WO2005026334A2 - Human tumour growth patterns - Google Patents

Abstract

The invention relates to an isolated tumour which comprises stem cells capable of asymmetric division and so giving rise to cells of a holoclone, meraclone or paraclone colony type which cells express characteristic markers and the exploitation of this information for the treatment of cancer.

Description

HUMAN TUMOUR GROWTH PATTERNS

The invention relates to methods and means for the diagnosis and treatment of cancer, in particular, the invention concerns especially, but not exclusively, an isolated tumour containing at least one population of homogenous cells which is characterised by the expression of at least one cell specific marker; said markers; antibodies and other reagents raised against said markers; and the use of said tumours, markers and antibodies in the diagnosis and treatment of cancers.

Cancer is a major cause of death in the UK with more than 267,000 new cases recorded in 1999. Moreover, the lifetime risk of developing cancer is more than 1 in 3 although the disease is more likely to develop in later life with around 65% of cancers diagnosed in people over the age of 65.

Breast, lung, large bowel (colorectal) and prostate cancers account for over half of all new cases. The most common cancer in men is prostate cancer which is responsible for almost a 5^th of all new cases. Bowel cancer is the third most common in men and together these two diseases account for half of the newly diagnosed cancers in men. In contrast, in women, breast cancer is by far the most common form accounting for 30% of all new cases. The second most common cancer is bowel cancer followed by lung cancer. These three cancers in women account for half of all the newly diagnosed cases. As mentioned above, cancer is more prevalent with advancing age but even in those under the age of 65 cancer is responsible for 36% of deaths. With these figures it is not surprising that there remains a considerable drive to develop new ways of treating the disease and so overcoming the illness.

It is well known that cancer is a disease where cells grow out of control and so expand in volume and invade and destroy neighbouring tissues. The causes of cancer are multifarious and the symptoms can vary enormously so, too, can the methods of treatment. However, if, despite this vicissitude, one could identify common growth patterns and mechanisms of growth control it would be possible to develop a common treatment programme to combat the disease. With this in mind we have investigated growth patterns in malignant tumours.

Normal tissue growth and renewal depends on a subpopulation of stem cells that has a high self renewal capacity. The stem cells of a normal epithelium divide, typically infrequently, both to renew themselves and to generate transit amplifying cells that enter the differentiation pathway and undergo a series of amplifying divisions before they become post-mitotic. This division is known as asymmetric division and is the mechanism that allows self-renewal of stem cells while generating hierarchies of amplifying cells that form the bulk of normal tissue. Maintenance of constant stem cell numbers results from a pattern of asymmetric stem cell division in which each cell division generates one stem cell that remains a stem cell and one that becomes committed to differentiation, but this homeostatic pattern is not fixed and can be shifted towards a greater frequency of symmetric divisions when it is necessary to increase stem cell numbers, as occurs during growth and wound healing.

Normal epithelia have an intrinsic ability to set up hierarchies of stem and amplifying cells in vivo and in vitro and, by plating at clonal densities, individual keratinocytes can be shown to have widely differing proliferative potentials. Some cells form round colonies composed of small compact cells that can be repeatedly passaged; others form irregular colonies capable of less extensive growth; yet others form colonies of large flattened cells that do not passage. These colony forms are referred to as holoclones, meroclones and paraclones and are thought to be derived from, and consist largely of, stem cells and early and late amplifying cells, respectively. The cells that form holoclones are also smaller and more rapidly adherent than other, less clonogenic cells.

The persistent growth of tumours indicates that they contain cells with the basic stem cell property of indefinite self renewal but it has been a matter of controversy whether malignant lesions also retain the hierarchical pattern of stem cell and amplifying cells typical of normal epithelia. The presence of two populations would have major implications concerning the behaviour and therapy of tumours.

There is some evidence for the persistence of stem cell patterns in haematological malignancies but whether, and to what extent, a hierarchical proliferative pattern is retained in solid tumours has been unclear. The need to postulate the existence of malignant stem cells has been denied by some but a recent report that tumour initiating cells can be identified as a small subpopulation of the cells isolated from breast tumours (Al Hajj et ai Proc. Natl. Acad. Sci. U.S.A. 100, 3983-3988 2003.) provides strong evidence for the presence of stem cells in a least one human solid tumour.

We anticipated that if stem cell patterns persist in malignancy, malignant cells should possess in vitro patterns of heterogeneity similar to those of normal cells. Moreover, if we were able to identify these patterns and show that they are consistently and repeatedly regenerated we anticipated that this information could be exploited in the treatment of cancer. More specifically, we anticipated that if we could identify the different cell populations within a tumour we would have the means for targeting those populations and thus more specifically attacking the disease. This reasoning is particularly important when one considers that, if a stem cell population is common to most tumours, then it will be possible to specifically target this population and so remove the regenerative properties of the tumour, both at the initial site and sites remote therefrom. Indeed, we hypothesise that one reason why cancer may reoccur is because of the existence of the stem cell population and the inability to annihilate same during treatment. This persistence of this cell type provides the basis for recurrent cancer.

Accordingly, our work has been undertaken in order to determine the heterogeneity of cells within tumours and, further, to reliably identify these cells using morphology and by determining the series of markers, both intracellularly and extracellularly (i.e. located on/in the cell membrane), that characterise these cells with a view to using, particularly, these markers as a means of studying the nature of the differences between these cells and of targeting to them agents for specific treatments.

STATEMENTS OF INVENTION

The invention is generally based upon the realisation that a tumour comprises a heterogenous population of cells whose growth patterns correspond to the clonal patterns one would expect to find in a normal cell population. This is an unexpected discovery because the prejudice in the art has, for many years, taught that a tumour contains a mass of cells that are essentially out of control. Thus the growth patterns of cells constituting the tumour should be erratic and ill-defined and so not pass through an organised sequence of events. Further, the prejudice in the art has, for many years, also taught that cell heterogeneity within a tumour arises from mutational changes associated with genetic instability rather than from an underlying physiological mechanism.

Accordingly, the invention provides an isolated tumour comprising any one or more of the following colony types: i. holoclone; ii. meraclone; or iii. paraclone; wherein cells of said holoclone colony are characterised by expression of any one or more of the markers shown in Tables 1A and 1 B and cells of the paraclone colony are characterised by expression of any one or more of the markers shown in Tables 2A and 2B.

Reference herein to the term marker includes reference to a gene or the product thereof that is shown, relatively consistently, to be highly expressed in one of the following colony types: holoclone, meraclone or paraclone having regard to its expression in any of the other colony types. Thus, whilst the product of the gene may be termed a marker in that it identifies a particular cell type the product of the gene may be involved in cell structure, cell signalling, cell metabolism or indeed any other cellular mechanism.

Reference herein to the colony types holoclone, meraclone or paraclone includes reference to a sub-population of cells within a tumour which are characterised by having the morphological or other properties that are common to these colony types, including expression of markers, or proliferative or migratory behaviours, characteristic of holoclone, meraclone or paraclone cells.

By examining cells isolated from tumours we have discovered that within each tumour there exists a development pattern which is parallel to that found in normal tissues and in cells from normal tissues in cell culture, i.e. cells in colonies that exhibit holoclone morphologies give rise to holoclones and also to colonies that exhibit meraclone morphologies which, in turn, give rise to colonies that exhibit paraclone morphologies. More preferably, the isolated tumour of the invention comprises a population of cells exhibiting holoclone morphologies and which are characterised by expressing any one or more of the markers shown in Tables 1A and 1 B. Most ideally, the holoclone population expresses the following markers which are expressed at a higher level, typically at least 50% higher, than in paraclone colonies:

Adaptor protein with pleckstrin homology. Aldehyde dehydrogenase;

ATP-binding cassette, sub-family A (ABC1 );

ABC transporter (WH1TE2);

ATP-binding cassette, sub-family B (ABCB3);

ATP-binding cassette transporter 1 (ABCA1 ); ATP-binding cassette (ABCG1 );

ATP-binding cassette, sub-family C (CFTRMRP);

ATP-binding cassette, sub-family D (ALD);

Binder of Rho GTPases 4 (CEP4);

Bisphosphoglycerate mutase (BPGM); Cadherin;

Cadherin-E (UVO);

Cadherin-K;

CCAAT enhancer binding protein (CEBP);

CDC42 GAP-related protein; Chloride channel 2; Chloride channel 3; Cytochrome P450; Cytochrome P-450mp; Cyclin-dependent kinase inhibitor p12; Death-associated protein kinase 1 (DAPK1 ); Dual specificity phosphatase 4 (DUSP4); Dual specificity phosphatase 6;

Dual-specificity tyrosine-(Y)-phosphorylation regulated kinase; EGF LAG seven-pass G-type receptor 2, flamingo homolog (CELSR2);

Epithelial membrane protein 1 ;

Erythroblasfic leukemia viral oncogene homolog 3;

Erythrocyte membrane protein band 4.1 -like;

Forkhead box C1 ; Frizzled homolog 3 (FZD3);

Frizzled homolog 10 (FZD10);

GTP cyclohydrolase I feedback regulatory protein (GCHFR);

HMG-box transcription factor TCF-3 (TCF-3);

HMG box factor SOX-13; Hurpin;

Hydroxyprostaglandin dehydrogenase;

Insulin-like growth factor-binding protein 3 (HGF dependent);

Insulin-like growth factor binding protein 3;

Insulin-like growth factor binding protein 6 (1GFBP6); Integrin alpha 10 subunit (ITGA10); Interleukin 2 receptor, beta (IL2RB); Kallikrein 10; Keratin 13; Keratin 15 (KRT15); Mesenchymal stem cell protein;

Mitogen-activated protein kinase 5 (MAP3K5);

Mucin 1 (MUC1 );

Mucin short variant (MUC1 ); Mucin (MUC3);

Mucin 4, tracheobronchial;

Myc protein;

Nebulette (NEBL);

Neuropilin 2; Notch homolog 3;

Pak5 protein;

Phosphoglucomutase 1 (PGM1 );

Pirin (PIR);

PML-1 ; Podocalyxin-like (PODXL);

Pregnancy-specific beta-1 -glycoprotein;

Prostate stem cell antigen (PSCA);

Proteoglycan 4;

Retinol dehydrogenase homolog (RDHL); Retinoic acid receptor responder (tazarotene induced) 3 (RARRES3); Retinoic acid receptor gamma; Ribosomal protein L3; Semaphorin III; Sialyltransferase (STHM); Signal transducer and activator of transcription 4 (STAT4);

Toll-like receptor 3 (TLR3);

Tumor necrosis factor (TNFSF10);

Tyrosine phosphatase, receptor type, R (PTPRR); Transcription factor 3, E2F;

Tumor necrosis factor, alpha-induced protein 2 (TNFAIP2);

Vav 1 oncogene (VAV1);

Vav 3 oncogene (VAV3);

V-erb-b2 avian erythroblastic leukemia viral oncogene homolog 3 (ERBB3); Zinc finger protein, X-linked (ZFX);

CD133;

CD44;

E-cadherin;

Alpha 6 and beta 1 integrins; Alpha and beta catenins;

Epidermal growth factor receptor; or

Proteoglycan 4.

