. (2007). Measurement of IL-13-induced iNOS-derived gas phase nitric oxide in human bronchial epithelial cells. American Journal of Respiratory Cell and Molecular Biology 37, 97-104.

van der Vliet, A., O'Neill, C.A., Halliwell, B., Cross, C.E., and Kaur, H. (1994). Aromatic hydroxylation and nitration of phenylalanine and tyrosine by peroxynitrite. Evidence for hydroxyl radical production from peroxynitrite. FEBS Lett 339, 89-92.

Vodovotz, Y., Chesler, L, Chong, H., Kim, S.J., Simpson, J.T., DeGraff, W., Cox, G.W., Roberts, A.B., Wink, D.A., and Barcellos-Hoff, M.H. (1999). Regulation of transforming growth factor β1 by nitric oxide. Cancer Research 59, 2142-2149.

Walton, K. L; Makanji, Y.; Chen, J.; Wilce, M. C; Chan, K. L; Robertson, D. M.; Harrison, C. A.: Two distinct regions of latency-associated peptide coordinate stability of the latent transforming growth factor-betal complex. J Biol Chem 2010, 285, 17029-37

Wills-Karp, M., and Finkelman, F.D. (2008). Untangling the complex web of IL-4- and IL-13- mediated signaling pathways. Sci Signal 1 , pe55.

Wouters, B.G., and Skarsgard, L.D. (1997). Low-dose radiation sensitivity and induced radioresistance to cell killing in HT-29 cells is distinct from the "adaptive response" and cannot be explained by a subpopulation of sensitive cells. Radiat Res 148, 435-442.

Xu, B., Kim, ST., Lim, D.S., and Kastan, M.B. (2002). Two molecularly distinct G(2)/M checkpoints are induced by ionizing irradiation. Molecular and Cellular Biology 22, 1049-1059. Yamamoto, O. (1982). Yields of hydroxyl radical and reducing species in tritiated water based on the radiolysis of tetranitromethane. International journal of radiation biology 42, 661-665.Xue, L; Yu, D. ; Furusawa, Y.; Cao, J.; Okayasu, R.; Fan, S.: ATM-dependent hyper-radiosensitivity in mammalian cells irradiated by heavy ions. Int J Radiat Oncol Biol Phys 2009, 75, 235-43.