Additionally or alternatively, the isolated tumor comprises a population of holoclone cells wherein the following gene products are found at generally or locally higher concentrations: CD133, CD44, e-cadherin, alpha 6 and beta 1 integrins, alpha and beta catenins, epidermal growth factor receptor and proteoglycan 4.

Additionally, or alternatively, the isolated tumour of the invention comprises a population of cells exhibiting paraclone morphologies and which are characterised by expressing any one of more of the markers shown in Tables 2A and 2B. Most ideally, the paraclone population expresses the following markers which are expressed at a higher level, typically at least 50% higher, than in holoclone colonies:

A kinase (PRKA) (gravin);

Alpha(1 ,3)-fucosyltransferase IV (FUTIV); Autophagy 12 (APG12L);

Avian myelocytomatosis viral oncogene;

Connective tissue growth factor;

Cysteine-rich, angiogenic inducer, 61 (CYR61 );

Cysteine-rich, angiogenic inducer, 61 ; Dickkopf homolog 1 (DKK1);

Dickkopf gene 3;

DNA-binding zinc finger(GBF);

Epithelial membrane protein 3 (EMP3);

Forkhead box D1 (FOXD1 ); Forkhead box F2 (FOXF2); Goliath protein; Growth arrest and DNA-damage-inducible 34 (GADD34);

GRO3 oncogene (GRO3); Hepatoma-derived growth factor;

HRP-3;

HSPC007 protein;

Inhibin, beta A (activin A, activin AB alpha polypeptide) (INHBA;

Inhibin ovarian beta-A; Insulin-like growth factor 1 receptor;

Insulin-like growth factor binding protein 7;

Integrin, beta 1 ;

Integrin, alpha E;

Interleukin 1-alpha (IL1A); Interleukin-1 receptor accessory protein (IL1 RAP);

Interleukin 6 (interferon, beta 2) (IL6);

Interleukin 8 (IL8);

Keratin 6A Type II;

Keratin 6A; Keratin 6B;

Keratin 6 isoform K6f (KRT6F);

Laminin, gamma 2 (nicein);

Low density lipoprotein receptor;

MAD1-like 1 (MAD1 L1 ); Mesenchymal stem cell protein DSC92; Meiotic recombination protein; Migration stimulation factor FN70; Mitotic arrest deficient-like 1 (MAD1 L1); Mortality factor 4 (MORF4); Nicotinamide N-methyltransferase (NNMT); NS1 -associated protein 1 (NSAP1); Nuclear localization signal deleted in velocardiofacial syndrome (NLVCF); Ornithine decarboxylase 1 (ODC1 ); Phospholipase C, beta 4; Potential tumor suppressor (ST7);

Regulated in glioma (RIG);

RNA binding motif, single stranded interacting protein 1 (RBMS1 );

SAM domain, SH3 domain and nuclear localisation signals, 1 (SAMSN1 ); Secreted frizzled-related protein 1 (SFRP1 );

Serumglucocorticoid regulated kinase (SGK);

Syntaxin 3A;

Thrombospondin 1 (THBS1 );

Thrombospondin 1 ; Thrombospondin 1 ;

Transcription factor Dp-2 (E2F dimerization partner 2);

Transcription factor IIF, polypeptide 2;

Vimentin;

Cytokeratins numbers 6, 14, 16, 17, 20; or Epcam.

Reference herein to gene products that are found at generally, or locally, higher concentrations includes reference to gene products whose concentrations are determined having regard to the total amount of a gene product within and/or withon a cell regardless of its particular location; or the concentration of a gene product at any one or more particular cellular loci, respectively.

Additionally, or alternatively, the isolated tumour comprises a population of paraclone cells wherein the following gene products are found at generally or locally higher concentrations: vimentin, DKK1 , DKK3, thrombospondin, cytokeratins number 6, 14, 16, 17 and 20.

In a yet further preferred embodiment of the invention said tumour is a carcinoma or other cancer derived from breast, prostate, lung, gastro-intestinal tract, head, neck, or other tissue. More preferably still said cancer is of human origin.

More preferably still said tumour has been treated to either increase or decrease expression of at least one of said aforementioned markers. Most typically, over- expression of a selected marker is undertaken by transfecting ideally, but not exclusively, a selected cell or cell population with multiple copies of a gene encoding said marker and/or transfecting said cell or cell population with a high expression promoter which is operatively linked to the gene encoding said marker whereby, in either case, the over-expression of said selected marker occurs in said selected cell population. Most ideally a cell, or cell population, displaying a holoclone morphology is transfected with at least one gene, or promoter controlling the expression of same, as listed in Table 1A or 1B. Alternatively, a cell or cell population displaying a paraclone morphology is transfected with at least one gene, or a promoter controlling the expression of same, listed in Table 2A or 2B.

Alternatively, and perhaps more conveniently, a tumour is transfected as afore described and in this instance the heterogeneous population of cells comprising said tumour is transfected with said gene or said promoter and, ideally, said high > expression promoter but, as will be apparent to those skilled in the art, the recombinant DNA is likely only to be expressed in the compatible cell type i.e. , the cells that would express the wild-type equivalent of that DNA.

In yet a further preferred aspect of the invention, an isolated tumour is treated so that it is transfected with a construct comprising the promoter of any one or more of the genes identified in Tables 1 A or 1 B, or 2A or 2B, and further wherein said promoter is operationally linked to a reporter molecule whereby the activity of said promoter in particular classes of cells is monitored by the expression of said reporter molecule in order to determine the action of drugs on the population of cells. Most typically, the increased activity of said promoter in said population of cells is used to drive the increased expression of said reporter molecule so that by monitoring the expression of said reporter molecule one can determine the action of drugs on the population of cells.

In a preferred embodiment of this aspect of the invention a homogeneous population of cells from said tumour is transfected with said construct and where the cells display a holoclone morphology they are transfected with a promoter of any one or more of the genes in Table 1A or 1B and where the cells display a paraclone morphology they are transfected with a promoter of any one or more of the genes in Table 2A or 2B. In both cases the promoter is operationally linked to at least one reporter molecule for monitoring the activity of drugs on the homogeneous population of cells. The reporter molecule may be endogenous or exogenous to said cell. Similarly, the promoter may be endogenous or exogenous to said cell.

Reference herein to the term endogenous includes reference to a molecule that would normally be found, or expressed, in/on a cell. Conversely, reference to the term exogenous includes reference to a molecule that would not normally be found, or expressed, in/on a cell.

According to a further aspect of the invention there is therefore provided a method for screening drugs to treat tumours wherein said method comprises:

(a) exposing a tumour, or heterogeneous population of tumour cells, which have been transfected with a promoter region of any one or more of the genes shown in Table 1A, Table 1B or Table 2A, Table 2B which promoter is operationally linked to its gene or a reporter molecule, to a test drug; and (b) monitoring the level of expression of said gene or reporter molecule in response to said drug and, optionally, where there is a reduction in expression concluding that the drug may be of use in treating said tumour, or, where there is a proportionate increase in the amount of the gene product or reporter molecule, due to the destruction or disabling of at least some cells of said tumour population, concluding that the drug may be of use in treating said tumour.

According to a further aspect of the invention there is provided a method for screening drugs to treat tumours wherein said method comprises: (a) exposing a cell, or population of cells, of either a holoclone or paraclone colony which has been transfected with a promoter region of any one or more of the genes shown in Tables 1A 1 B or Table 2A 2B, respectively, which promoter is operationally linked to its gene or a reporter molecule, to a test drug; and (b) monitoring the level of the expression of said gene or reporter molecule in response to said drug and, optionally, where there is a reduction in expression concluding that the drug may be of use in treating said tumour.

As mentioned above, where said cell or homogeneous population of cells has/have holoclone characteristics said promoter region(s) is selected from those linked to the genes in Tables 1A or 1 B. Alternatively, where said cell or homogeneous population of cells has/have paraclone characteristics the promoter molecule(s) is selected from those linked to the genes in Tables 2A or 2B. According to a further aspect of the invention there is provided a transfected cell or transfected population of cells adapted for monitoring the effects of drugs thereon. Ideally said population of cells comprises either a heterogeneous or a homogeneous population of cells, in the latter instance said homogeneous populations of cells is either of the holoclone, meraclone or paraclone colony type.

According to a further aspect of the invention there is provided a relatively homogenous population of cells derived from a tumour wherein said cells comprise cells derived from one of the following colony types: holoclone; meraclone; or iii. paraclone; wherein cells of the said holoclone colony are characterised by expressing any one or more of the markers shown in Tables 1A or 1 B and said cells of paraclone colony are characterised by expressing any one or more of the markers shown in Table 2A or 2B.

Reference herein to the term relatively homogenous includes reference to a population of cells that substantially comprise a singular cell type and thus, to all intents and purposes, can be said to comprise, and/or function, as though the population comprised only one cell type.

It will be apparent that a homogenous population of cells, as afore described, will have use as a tool for the identification of agents that are effective in treating cancer. This is because the homogenous population of cells will comprise markers which can be exploited for the purpose of either targeting a specific agent to a specific cell type and/or exploiting the characteristics of the marker(s) in the destruction or disabling of the cell.

Reference herein to the term 'destruction' or 'destroy' includes reference to doing away with, extinguishing, killing or annihilating a cell.

Reference herein to the term disabling includes reference to an instance where the cell may persist but its harmful effects may be annihilated or considerably reduced.

According to a further aspect of the invention there is provided an antibody to any one or more of the aforementioned markers. Most preferably, said antibody is monoclonal, although polyclonal antibodies may also be used in the exploitation of the invention. Most preferably still, said antibodies are humanised.

According to a further aspect of the invention there is provided use of said antibodies to treat cancer.

According to a further aspect of the invention there is provided a method for screening an agent in order to determine its ability to treat cancer comprising exposing an isolated solid tumour, as afore described, or a homogenous population of cells, as afore described, to said agent and then determining the effects of said agent on said tumour or said homogeneous cell population with a view to concluding, where said isolated tumour or said cell population is destroyed or disabled, that the agent has utility in the treatment of cancer.

According to a further aspect of the invention there is provided a non-human animal for use in identifying agents that are effective in treating cancer comprising a non-human animal that has a tumour wherein sa id tumour comprises any one or more cells corresponding to the following colony types: i. holoclone; ii. meraclone; or iii. paraclone; wherein cells of said holoclone colony are characterised by the exp ression , of any one or more of the markers shown in Table 1A and 1 B and cells of the paraclone colony are characterised by expression of any one or more of the markers shown in Tables 2A and 2B.

In a preferred embodiment said tumour in said non-human animal comprises an isolated tumour, as afore described, or a population of cells as afore described.

According to a further aspect of the invention there is provided an isolated cell- specific marker comprising any one or more of the markers shown in Tables 1 A or 1 B or Tables 2A or 2B. According to a further aspect of the invention there is provided the use of any one or more of the markers shown in Tables 1 A or 1 B or Tables 2A or 2B for the identification or development of an agent that treats cancer.

According to a further aspect of the invention there is provided the use of any one or more of the markers shown in Tables 1A or 1 B or Tables 2A or 2B for targeting agents to a specific cancer cell.

According to a further aspect of the invention there is provided the use of any one or more of the markers shown in Tables 1 A or 1 B or Tables 2A or 2B for use in destroying or disabling a cancer cell.

It will be apparent from the above that the invention provides for the identification of different cell types within a tumour and thus the efficient targeting of cancer treatments to these cell types. This is particularly important when one considers the proliferative nature of the stem cell population within the tumour. It follows that targeting this particular population of cells is desirable for the treatment of cancer.

According to a further aspect of the invention there is therefore provided a medicament for treating cancer wherein said medicament acts, either directly or indirectly, via one or more of the markers shown in Tables 1A or 1 B or Tables 2A or 2B. In this respect, the medicament exploits one or more of the markers shown in Tables 1 A or 1 B or Tables 2A or 2B in order to target the medicament to the cell type, or to destroy or disable said cell type.

According to a yet further aspect of the invention there is provided the use of any one or more of the markers shown in Tables 1 A or 1 B or Tables 2A or 2B or antibodies raised thereto in the manufacture of a medicament for the treatment of cancer.

According to a yet further aspect of the invention there is provided the use of any one or more of the markers, or a molecule that acts upstream or downstream thereof in a known biological pathway, in Tables 1A or 1 B or Tables 2A or 2B in the manufacture of a medicament for the treatment of cancer.

According to a yet further aspect of the invention there is provided a pharmaceutical composition comprising an agent that interacts with or interferes with, either directly or indirectly, any one or more of the markers shown in Tables 1A or 1 B or Tables 2A or 2B.

In all the afore aspects of the invention said holoclone or paraclone cell types may be characterised by any selected combination of the markers shown in Tables 1A, 1 B or Tables 2A, 2B, respectively. The invention will now be described by way of example only with reference to the following Figure, materials and methods.

FIGURE 1 : Panel A shows 3 OSCC cell lines plated at low density 3 days previously. Each shows a spectrum of colony morphologies from holoclones (stem, to the left) and paraclones (late amplifying, to the right).

FIGURE 2: Panel B shows a population in which half the cells were transduced with Enhanced Green Fluorescent Protein (EGFP), a holoclone is EGFP positive, and a paraclone negative, a typical finding and suggesting that colonies develop as clones.

FIGURE 3: Panel C shows a typical plot from experiments relating cell adhesion to cell size. Tumour cells were plated onto plastic dishes and the supernatant with non-adherent cells twice sequentially removed and replated.

The early adherent cells are smaller than those that adhere later, a property similar to that of normal cells in vitro.

FIGURE 4: Panel D. Plating early adherent cells produces H holoclones (D1 ) that can be isolated with cloning rings and replated to produce initially more H holoclones (D2). However, if these are passaged further, M meroclones (D3) and P paraclones (D4) begin to appear.

FIGURE 5: Panel E. Cloned populations, passaged and stained with antibodies against molecules reported to be expressed at higher levels on stem cells (CK19, β 1 integrin, desmoplakin, EGF receptor, E-cadherin). In each

case, high expression is seen on compact holoclones. The reverse is true for CK14 which is expressed by differentiating cells in paraclones.

FIGURE 6: Immunofluorescent staining of colonies, (a) shows strong staining of a large holoclone (left) for cytokeratin 15 but little reactivity of a paraclone (right), (b) the same field viewed to display nuclear staining with Hoechst 33258 clearly illustrates the larger nuclei and looser packing of the cells forming paraclones. (c) shows staining of the MCF7 line for CD44; staining is restricted to a tight cluster of holoclone cells, (d) Hoechst staining of the same field shows the presence of large, spaced, unstained paraclone cells above and to the right, (e) shows strong reactivity of the borders of holoclone cells of the CA1 line stained for CD44 and lack of staining of paraclones (to the right), (f) shows staining of the C1 line for β1 integrin with stronger reactivity of the peripheries of a holoclone

(left). (g,h) show staining of paraclone colonies of normal oral keratinocytes (g) and the CA1 cell line (h) for cytokeratin 6. Both colonies lack staining of small "basal" cells but larger flattened cells over the centres of the colonies are similarly stained for this early differentiation marker.

FAD MEDIUM PREPARATION

(Used for cancer cell growth) Insulin preparation Dilute 100mg Insulin (whole bottle) with 10ml distilled water and lOOmicroi of glacial acid. This is a 10mg/ml stock solution. Prepare aliquots of stock solution and store at 4°C for 3-6 months. Add O .25ml stock solution per 500ml of medium. Dissolve one aliquot of stock solution in 500ml of medium. Final cone

5microl/ml.

Solution A:

375ml Dulbecco's MEM high glucose 125ml Ham's F-12

In a glass beaker weigh 9.1 mg adenine. Add 4ml of DMEM/F12 to dissolve. Then add back to 500ml DMEM F-12 mixture. This is now solution "A".

Working Solution:

450ml Solution A

50ml Foetal Bovine Serum

0.5ml Hydrocortisone stock solution

0.5ml Cholera toxin stock solution 0.5ml Epidermal Growth Factor

0.250ml "New" Insulin stock solution

5ml Antibiotic/Antimycotic

Mix well and filter sterilise. Final Concentrations of Components are: DMEM 67.5% Ham's F12 22.5% FBS 10.0% Hydrocortisone 400ng/ml Cholera toxin 10-10M EGF 10ng/ml

Adenine 0.089mM Insulin 5microg/ml

Preparation of Stock Solutions: 1. Hydrocortisone 1000 x stock solution Sigma - H0888

4mg hydrocortisone in 10ml solution A = 400μg/ml

(Add 1 ml ETOH to dissolve hydrocortisone to total 10mls of solution A) Filter sterilise and store aliquots at 4°C

2. Cholera toxin 1000 x stock solution Sigma - C3012 1 mg cholera toxin in 119mls solution A Filter sterilise and store aliquots at -20°C

3. Epidermal Growth Factor 1000 x stock solution Sigma - E4127

100μg EGF in 10ml solution A Filter sterilise and store aliquots at -20°C

Colony morphologies in malignant cell lines. Fifteen different cell lines derived from oral squamous cell carcinoma (SCC) were plated at clonal densities and the morphology of the colonies formed was recorded by phase contrast photo-microscopy (please see Figure 1 ). Differences in cellular morphologies were apparent between individual lines but, superimposed on such differences, each line developed a similar range of colony patterns. At one extreme were round colonies with a smooth outline consisting of small packed cells equivalent to holoclones please see the first image [left-hand side] of each cell line in Figure 1). At the other extreme were irregular colonies consisting of spaced flattened or spindle shaped cells (termed Type 3 colonies, equivalent to paraclones, please see the last image [right-hand side] of the three cell lines in Figure 1). A spectrum of intermediate forms was also present (termed Type 2 colonies, equivalent to meroclones, please see images 2 and 3 of each cell line in Figure 1). These colony morphologies were identifiable at the early stages of colony formation and were retained as the colonies increased in size. Thus, a range of colonies corresponding approximately to those present in cultures of normal epithelia could be consistently found within malignant cell lines.

Regeneration of colony morphologies. Cellular heterogeneity is a common feature of malignant cell lines and has typically been attributed to clonal evolution associated with genetic instability and high mutation rates. Possibly, therefore, the apparent sequence of colony patterns identified could have resulted simply from an arbitrary arrangement of genetically different clones. It was anticipated that if variable coiony morphologies develop because of genetic changes then a particular colony morphology would initially persist after cloning, new colony morphologies would be relatively slow to develop and would not necessarily correspond to those previously present. On the other hand, colony heterogeneities resulting from a stem cell hierarchy would be expected to arise more rapidly and consistently recapitulate the same variety of patterns of colony morphology.

To test this, cell lines were plated at very low numbers, colonies grown up and then classified by morphology. Type 1 colonies, thought to correspond to holoclones, were isolated with cloning rings and trypsinized to release cells that were again re-plated at clonal densities. Such cultures formed mainly Type 1 colonies but some Type 2 and 3 colonies were formed with morphologies corresponding to those present prior to cloning. On further passage, the proportion of Type 2 and 3 colonies increased. A feature of stem cell related patterns in normal epithelial cells is directionality, i.e., holoclones to meroclones to paraclones. We therefore examined the range of colony morphologies that developed after isolation and plating of cells isolated from Type 2 and Type 3 colonies. It was shown that Type 2 colonies formed both Type 2 and Type 3 colonies, whereas Type 3 colonies formed only Type 3 colonies or failed to grow.

Relationships between cell size, cell adhesion and colony formation.

When cells are isolated from normal epithelia, the cells that form holoclones are typically smaller than the majority of cells and more rapidly adherent to a range of substrates. To see whether a similar relationship held for malignant cell lines, cells from five lines of oral SCC were plated in paired uncoated T75 flasks until 10-20% of the cells were adherent. Then medium and non-adherent cells were removed and added to new flasks for one hour. The supernatant and non- adherent cells were again removed and re-plated for four hours. At the end of each plating period the flasks were washed well to remove any non-adherent cells. The cells in one of each pair of flasks were incubated in medium to observe the colony morphologies that developed. The cells in the other flasks were removed, prepared as cell smears, and digital images captured for analysis of cell sizes using Scion Software. It was found that the rapidly adherent cells were smaller than the more slowly adherent cells. Type 1 colonies had been found to be made up of closely packed round cells and Type 3 colonies of cells that were spaced, larger, and either flattened or elongated. Classified by these criteria, the colonies developed from rapidly adherent cells contained a greater proportion of Type 1 colonies than those developing from more slowly adherent cells. The relationship between cell size, cell adhesion and colony formation for malignant cells therefore paralleled that of normal keratinocytes.

Immunofluorescent analysis of patterns of macromolecular expression associated with differing colony morphologies. Several malignant cell lines were plated at low cell densities, grown to form colonies, fixed in acetone/methanol or formalin and stained by standard methods using antibodies against a wide range of macromolecules reported to be expressed in relation to epithelial differentiation. Individual cell lines showed some differences in their particular patterns of staining but different molecular expression profiles between holoclone and paraclone colonies were found for all cell lines. Affymetrix gene expression data. We have performed a Affymetrix gene chip experiment in order to test whether differential gene expression can be detected between populations derived from holoclones and paraclones. Five cloned OSCC cell lines were plated at low density, grown for 10 days and for each line a holoclone and a paraclone identified, isolated, trypsinised, and passaged into T75 flasks. As the cells proliferated to fill the flask, maintenance within the majority of the population of the cellular features of the holoclones and paraclones of origin was confirmed by microscopy. As cells approached confluence, RNA was extracted and used for Affymetric analysis on gene chips.

Examples of genes differentially expressed in holoclones and in paraclones are listed in Tables 1 and 2.

Patterns of colony morphology in malignant cell lines derived from prostate and breast carcinoma. We have investigated the degree to which the stem and amplification patterns identified in oral SCC are also present in other major tumours. We have grown up and examined two prostate cell lines and four breast cell lines (two well differentiated, two not). Patterns of colony morphology paralleling the holoclone, meraclone and paraclone present in normal and oral SCC lines were apparent. Staining indicates differential expression of macromolecules between holoclone and paraclone colonies for both types of tumours. Methods for RNA extraction: General Culture Conditions Cells should ideally be cultured in DMEM + 10% FCS or other supplemented medium. They should be trypsinised and harvested and then passaged at a cell count sufficient to give approximately 2 x 105 - 1 x 106 cells after 72 hours. The cells should be plated out in numbers such that they are still growing and are not contact-inhibited after 72 hours. This may require a number of dishes to be set up. The cells should then be left for 72 hours, ideally over a weekend, when no one will disturb the incubator.

General Reagents:

PBS

TRIzol

Phase lock gel (heavy for TRIzol) Chloroform

Isopropanol

Glycogen (5mg/ml)

Ethanol

3M NaOAc

Extraction method

Wash cells with PBS

Add 1.5ml TRIzol to each flask (10cm dish) and leave to lyse for 5-30mins

Scrape off cells using cell scraper and put 750μl into each eppendorf Add 175μl chloroform to each and shake

Spin down phase lock gel in microfuge 13K, 25sec Add chloroform/TRIzol mix to phase lock gel (DO NOT SHAKE) and centrifuge13K, 2 mins Transfer upper colourless aqueous layer into new eppendorfs

Precipitate RNA by mixing with 500μl isopropanol and 1 μl glycogen. Leave on

bench for 10 mins+

Centrifuge 12,000g, 10 mins, 40°C

Remove supernatant

Wash pellet with 500μl of 75% EtOH

Centrifuge 7,500g, 5mins, 40°C

Remove supernatant and air-dry pellet for 5-10 mins

Dissolve pellet in 200μl H2O, 650C, 10 mins (use heat block)

Run 1μl on Agilent RNA chip

Dilute 1 μl in 69μl H2O and OD on Spec

Add : 1/10th vol of 3M NaOAc (20μl)

2.5 x vol of 100%, EtOH (500μl)

1 μl of glycogen

EtOH precipitate -200C, o/n Centrifuge 12,000g, 20 mins, 40°C

Remove supernatant and wash with 500μl of 80% EtOH

Centrifuge 12,000g, 10 mins, 40°C

Remove supernatant and wash with 500μl of 80% EtOH Centrifuge 12,000g 10 mins, 40°C Remove supernatant and resuspend in required volume of H2O, 65°C,10 mins

Dilute 1μl in 69μl of H2O and OD on spec

Dilute RNA to < 500ng/μl and run 1 μl on an Agilent RNA chip.

Methods used to identify genes that are expressed at a higher level in holoclone or paraclone colonies, or cells thereof

The levels of gene expression in holoclones and paraclones were determined by three methods:

1. Immunofluorescent staining of acetone/methanol fixed colonies with antibodies specific for gene products of interest, using our previously published methods (Gao, Z., Mackenzie, I.C. (1996) Influence of retinoic acid on the expression of cytokeratins, vimentin and ICAM-1 in human gingival epithelia in vitro. J. Perio Res. 31 :81-89).

2. Affymetric analysis of RNA extracted from holoclone or paraclone cells as described above. Analysis of the data so obtained was undertaken by a) elimination of genes or gene regions not uniformly showing greater or lesser holoclone expression values for four of the five tumour lines. Levels of expression were assessed to be significantly different if holoclone levels were determined to be either 50% higher or lower than paraclones levels, b) elimination of genes or gene regions that, as assessed by Affymetric statistics, did not show true expression associated with the higher of the pair of values, c) examining expression patterns, using manual searches or commercial programmes, for genes previously identified to be related to stem cell behaviour. 3. Q-PCR of RNA extracted from holoclone or paraclone cells to confirm in detail expression patterns identified by Affymetric analysis.

Table 1A. Genes expressed at higher levels in holoclones.

adaptor protein with pleckstrin homology. gb:NM_020979.1 /DEF=Homo sapiens adaptor protein with pleckstrin homology and src homology 2 domains (APS), /DB_XREF=gi: 10280625 /UG=Hs.105052 gb:AB000520.1

aldehyde dehydrogenase. gb:NM_000695.2 /DEF=Homo sapiens aldehyde dehydrogenase 3 family, member B2 (ALDH3B2), aldehyde dehydrogenase 3B2 /DB_XREF=gi:4580414

ATP-binding cassette, sub-family A (ABC1). gb:NM_019112.1 /DEF=Homo sapiens ATP-binding cassette, sub-family A (ABC1 ), member 7 (ABCA7), /DB_XREF=gi:9506364 /UG=Hs.134514 /FL=gb:AF328787.1 gb:AF250238.1

ABC transporter (WHITE2). gb:NM_022169.1 /DEF=Homo sapiens putative

ABC transporter (WHITE2), /DB_XREF=gi:11545923 /UG=Hs.126378

ATP-binding cassette, sub-family B (ABCB3). gb:NM_000544.2 /DEF=Homo sapiens (MDRTAP), member 3, transcript variant 1 , /DB_XREF=gi:9961245 /UG=Hs.502 /FL=gb:BC002751.1 gb:M74447.1 gb:Z22935.1 gb:Z22936.1

ATP-binding cassette transporter 1 (ABCA1). gb:AF285167.1 /DEF=Homo sapiens ATP-binding cassette transporter 1 (ABCA1 ) DB_XREF=gi:9755158 /UG=Hs.211562 /FL=gb:AF165281.1 gb:NM_005502.1 ATP-binding cassette (ABCG1). gb:NM_004915.2 /DEF=Homo sapiens ATP- binding cassette, sub-family G (WHITE), member 1 (ABCG1), transcript variant 1 , /DB_XREF=gi:8051574 /UG=Hs.10237,

ATP-binding cassette, sub-family C (CFTRMRP). gb:NM_020037.1 /DEF=Homo sapiens ATP-binding cassette, sub-family C (CFTRMRP), member 3 (ABCC3), transcript variant MRP3A. /DB_XREF=gi:9955971 /UG=Hs.90786 /FL=gb:AF085691.1

ATP-binding cassette, sub-family D (ALD). gb:NM_000033.2 /DEF=Homo sapiens ATP-binding cassette, sub-family D (ALD), member 1 (ABCD1 ), /DB_XREF=gi:7262392 /UG=Hs.159546

binder of Rho GTPases 4 (CEP4). gb:NM_012121.2 /DEF=Homo sapiens

Cdc42 effector protein 4; binder of Rho GTPases 4 (CEP4), /DB_XREF=gi: 13786126 /UG=Hs.3903 Cdc42 effector protein 4; /FL=gb:AB042237.1 gb:AF099664.1

bisphosphoglycerate mutase (BPGM). gb:NM_001724.1 /DEF=Homo sapiens

2,3-bisphosphoglycerate mutase (BPGM), /DB_XREF=gi:4502444 /UG=Hs.198365 cadherin. gb:NM_001407.1 /DEF=Homo sapiens cadherin, EGF LAG seven- pass G-type receptor 3, flamingo (Drosophila) homolog (CELSR3), /DB_XREF=gi: 13325065 /UG=Hs. gb:AF231023.1

cadherin-E (UVO). gb:L08599.1 /DEF=Human uvomorulin (E-cadherin) (UVO /DB_XREF=gi:340184 /UG=Hs.194657 cadherin 1 , type 1 , E-cadherin (epithelial) /FL= gb:NM_004360.1

cadherin-K. gb:BC000019.1 /DEF=Homo sapiens, Similar to cadherin 6, type 2, K-cadherin (fetal kidney), clone MGC:1470, /DB_XREF=gi: 12652558

/UG=Hs.32963 /FL=gb:BC000019.1

CCAAT enhancer binding protein (CEBP). gb:NM_004364.1 /DEF=Homo sapiens CCAATenhancer binding protein (CEBP), alpha (CEBPA), /DB_XREF=gi:4757971 /UG=Hs.76171

CDC42 GAP-related protein. Consensus includes gb:U62794.1 /DEF=Human CDC42 GAP-related protein mRNA, partial eds. /FEA=mRNA /PROD=CDC42 GAP-related protein /DB_XREF=gi:2326170 /UG=Hs.138860 Rho GTPase activating protein 1

chloride channel 2. gb:NM_006536.2 /DEF=Homo sapiens chloride channel, calcium activated, family member 2 (CLCA2), /DB_XREF=gi: 12025665 /UG=Hs.241551 gb:AB026833.1 gb:AF043977.1 gb:AF127980.1 chloride channel 3. Consensus includes gb:AI760629 /FEA=EST /DB_XREF=gi:5176296 /DB_XREF=est:wi66e06.x1 /CLONE=IMAGE:2398306 /UG=Hs.174139 /FL=gb:AF029346.1 gb:NM_001829.1 gb:AF172729.1

cytochrome P450. gb:NM_000896.1 /DEF=Homo sapiens cytochrome P450, subfamily IVF, polypeptide 3 (leukotriene B4 omega hydroxylase) (CYP4F3) /DB_XREF=gi:4503240 /UG=Hs.106242 /FL=gb:AB002454.1 gb:D12620.1

cytochrome P-450mp. Consensus includes gb:M15331.1 /DEF=Human liver cytochrome P-450 S-mephenytoin 4-hydroxylase (P-450mp). /FEA=mRNA /GEN=CYP2C /DB_XREF=gi:181361 /UG=Hs.167529 cytochrome polypeptide 9.

cyclin-dependent kinase inhibitor p12. gb:AF115544.1 /DEF=Homo sapiens cyclin-dependent kinase inhibitor p12 (p16INK4a) mRNA, alternatively spliced form, complete eds. /FEA=mRNA /GEN=p16INK4a /PROD= /DB_XREF=gi:4206166 /UG=Hs.1174 (melanoma, p16, inhibits CDK4)

death -associated protein kinase 1 (DAPK1 ). gb:NM_004938.1 /DEF=Homo sapiens death-associated protein kinase 1 (DAPK1 ), mRNA. /FEA=mRNA /GEN=DAPK1 /PROD= /DB_XREF=gi:4826683 /UG=Hs.153924 dual specificity phosphatase 4 (DUSP4), gb:NM_001394.2 /DEF=Homo sapiens /DB_XREF=gi:12707552 /UG=Hs.2359 4 /FL=gb:U48807.1 gb:BC002671.1 gb:U21108.1

dual specificity phosphatase 6. gb:BC003143.1 /DEF=Homo sapiens, dual specificity phosphatase 6, clone MGC:3789, /FL=gb:BC003562.1 gb:BC003143.1 gb:BC005047.1 gb:AB013382.1 gb:NM_001946.1

dual-specificity tyrosine-(Y)-phosphorylation regulated kinase. Consensus includes gb:A1192838 /FEA=EST /DB_XREF=gi:3744047

/DB_XREF=est:qe63c04.x1 /CLONE=IMAGE: 1743654 /UG=Hs.173135 /FL=gb:NM_006482.1

EGF LAG seven-pass G-type receptor 2, flamingo homolog (CELSR2), gb:NM_001408.1 /DEF=Homo sapiens cadherin, /DB_XREF=gi: 13325063 /UG=Hs.57652

epithelial membrane protein 1. Consensus includes gb:BF445047 /FEA=EST /DB_XREF=gi:11510185 /DB_XREF=est:nad20g10.x1

/CLONE=IMAGE:3366330 /UG=Hs.79368

erythroblastic leukemia viral oncogene homolog 3. /DB_XREF=gi:4503596 /UG=Hs.199067 v-erb-b2 /FL=gb:M29366.1 gb:M34309.1 gb:NM_001982.1 erythrocyte membrane protein band 4.1 -like. Consensus includes gb:AA912711 /FEA=EST /DB_XREF=gi:3052103 /DB_XREF=est:ol30f08.s1 /CLONE=ΪMAGE:1524999 /UG=Hs.26395 erythrocyte membrane protein band 4.1-like 1

forkhead box C1. Consensus includes gb:AU 145890 /FEA=EST /DB_XREF=gi:11007411 /DB_XREF=est:AU 145890 /CLONE=HEMBA1006158 /UG=Hs.284186 f /FL=gb:NM_001453.1

frizzled homolog 3 (FZD3). gb:NM_017412.1 /DEF=Homo sapiens frizzled (Drosophila) homolog 3 (FZD3), /DB_XREF=gi:8393377 /UG=Hs.40735 /FL=gb:AY005130.3 gb:AB039723.1

frizzled homolog 10 (FZD10). gb:NM_007197.1 /DEF=Homo sapiens frizzled

(Drosophila) homolog 10 (FZD10), /DB_XREF=gi:6005761 /UG=Hs.31664 /FL=gb:AB027464.1

GTP cyclohydrolase I feedback regulatory protein (GCHFR). gb:NM_005258.2 /DEF=Homo sapiens, /DB_XREF=gi:6382072 /UG=Hs.

HMG-box transcription factor TCF-3 (TCF-3). gb:NM_031283.1 /DEF=Homo sapiens DB_XREF=gi: 13786122 / HMG box factor SOX-13. Cluster Incl. AF083105:Homo sapiens /gi=3982828 /ug=Hs.201671 /len=3583

hurpin. Consensus includes gb:AJ001696.2 /DEF=Homo sapiens mRNA for hurpin, clone R7-1.1. /FEA=mRNA /GEN=PI13 /PROD=hurpin /DB_XREF=gi:6018505 /UG=Hs.241407 serine (or cysteine) proteinase inhibitor, clade B (ovalbumin), member 13 /FL=gb:AF169949.1 gb:NM_012397.1

hydroxyprostaglandin dehydrogenase. gb:NM_000860.1 /DEF=Homo sapiens hydroxyprostaglandin dehydrogenase 15-(NAD) (HPGD),

/FL=gb:L76465.1

insulin-like growth factor-binding protein 3. gb:M31159.1 /DEF=Human growth hormone-dependent insulin-like growth factor-binding protein 3 mRNA, complete eds. /FEA=mRNA /GEN=IGFBP1 /DB_XREF=gi:183115

/UG=Hs.77326 FL=gb:BC000013.

insulin-like growth factor binding protein 3. Consensus includes gb:BF340228 /FEA=EST /DB_XREF=gi:11286690 /DB_XREF=est:602036816F1 /CLONE=IMAGE:4185050 /UG=Hs.77326 insulinlike growth factor binding protein 3 /FL=gb:NM_000598.1

insulin-like growth factor binding protein 6 (IGFBP6). gb:NM_002178.1 /DEF=Homo sapiens insulin-like growth factor binding protein 6 (IGFBP6), /DB_XREF=gi:11321592 /UG=Hs.274313 FL= gb:BC003507.1 gb:BC005007.1 gb:M62402.1

integrin alpha 10 subunit (ITGA10). gb:AF112345.1 /DEF=Homo sapiens /DB_XREF=gi:6650627 /UG=Hs.158237 integrin, alpha 10 /FL=gb:AF074015.1 gb:NM_003637.2

interleukin 2 receptor, beta (IL2RB). gb:NM_000878.1 /DEF=Homo /DB_XREF=gi:4504664 /UG=Hs.75596 /FL=gb:M26062.1

kallikrein 10. gb:BC002710.1 /DEF=Homo sapiens, kallikrein 10, clone MGC:3667/DB_XREF=gi: 12803744 /UG=Hs.69423

keratin 13. gb:NM_002274.1 /DEF=Homo sapiens keratin 13 (KRT13),

/DB_XREF=gi:4504910 /UG=Hs.74070

keratin 15 (KRT15). gb:NM_002275.1 /DEF=Homo sapiens keratin 15 (KRT15), /DB_XREF=gi:4504914 /UG=Hs.80342 FL=gb:BC002641.1

mesenchymal stem cell protein. gb:NM_016647.1 /DEF=Homo sapiens mesenchymal stem cell protein DSCD75 (LOC51337), /DB_XREF=gi:7706199 /UG=Hs.25237 /FL=gb:BC001311.1 gb:AF242773.1 mitogen-activated protein kinase kinase kinase 5 (MAP3K5). gb:NM_005923.2 /DEF=Homo sapiens /DB_XREF=gi:6031181 /UG=Hs.151988 /FL=gb:U67156.1 gb:D84476.1

mucin 1(MUC1). gb:NM_002456.1 /DEF=Homo sapiens mucin 1 , transmembrane (MUC1 ), /DB_XREF=gi:4505282 /UG=Hs.89603

mucin short variant (MUC1). gb:AF348143.1 /DEF=Homo sapiens mucin short variant (MUC1 ) /DB_XREF=gi: 13560650

mucin (MUC3). Consensus includes gb:AF007194.1 /DEF=Homo sapiens mucin (MUC3) DB_XREF=gi:2853300 /UG=Hs.129782 mucin 3A, intestinal

mucin 4, tracheobronchial. Consensus includes gb:AJ242547.1 /DEF=Homo sapiens partial mRNA for sv7-MUC4 apomucin. /FEA=mRNA /GEN=MUC4

/PROD=sv7-MUC4 apomucin /DB_XREF=gi:7406614 /UG=Hs.198267

myc protein. Consensus includes gb:M19720 /DEF=Human L-myc protein gene, complete eds /FEA=mRNA_2 /DB_XREF=gi: 188906 /UG=Hs.92137 v- myc avian myelocytomatosis viral oncogene homolog 1 , lung carcinoma derived

nebulette (NEBL). gb:NM_006393.1 /DEF=Homo sapiens nebulette (NEBL). /FEA=mRNA /GEN=NEBL /PROD=nebulette /DB_XREF=gi:5453757 /UG=Hs.5025 neuropilin 2. Consensus includes gb:AA295257 /FEA=EST

/DB_XREF=gi: 1947819 /DB_XREF=est:EST100654 /UG=Hs.17778 neuropilin 2

/FL=gb:AF022860.1 gb:AF016098.1 gb:NM_003872.1

Notch homolog 3. gb:NM_000435.1 /DEF=Homo sapiens Notch (Drosophila) homolog 3 (NOTCH3), /DB_XREF=gi:4557798 /UG=Hs.8546 /FL=gb:U97669.1

pak5 protein. Consensus includes gb:AJ236915.1 /DEF=Homo sapiens. /DB_XREF=gi:11691854 /UG=Hs.21420 p21 -activated protein kinase 6

/FL=gb:AF276893.1 gb:NM_020168.1

phosphoglucomutase 1 (PGM1. gb:NM_002633.1 /DEF=Homo sapiens /DB_XREF=gi:4505764 /UG=Hs.1869 phosphoglucomutase 1 /FL=gb:BC001756.1 gb:M83088.1

Pirin (PIR). gb:NM_003662.1 /DEF=Homo sapiens /DB_XREF=gi:4505822 /UG=Hs.279663

PML-1. gb:M79462.1 /DEF=Human PML-1 mRNA, complete CDS. /FEA=mRNA

/GEN=PML /DB_XREF=gi:190116 /UG=Hs.89633 promyelocytic leukemia podocalyxin-like (PODXL). gb:NM_005397.1 /DEF=Homo sapiens podocalyxin-like (PODXL), /DB_XREF=gi:4885556 /UG=Hs.16426 podocalyxin- like /FL=gb:U97519.1

pregnancy-specific beta-1 -glycoprotein. gb:M34715.1 /DEF=

/DB_XREF=gi: 190590 /UG=Hs.173609 gb:J04539.1 gb:M33663.1

prostate stem cell antigen (PSCA). gb:NM_005672.1 /DEF=Homo sapiens prostate stem cell antigen (PSCA), /DB_XREF=gi:5031994 /UG=Hs.20166 /FL=gb:AF043498.1

proteoglycan 4, (megakaryocyte stimulating factor, articular superficial zone protein) (PRG4), DB_XREF=gi:5031924 /UG=Hs.218791 proteoglycan 4, /FL=gb:U70136.1 gb:NM_005807.1

retinol dehydrogenase homolog (RDHL). gb:NM_005771.1 /DEF=Homo sapiens retinol dehydrogenase homolog (RDHL), mRNA. /FEA= /DB_XREF=gi:5032034 /UG=Hs.179608 /FL=gb:AF067174.1

retinoic acid receptor responder (tazarotene induced) 3 (RARRES3). gb:NM_004585.2 /DEF=Homo sapiens /DB_XREF=gi:8051633 /UG=Hs.17466 /FL=gb:AF060228.1 gb:AF092922.1 gb:AB030815.1 retinoic acid receptor gamma. gb:M57707.1 /DEF=Human retinoic acid receptor gamma mRNA, complete eds. DB_XREF=gi:190867 /UG=Hs.1497 /FL=gb:M24857.1 gb:M38258.1 gb:NM_000966.1

ribosomal protein L3. Consensus includes gb:AW517464 /FEA=EST

/DB_XREF=gi:7155546 /DB_XREF=est:xp94g06.x1 /CLONE=IMAGE:2748058 /UG=Hs.119598

semaphorin III. Cluster Incl. U38276:Human semaphorin III family homolog mRNA, complete eds /cds=(315,2576) /gb=U38276 /gi=1061350 /ug=Hs.32981

/len=3534

sialyltransferase (STHM). gb:NM_006456.1 /DEF=Homo sapiens DB_XREF=gi:5454091 /UG=Hs.288215 sialyltransferase gb:U14550.1

signal transducer and activator of transcription 4 (STAT4). gb:NM_003151.1 /DEF=Homo sapiens /DB_XREF=gi:4507254 /UG=Hs.80642 signal transducer and activator of transcription 4 /FL=gb:L78440.1

toll-like receptor 3 (TLR3). gb:NM_003265.1 /DEF=Homo sapiens toll-like receptor 3 (TLR3), /DB_XREF=gi:4507530 /UG=Hs.29499 /FL=gb:U88879.1 tumor necrosis factor (TNFSF10). gb:NM_003810.1 /DEF=Homo sapiens tumor necrosis factor (ligand) superfamily, member 10 (TNFSF10), DB_XREF=gi:4507592 /UG=Hs. /FL=gb:U37518.1 gb:U57059.1

tyrosine phosphatase, receptor type, R (PTPRR). gb:NM_002849.1

/DEF=Homo sapiens protein tyrosine phosphatase, receptor type, R (PTPRR), R /DB_XREF=gi:4506324 /UG=Hs. /FL=gb:U77916.1 gb:U42361.1 gb:D64053.1

transcription factor 3, E2F. Consensus includes gb:AI640363 /FEA=EST /DB_XREF=gi:4703472 /DB_XREF=est:wa17d03.x1 /CLONE=IMAGE:2298341

/UG=Hs.1189 E2F transcription factor 3 /FL=gb:NM_001949.2

tumor necrosis factor, alpha-induced protein 2 (TNFAIP2). gb:NM_006291.1 /DEF=Homo sapiens /DB_XREF=gi:5454133 /UG=Hs.101382 tumor necrosis factor, alpha-induced protein 2 /FL=gb:M92357.1

vav 1 oncogene (VAV1). gb:NM_005428.2 /DEF=Homo sapiens DB_XREF=gi:7108366 /UG=Hs.116237 vav 1 oncogene

vav 3 oncogene (VAV3). gb:NM_006113.2 /DEF=Homo sapiens

/DB_XREF=gi:7262390 /UG=Hs.267659 vav 3 oncogene /FL=gb:AF067817.1 gb:AF118887.1 v-erb-b2 avian erythroblastic leukemia viral oncogene homolog 3 (ERBB3). gb:NM_001982.1 /DEF=Homo sapiens

zinc finger protein, X-lϊnked (ZFX). gb:NM_003410.1 /DEF=Homo sapiens /DB_XREF=gi:4507964 /UG=Hs.2074 zinc finger protein, X-linked /FL=gb:M30608.1

Table 1B. Gene products acting as markers of holoclones.

CD133, CD44, e-cadherin, alpha 6 and beta 1 integrins, alpha and beta catenins, epidermal growth factor receptor and proteoglycan 4.

Table 2A. Genes expressed at higher levels in paraclones

A kinase (PRKA) (gravin). gb:AB003476.1 /DEF=Homo sapiens /DB_XREF=gi:2081606 /UG=Hs.788 12

alpha(1,3)-fucosyltransferase IV (FUTIV). Consensus includes gb:AF305083.1 /DEF=Homo sapiens alpha(1 ,3)-fucosyltransferase IV (FUTIV) gene, 3 UTR. /FEA=mRNA /DB_XREF=gi:11096240 /UG=Hs.2173 /FL=gb:M58596.1 gb:M58597.1 gb:NM_002033.1

autophagy 12 (APG12L). gb:NM_004707.1 /DEF=Homo sapiens Apg12 (autophagy 12, S. cerevisiae)-like (APG12L), /DB_XREF=gi:4757757 /UG=Hs.264482 /FL=gb:AB017507.1

avian myelocytomatosis viral oncogene. gb:NM_002467.1 /DEF=Homo sapiens v-myc avian myelocytomatosis viral oncogene homolog (MYC), /DB_XREF=gi:12962934 /UG=Hs.79070 /FL=gb:BC000141.1 gb:BC000917.2

connective tissue growth factor. gb:M92934.1 /DEF=Human /DB_XREF=gi: 180923 /UG=Hs.75511 connective tissue growth factor gb:NM_001901.1 cysteine-rich, angiogenic inducer, 61 (CYR61). gb:NM_001554.1 /DEF=Homo sapiens /DB_XREF=gi:4504612 /UG=Hs.8867 /FL=gb:BC001271.1 gb:U62015.1 gb:AF003594.1 gb:AF031385.1

cysteine-rich, angiogenic inducer, 61. gb:AF003114.1 /DEF=Homo sapiens

/DB_XREF=gi:6649848 /UG=Hs.8867 /

dickkopf homolog 1 (DKK1). gb:NM_012242.1 /DEF=Homo sapiens dickkopf (Xenopus laevis) homolog 1 /DB_XREF=gi:7110718 /UG=Hs.40499 (DKK1 ), /FL=gb:AF127563.1 gb:AF177394.

Dickkopf gene 3. gb:NM_013253.1 /DEF=Homo sapiens dickkopf (Xenopus laevis) homolog 3 (DKK3), /DB_XREF=gi:7019362 /UG=Hs.4909 /FL=gb:AF177396.1 gb:AB033421.1 gb:AB034203.1

DNA-binding zinc finger(GBF). gb:AB017493.1 /DEF=Homo /DB_XREF=gi:3582142 /UG=Hs.285313 core promoter element binding protein /FL=gb:BC000311.1 gb:BC004301.1 gb:AF001461.1 gb:NM_001300.2

epithelial membrane protein 3 (EMP3). gb:NM_001425.1 /DEF=Homo sapiens

PROD=epithelial membrane protein 3 /DB_XREF=gi:4503562 /UG=Hs.9999 /FL=gb:U52101.1 gb:U87947.1 forkhead box D1 (FOXD1). gb:NM_004472.1 /DEF=Homo sapiens forkhead box D1 (FOXD1 ), /DB_XREF=gi:4758391 /UG=Hs.96028 /FL=gb:U59832.1

forkhead box F2 (FOXF2). gb:NM_001452.1 /DEF=Homo sapiens forkhead box F2 (FOXF2), /DB_XREF=gi:4557594 /UG=Hs.44481 forkhead box F2 /FL=gb:U13220.1

goliath protein. gb:NM_018434.1 /DEF=Homo sapiens (LOC55819) GEN=LOC55819 /DB_XREF=gi: 10092650 /UG=Hs.102737

growth arrest and DNA-damage-inducible 34 (GADD34). gb:NM_014330.2 /DEF=Homo sapiens DB_XREF=gi:9790902 /UG=Hs.76556 FL=gb:BC003067.1 gb:U83981.1

GRO3 oncogene (GRO3). gb:NM_002090.1 /DEF=Homo sapiens

/DB_XREF=gi:4504156 /UG=Hs.89690 /FL=gb:M36821.1

hepatoma-derived growth factor. Consensus includes gb:AK001280.1 /DEF=Homo sapiens cDNA FLJ10418 fis, clone NT2RP1000130, moderately similar to HEPATOMA-DERIVED GROWTH FACTOR. /FEA=mRNA

/DB_XREF=gi:7022435 /UG=Hs.127842 CGI-142 /FL=gb:AF151900.1 gb:AB029156.1 gb:NM_016073.1 HRP-3. gb:AB029156.1 /DEF=Homo sapiens /DB_XREF=gi:6855467 /UG=Hs.127842 CGI-142 /FL=gb:AF151900.1 gb:NM_016073.1

HSPC007 protein. gb:NM_014018.1 /DEF=Homo sapiens /DB_XREF=gi:7661729 /UG=Hs.55097 HSPC007 protein /FL=gb:AF070663.

inhibin, beta A (activin A, activin AB alpha polypeptide) (INHBA). gb:NM_002192.1 /DEF=Homo sapiens FEA=mRNA /GEN=INHBA /PROD=inhibin beta A subunit precursor /DB_XREF=gi:4504698 /UG=Hs. FL=gb:J03634.1

inhibin ovarian beta-A. gb:M13436.1 /DEF=Human mRNA, complete eds. /FEA=mRNA /GEN=INHBA /DB_XREF=gi:186414 /UG=Hs.727 inhibin, beta A (activin A, activin AB alpha polypeptide) /FL=gb:M13436.1

insulin-like growth factor 1 receptor. Consensus includes gb:AI830698 /FEA=EST /DB_XREF=gi:5451454 /DB_XREF=est:wj52f06.x1 /CLONE=IMAGE:2406467 /UG=Hs.239176 /FL=gb:NM_000875.2

insulin-like growth factor binding protein 7. gb:NM_001553.1 /DEF=Homo sapiens DB_XREF=gi:4504618 /UG=Hs.119206 FL=gb:L19182.1

integrin, beta 1. Consensus includes gb:BG500301 /FEA=EST /DB_XREF=gi:13461818 /DB_XREF=est:602546969F1 /CLONE=IMAGE:4669168 /UG=Hs.287797 Homo sapiens integrin, beta 1 (fibronectin receptor, beta polypeptide, antigen CD29 includes MDF2, MSK12) (ITGB1 ), Consensus includes gb:AA215854 /FEA=EST /DB_XREF=gi:1815801 /DB_XREF=est:csh0013.seq.F /UG=Hs.

integrin, alpha E. gb:NM_002208.3 /DEF=Homo sapiens integrin, alpha E (antigen CD103, human mucosal lymphocyte antigen 1 ; alpha polypeptide) (ITGAE), /DB_XREF=gi:6007850 /UG=Hs.851 /FL=gb:L25851.2 interleukin 1-alpha (IL1A). gb:M15329.1 /DB_XREF=gi: 186277 /UG=Hs.1722

interleukin-1 receptor accessory protein (IL1RAP). gb:AF167343.1 /DEF=Homo sapiens soluble /DB_XREF=gi:8050486 /UG=Hs.173880

interleukin 6 (interferon, beta 2) (IL6). gb:NM_000600.1 /DEF=Homo sapiens /DB_XREF=gi: 10834983 /UG=Hs.93913 gb:M14584.1 gb:M18403.1 gb:M29150.1 gb:M54894.1

interleukin 8 (IL8). gb:NM_000584.1 /DEF=Homo sapiens DB_XREF=gi: 10834977 /UG=Hs.624 gb:M17017.1 gb:M26383.1

keratin 6A. Consensus includes gb:J00269.1 /DEF=Human 56k cytoskeletal type II keratin /FL=gb:NM_005554.1 gb:L42611.1 gb:L42612.1 keratin 6A. Consensus includes gb:AI831452 /FEA=EST /DB_XREF=gi:5452123 /DB_XREF=est:wj49b03.x1 /CLONE=IMAGE:2406125 /UG=Hs.111758 keratin 6A

keratin 6B. Consensus includes gb:AL569511 /FEA=EST

/DB_XREF=gi:12924921 /CLONE=CS0DE014YF20 (3 prime) /UG=Hs.321223 /FL=gb:NM_005555.1

keratin 6 isoform K6f (KRT6F). gb:L42612.1 /DEF=Homo sapiens /DB_XREF=gi:908804 /UG=Hs.111758 /FL=gb:NM_005554.1 gb:L42611.1

laminin, gamma 2 (nicein). gb:NM_005562.1 /DEF=Homo sapiens laminin, gamma 2 (nicein (100kD), kalinin (105kD), BM600 (100kD), Herlitzjunctional epidermolysis bullosa)) (LAMC2), transcript variant 1 , /DB_XREF=gi:5031846 /UG=Hs.54451

low density lipoprotein receptor. Consensus includes gb:AI861942 /FEA=EST /DB_XREF=gi:5526049 /DB_XREF=est:td18b10.x1 /CLONE=IMAGE:2075995 /UG=Hs.213289 (familial hypercholesterolemia) /FL=gb:NM_000527.2

MAD1-like 1 (MAD1 L1), gb:NM_003550.1 /DEF=Homo sapiens MAD1 (mitotic arrest deficient, yeast, hoιmolog)-like 1 (MAD1 L1 ), /DB_XREF=gi:4505064 /UG=Hs.7345 /FL=gb:U33822.1 gb:AF123318.1 gb:AF083811.1 mesenchymal stem cell protein DSC92. gb:NM_016645.1 /DEF=Homo sapiens mesenchymal stem cell protein DSC92 (LOC51335), /DB_XREF=gi:7706195 /UG=Hs.323467 /FL=gb:AB029315.1 gb:AF242770.1

meiotic recombination protein. gb:AF309553.1 /DEF=Homo sapiens REC14 mRNA, complete eds. /FEA=mRNA /PROD= /DB_XREF=gi:11139241 /UG=Hs.296242 gb:NM_025234.1

migration stimulation factor FN70. Consensus includes gb:AJ276395.1 /DEF=Homo sapiens mRNA /DB_XREF=gi:12053816 /UG=Hs.321592

mitotic arrest deficient-like 1 (MAD1 L1), gb:NM_003550.1 /DEF=Homo sapiens MAD1 (mitotic arrest deficient, yeast, homolog)-like 1 (MAD1 L1 ), /DB_XREF=gi:4505064 /UG=Hs. /FL=gb:U33822.1 gb:AF123318.1 gb:AF083811.1

mortality factor 4 (MORF4). gb:NM_006792.2 /DEF=Homo sapiens /DB_XREF=gi:6996927 /UG=Hs.251396

nicotinamide N-methyltransferase (NNMT). gb:NM_006169.1 /DEF=Homo sapiens /DB_XREF=gi:5453789 /UG=Hs.76669 /FL=gb:BC000234.1 gb:U08021.1 NS1 -associated protein 1 (NSAP1). gb:NM_006372.1 /DEF=Homo sapiens /DB_XREF=gi:5453805 /UG=Hs.155489 NS1 -associated protein 1 /FL=gb:AF155568.1

nuclear localization signal deleted in velocardiofacial syndrome (NLVCF), gb:NM_003776.1 /DEF=Homo sapiens /DB_XREF=gi:4505402 /UG=Hs.19500 /FL=gb:AF034091.1

ornithine decarboxylase 1 (ODC1). gb:NM_002539.1 /DEF=Homo sapiens /DB_XREF=gi:4505488 /UG=Hs.75212 1 /FL=gb:M16650.1

phospholipase C, beta 4. Consensus includes gb:AL535113 /FEA=EST /DB_XREF=gi: 12798606 CLONE=CS0DF008YC23 (3 prime) /UG=Hs.283006 /FL=gb:NM_000933.1 gb:L41349.1

potential tumor suppressor (ST7). gb:NM_013437.1 /DEF=Homo sapiens /DB_XREF=gi:7305524 /UG=Hs. /FL=gb:AF166350.1

regulated in glioma (RIG). gb:NM_006394.1 /DEF=Homo sapiens regulated in glioma (RIG), /DB_XREF=gi:5454007 /UG=Hs.278503 gb:U32331.1

RNA binding motif, single stranded interacting protein 1 (RBMS1). gb:NM_016837.1 /DEF=Homo sapiens RNA binding motif, single stranded interacting protein 1 (RBMS1 ), transcript variant MSSP-3, /DB_XREF=gi:8400719 /UG=Hs.

SAM domain, SH3 domain and nuclear localisation signals, 1 (SAMSN1). gb:NM_022136.1 /DEF=Homo sapiens /DB_XREF=gi:11545870 /UG=Hs.24633 /FL=gb:AF222927.1

secreted frizzled-related protein 1 (SFRP1). gb:NM_003012.2 /DEF=Homo sapiens DB_XREF=gi:8400731 /UG=Hs.7306 /FL=gb:AF001900.1 gb:AF017987.1 gb:AF056087.1

serumglucocorticoid regulated kinase (SGK). gb:NM_005627.1 /DEF=Homo sapiens /DB_XREF=gi:5032090 /UG=Hs. /FL=gb:BC001263.1 gb:AF153609.1

syntaxin 3A. Consensus includes gb:BE966922 /FEA=EST

/DB_XREF=gi:11772839 /DB_XREF=est:601660942R1

/CLONE=IMAGE:3915610 /UG=Hs.82240 /FL=gb:AJ002076.1 gb:NM_004177.1 gb:U32315.1 thrombospondin 1 (THBS1). gb:NM_003246.1 /DEF=Homo sapiens /DB_XREF=gi:4507484 /UG=Hs.87409 thrombospondin 1 /FL=gb:NM_003246.1

Consensus includes gb:AI812030 thrombospondin 1. Consensus includes gb:AV726673 /FEA=EST /DB_XREF=gi: 10836094 /DB_XREF=est: /CLONE=HTCBGC12 /UG=Hs.87409 /FL=gb:NM_003246.1

thrombospondin 1. Consensus includes gb:BF055462 /FEA=EST

/DB_XREF=gi: 10809358 /DB_XREF=est:7j80e06.x1 /CLONE=IMAGE:3392770 /UG=Hs.87409 /FL=gb:NM_003246.1

transcription factor Dp-2 (E2F dimerization partner 2). Consensus includes gb:BG034328 /FEA=EST /DB_XREF=gi: 12427528

/DB_XREF=est:602302213F1 /CLONE=IMAGE:4403787 /UG=Hs.19131 /FL=gb:NM_006286.1 gb:U18422.1

transcription factor IIF, polypeptide 2. gb:BC001771.1 /DEF=Homo sapiens, general transcription factor IIF, polypeptide 2 (30kD subunit), clone MGC:1502,

/DB_XREF=gi: 12804688 /UG=Hs.58593 gb:NM_004128.1

Vimentin. Consensus includes gb:AI922599 /FEA=EST /DB_XREF=gi:5658563 /DB_XREF=est:wm90b11.x1 /CLONE=IMAGE:2443197 /UG=Hs.297753 /FL=gb:BC000163.2 gb:NM_003380.1 Table 2B. Gene products acting as markers of paraclones.

Vimentin, DKK1 , thrombospondin, cytokeratins numbers 6, 14, 16, 17, 20. epcam,.

Claims

CLAIMS 1. An isolated tumour comprising cells of any one of more of the following colony-types: (i) holoclone; (ii) meraclone; or (iii) paraclone wherein cells of said holoclone colony are characterised by the expression of any one or more of the markers shown in TablelA and 1 B and cells of the paraclone colony are characterised by expression of any one or more of the markers shown in Tables 2A and 2B.

2. An isolated tumour according to Claim 1 wherein cells of said holoclone colony express at least one of the following markers at a level at least 50%_> higher than the expression, if any, of the same marker in a paraclone colony: Adaptor protein with pleckstrin homology.

Aldehyde dehydrogenase;

ATP-binding cassette, sub-family A (ABC1 );

ABC transporter (WHITE2);

ATP-binding cassette, sub-family B (ABCB3); ATP-binding cassette transporter 1 (ABCA1 );

ATP-binding cassette (ABCG1 );

ATP-binding cassette, sub-family C (CFTRMRP);

ATP-binding cassette, sub-family D (ALD);

Binder of Rho GTPases 4 (CEP4); Bisphosphoglycerate mutase (BPGM); Cadherin; Cadherin-E (UVO); Cadherin-K; CCAAT enhancer binding protein (CEBP); CDC42 GAP-related protein;

Chloride channel 2;

Chloride channel 3;

Cytochrome P450; Cytochrome P-450mp;

Cyclin-dependent kinase inhibitor p12;

Death-associated protein kinase 1 (DAPK1 );

Dual specificity phosphatase 4 (DUSP4);

Dual specificity phosphatase 6; Dual-specificity tyrosine-(Y)-phosphorylation regulated kinase;

EGF LAG seven-pass G-type receptor 2, flamingo homolog (CELSR2);

Epithelial membrane protein 1 ;

Erythroblastic leukemia viral oncogene homolog 3;

Erythrocyte membrane protein band 4.1 -like; Forkhead box C1 ;

Frizzled homolog 3 (FZD3);

Frizzled homolog 10 (FZD10);

GTP cyclohydrolase I feedback regulatory protein (GCHFR);

HMG-box transcription factor TCF-3 (TCF-3); HMG box factor SOX-13; Hurpin; Hydroxyprostaglandin dehydrogenase; Insulin-like growth factor-binding protein 3 (HGF dependent); Insulin-like growth factor binding protein 3; Insulin-like growth factor binding protein 6 (IGFBP6); Integrin alpha 10 subunit (ITGA10);

Interleukin 2 receptor, beta (IL2RB);

Kallikrein 10; Keratin 13;

Keratin 15 (KRT15);

Mesenchymal stem cell protein;

Mitogen-activated protein kinase 5 (MAP3K5);

Mucin 1 (MUC1); Mucin short variant (MUC1 );

Mucin (MUC3);

Mucin 4, tracheobronchial;

Myc protein;

Nebulette (NEBL); Neuropilin 2;

Notch homolog 3;

Pak5 protein;

Phosphoglucomutase 1 (PGM1 );

Pirin (PIR); PML-1 ; Podocalyxin-like (PODXL); Pregnancy-specific beta-1 -glycoprotein; Prostate stem cell antigen (PSCA); Proteoglycan 4; Retinol dehydrogenase homolog (RDHL); Retinoic acid receptor responder (tazarotene induced) 3 (RARRES3);

Retinoic acid receptor gamma;

Ribosomal protein L3; Semaphorin III;

Sialyltransferase (STHM);

Signal transducer and activator of transcription 4 (STAT4); Toll-like receptor 3 (TLR3); Tumor necrosis factor (TNFSF10); Tyrosine phosphatase, receptor type, R (PTPRR);

Transcription factor 3, E2F;

Tumor necrosis factor, alpha-induced protein 2 (TNFAIP2); Vav 1 oncogene (VAV1 ); Vav 3 oncogene (VAV3); V-erb-b2 avian erythroblastic leukemia viral oncogene homolog 3 (ERBB3);

Zinc finger protein, X-linked (ZFX); DKK1 ; CD133; CD44; E-cadherin; Alpha 6 and beta 1 integrins; Alpha and beta catenins; Epidermal growth factor receptor; or Proteoglycan 4.

3. An isolator tumour according to Claim 1 wherein cells of said paraclone colony express at least one of the following markers at a level at least 50% higher than the expression, if any, of the same marker in a holoclone colony: A kinase (PRKA) (gravin);

Alpha(1 ,3)-fucosyltransferase IV (FUTIV);

Autophagy 12 (APG12L);

Avian myelocytomatosis viral oncogene;

Connective tissue growth factor; Cysteine-rich, angiogenic inducer, 61 (CYR61 );

Cysteine-rich, angiogenic inducer, 61 ;

Dickkopf homolog 1 (DKK1 );

Dickkopf gene 3;

DNA-binding zinc finger(GBF); Epithelial membrane protein 3 (EMP3);

Forkhead box D1 (FOXD1);

Forkhead box F2 (FOXF2);

Goliath protein;

Growth arrest and DNA-damage-inducible 34 (GADD34); GRO3 oncogene (GRO3); Hepatoma-derived growth factor; HRP-3; HSPC007 protein; Inhibin, beta A (activin A, activin AB alpha polypeptide) (INHBA; Inhibin ovarian beta-A; Insulin-like growth factor 1 receptor; Insulin-like growth factor binding protein 7; Integrin, beta 1 ; Integrin, alpha E;

Interleukin 1-alpha (IL1A);

Interleukin-1 receptor accessory protein (IL1 RAP);

Interleukin 6 (interferon, beta 2) (IL6);

Interleukin 8 (IL8); Keratin 6A Type II;

Keratin 6A;

Keratin 6B;

Keratin 6 isoform K6f (KRT6F);

Laminin, gamma 2 (nicein); Low density lipoprotein receptor;

MAD1-like 1 (MAD1 L1);

Mesenchymal stem cell protein DSC92;

Meiotic recombination protein;

Migration stimulation factor FN70; Mitotic arrest deficient-like 1 (MAD1L1 ); Mortality factor 4 (MORF4); Nicotinamide N-methyltransferase (NNMT); NS1 -associated protein 1 (NSAP1); Nuclear localization signal deleted in velocardiofacial syndrome (NLVCF); Ornithine decarboxylase 1 (ODC1 ); Phospholipase C, beta 4; Potential tumor suppressor (ST7); Regulated in glioma (RIG); RNA binding motif, single stranded interacting protein 1 (RBMS1 );

SAM domain, SH3 domain and nuclear localisation signals, 1 (SAMSN1 ); Secreted frizzled-related protein 1 (SFRP1); Serumglucocorticoid regulated kinase (SGK); Syntaxin 3A; Thrombospondin 1 (THBS1 );

Thrombospondin 1; Thrombospondin 1 ;

Transcription factor Dp-2 (E2F dimerization partner 2); Transcription factor IIF, polypeptide 2; Vimentin;

Cytokeratins numbers 6, 14, 16, 17, 20; or Epcam.

4. An isolated tumour according to Claim 1 or Claim 2, wherein cells of said holoclone colony express the following markers at a higher level compared with the expression of these markers in cells of said paraclone colonies: CD133, CD44, e-cadherein, alpha 6 and beta 1 integrins, alpha and beta catenins, epidermal growth factor receptor and proteoglycan 4.

5. An isolated tumour according to Claim 1 or Claim 3 wherein cells of said paraclone colony express the following markers at a higher level compared with the expression of these markers in cells of said holoclone colonies: Vimentin, DKK1 , DKK3, thrombospondin, cytokeratins number 6, 14, 16, 17 and 20.

6. An isolated tumour according to any preceding claim wherein said tumour is derived from breast, prostate, lung, gastro-intestinal tract, head or neck.

7. An isolated tumour according to any preceding claim wherein said tumour is of human origin.

8. An isolated tumour according to any preceding claim wherein said tumour comprises recombinant DNA encoding at least one copy of at least one of the genes listed in Tables 1A, 1 B or Tables 2A, 2B, or, at least one promoter corresponding thereto.

9. An isolated tumour according to Claim 8 wherein said tumour comprises multiple copies of said recombinant DNA.

10. An isolated tumour according to Claims 8 or 9 wherein said tumour comprises a homogeneous population of holoclone cells and said recombinant DNA comprises at least one gene encoding at least one of the products listed in Tables 1 A or Table 1 B, or at least one promoter corresponding thereto.

11. An isolated tumour according to Claims 8 or 9 wherein said tumour comprises a homogeneous population of paraclone cells and said recombinant DNA comprises at least one gene encoding at least one of the products listed in Tables 2A or 2B, or at least one promoter corresponding thereto.

12. An isolated tumour according to Claims 8-11 wherein said recombinant DNA comprises said promoter or a selected high expression promoter which is operationally linked to a reporter molecule whereby the promoter drives the expression of the reporter molecule.

13. A method for screening drugs for their util ity to treat tumours wherein said method comprises:

(a) exposing a population of tumour cells, that have been transfected with a recombinant molecule comprising at least one promoter region of any one or more of the genes encoding the products shown in Table 1A, Table 1 B or Table 2A, Table 2B and, operationally linked thereto, the corresponding gene or a selected reporter molecule, to a test drug; and

(b) monitoring the level of expression of said gene or reporter molecule in response to said drug and, optionally, where there is a reduction in expression concluding that the drug may be of use in treating said tumour, or, where there is a proportionate increase in the amount of said gene product or reporter molecule, due to the destruction or disabling of at least some of the tumour cells, concluding that the drug may be of use in treating said tumour.

14. A method for screening drugs to treat tumours wherein said method comprises: (a) exposing a cell, or population of cells, of either a holoclone or paraclone colony which has been transfected with a recombinant molecule comprising at least one promoter region of any one or more of the genes encoding the products shown in Tables 1A, 1B or Tables 2A, 2B, respectively, which promoter is operationally linked to its gene or a selected reporter molecular, to a test drug; and (b) monitoring the level of expression of said gene or reporter molecule in response to said drug and, optionally, where there is a reduction in expression concluding that the drug may be of use in treating said tumour.

15. A method for screening drugs for their utility to treat tumours wherein said method comprises:

(a) exposing a population of tumour cells, that have been transfected with a recombinant molecule comprising at least one high expression promoter operationally linked to of any one or more of the genes encoding the products shown in Table 1A, Table 1B or Table 2A, Table 2B or a selected reporter molecule, to a test drug; and (b) monitoring the level of expression of said gene or reporter molecule in response to said drug and, optionally, where there is a reduction in expression concluding that the drug may be of use in treating said tumour, or, where there is a proportionate increase in the amount of said gene product or reporter molecule, due to the destruction or disabling of at least some of the tumour cells, concluding that the drug may be of use in treating said tumour.

16. A method for screening drugs to treat tumours wherein said method comprises:

(a) exposing a cell, or population of cells, of either a holoclone or paraclone colony which has been transfected with a recombinant molecule comprising at least one high expression promoter which is operationally linked to any one or more of the genes encoding the products shown in Tables 1A, 1B or Tables 2A, 2B, respectively, or a selected reporter molecular, to a test drug; and

(b) monitoring the level of expression of said gene or reporter molecule in response to said drug and, optionally, where there is a reduction in expression concluding that the drug may be of use in treating said tumour.

17. A transfected tumour cell or a transfected population of tumour cells comprising recombinant DNA encoding at least one of the markers listed in Tables 1A, 1 B or Tables 2A, 2B and/or the corresponding promoter.

18. A transfected cell or a transfected cell population according to Claim 17 wherein said cell or said population of cells comprises cells exhibiting a holoclone morphology and said recombinant DNA encodes at least one of the markers listed in Tables 1A, 1 B and/or at least one corresponding promoter.

19. A transfected cell or a transfected cell population according to Claim 17 wherein said cell or population of cells comprises cells exhibiting a paraclone morphology and said recombinant DNA encodes at least one of the markers listed in Tables 2A, 2B and/or at least one corresponding promoter .

20. A homogeneous population of cells derived from a tumour wherein said cells comprise one of the following colony types: (i) holoclone; (ii) meraclone; or (iii) paraclone; wherein cells of said holoclone colony are characterising by expressing any one or more of the markers shown in Tables 1A or 1 B and cells of said paraclone colony are characterised by expressing any one or more of the markers shown in Tables 2A or 2B.

21. An antibody raised against any one or more of the markers shown in

Tables 1A, 1 B or Tables 2A, 2B.

22. An antibody according to Claim 21 wherein said antibody is monoclonal.

23. The use of antibodies according to Claims 21 or 22 to treat cancer.

24. A method for screening an agent in order to determine its ability to treat cancer comprising: (a) exposing an isolated tumour cell or tumour according to Claim 17, or a cell or homogeneous population of cells according to Claims 18 or 19 to said agent; and (b) determining the effects of said agent on said tumour cell or tumour or said cell or homogeneous populations of cells with a view to concluding, where said isolated tumour cell or tumour or said cell or homogeneous cell population is destroyed or disabled, that the agent has utility in the treatment of cancer.

25. A non-human animal for use in identifying agents that are effective in treating cancer comprising a non-human animal that has a tumour wherein said tumour comprises any one or more of cells corresponding to the following colony types: (i) holoclone; (ii) meraclone; or (iii) paraclone wherein cells of said holoclone colony are characterised by the expression of any one or more of the markers shown in Table 1A and 1 B and cells of the paraclone colony are characterised by expression of any one or more of the markers shown in Table 2A and 2B.

26. An isolated cell-specific marker comprising any one or more of the markers shown in Tables 1 A or 1 B or Tables 2A or 2B.

27. The use of any one or more of the markers shown in Tables A or 1 B or

Tables 2A or 2B for the identification, or development, of an agent to treat cancer.

28. The use of any one or more of the markers shown in Tables 1A or 1 B or Tables 2A or 2B for targeting agents to a specific cancer or cell.

29. The use of any one or more of the markers shown in Tables 1A or 1 B or Tables 2A or 2B for use in destroying or disabling a cancer cell.

30. A medicament for treating cancer wherein said medicament acts, either directly or indirectly, via one or more of the markers shown in Tables 1 A or 1 B or Tables 2A or 2B.

31. The use of any one or more of the markers shown in Tables 1 A or 1 B or Tables 2A or 2B, or antibodies raised thereto, in the manufacture of a medicament for the treatment of cancer.

32. A pharmaceutical composition comprising an agent that interacts with, or interferes with, either directly or indirectly, any one or more of the markers shown in Tables 1A or 1B or Tables 2A or 2B.