WO2010016766A2 - Antibodies recognizing endogenous human lgr5 and/or lgr6 - Google Patents

Abstract

The present invention provides novel antibodies that bind human Leucine-rich-repeat- containing G-protein-coupled Receptor 5 (Lgr5) and/or Lgr6. More specifically, the invention provides antibodies that recognize endogenous Lgr5 and/or Lgr6. The invention further provides methods for generating the antibodies and uses of the antibodies.

Description

Antibodies recognizing endogenous human Lgr5 and/or Lgr6.

Field The invention relates to the field of biochemistry, pharmacy and oncology. The invention particularly relates to the use of novel binding bodies that are specific for stem cell markers. Said antibodies are used for the isolation of adult tissue stem cells. The invention further relates to means suitable for cancer treatment and even more specific for the treatment of cancer stem cells. The invention also relates to means suitable for cancer diagnosis.

Background

Stem cells are undifferentiated cells that can renew themselves through mitotic cell division and that can differentiate into a diverse range of specialized cell types. Two broad types of mammalian stem cells are: embryonic stem cells that are isolated from the inner cell mass of blastocysts, and adult stem cells that are found in adult tissues.

Embryonic stem cells can differentiate into all of the specialized embryonic tissues.

Adult stem cells serve to maintain and repair the tissue in which they are found.

Adult stem cells act as a repair system for the body, replenishing specialized cells, but also maintain the normal turnover of regenerative organs, such as blood, skin or intestinal tissues.

Some tissues, such as the epidermis of the skin, the lining of the small intestine, and bone marrow, undergo continuous cellular turnover. They contain epithelial stem cells, which persist indefinitely, and a much larger number of "transit amplifying cells," which arise from the stem cells and divide a finite number of times until they become differentiated.

The self-renewing epithelium of the small intestine is ordered into crypts and villi (Gregorieff and Clevers, 2005. Genes Dev 19, 877-90). Cells are newly generated in the crypts and are lost by apoptosis at the tips of the villi, with a resulting epithelial turn-over time of 5 days in the mouse. Self-renewing stem cells have long been known to reside near the crypt bottom and produce the rapidly proliferating transit amplifying (TA) cells capable of differentiating towards all lineages. The estimated number of stem cells is between 4 and 6 per crypt (Bjerknes and Cheng, 1999. Gastroenterology 116, 7-14). Three differentiated cell types (enterocytes, goblet cells and enteroendocrine cells) form from TA cells and continue their migration in coherent bands along the crypt-villus axis. Each villus receives cells from multiple different crypts. The fourth major differentiated cell-type, the Paneth cell, resides at the crypt bottom.

Stem cells are also present in cancerous tissue. The cancer stem cell hypothesis postulates that a small reservoir of self-sustaining cells is exclusively able to self- renew and maintain a tumor. These cancer stem cells can expand the cancer stem cell pool, but will also generate the heterogeneous cell types that constitute the bulk of the tumor. Cancer stem cells are relatively refractory to therapies that have been developed to eradicate the rapidly dividing cells that constitute the bulk of a tumor. Cancer stem cells may also be the most likely cells to metastasize. Thus, the cancer stem cell hypothesis would require that we re-think the way tumors are diagnosed and treated. Therapy would have to target also said stem cell population that fuels tumor growth and metastasis, in addition to the bulk of the tumor. The cancer stem cell hypothesis is at the centre of a rapidly evolving field and may dictate changes in how basic and clinical researchers view cancer.

There is some confusion in the literature as to the definition of a cancer stem cell. Here, we follow the consensus reached at a recent AACR workshop (Clarke et al., 2006. Cancer Res. 66:9339-44), which states that the cancer stem cell "is a cell within a tumor that possesses the capacity to self-renew and to cause the heterogeneous lineages of cancer cells that comprise the tumor. Cancer stem cells can thus only be defined experimentally by their ability to recapitulate the generation of a continuously growing tumor". Alternative terms in the literature include tumor- initiating cell and tumorigenic cell. Assays for cancer stem cell activity need to address the potential of self-renewal and of tumor propagation. The gold- standard assay currently is serial xeno-transplantation into immunodeficient mice.

Recent studies (O'Brien et al., 2007. Nature 445:106-10 ; Ricci-Vitiani et al., 2007. Nature. 445:111-5) imply that colon cancer may represent an example of a solid tumor in which a small number of cancer stem cells is responsible for maintenance of the tumor. Sorting for expression of a presumed stem cell marker, CD 133, enriches significantly for the cancer stem cell, but the resulting cell mixture remains far from pure. It therefore remains unclear what the exact properties are of the cancer stem cells within the sorted cell preparation, such as their cell cycle status, or their resistance to chemotherapy or radiation.

US 2004/0058392 and EP 1 400 807 describe a list of TCF target genes that were defined in the colon cancer cell line Lsl74T (van de Wetering et al, 2002. Cell 111 : 241-250). In the applications, it was postulated that these molecules expressed in colon cancer cells and in intestinal crypts would represent stem cell markers. Several of the markers encode cell-surface proteins. The inventors of US 2004/0058392 and EP 1 400 807 contemplated that these proteins can be used as markers for selection of the abundant stem cell population in the gut. However, it turned out that subsequent research by applicants showed that a subset of these proteins show the most promise as stem cell markers because of their highly selected expression pattern on stem cells, whereas other proteins were less promising. Examples of the latter category include e.g. the CD44 protein, which is expressed by all dividing crypt cells (Wielenga et al., 1999. Am J Pathol 54: 515-523) as is cMyb (Malaterre et al., 2007. Proc. Natl. Acad. Sci USA 104, 3829-3834) and GPX2 (Brigelius-Flohe, (2006) Biological Chemistry, 387 (10-11): 1329-1335). The c-Kit protein is expressed on non-dividing entero-endocrine cells (Neid and Wittekind, 2007. Chirurg Gastroenterol Interdisz 23 (2): 108-112). EphB2 is expressed by all dividing cells and EphB3 is expressed by the non-dividing Paneth cells (Battle et al., 2002. Cell 111 : 251-263). BMP4 is expressed by stromal cells in the villus (Haramis et al., 2004. Science. 303: 1684-1686). And Claudinl is expressed almost ubiquitously (Hewitt et al., 2006. BMC Cancer, 6, art. no. 186). Examples of proteins with a stem cell specific expression pattern are human Leucine-rich-repeat-containing G-protein- coupled Receptor 5 (Lgr5) and/or Lgr6.

A clear need exists for methods and means that allow identification of adult stem cells and cancer stem cells. Adult stem cells are potentially useful in the repair of damaged or diseased tissue, and therefore there is a clear need for a method that allows for the isolation of adult stem cells. Furthermore, there is a clear need for methods that allow the identification and eradication of cancer stem cells. The goal of the present invention is to provide an antibody or epitope-binding fragment that identifies adult stem cells and/or cancer stem cells. Yet another goal of the present invention is to eradicate cancer stem cells by using said antibody or epitope-binding protein.

Description of the invention

In a first aspect, there is provided an antibody which binds a human Leucine-rich- repeat-containing G-protein-coupled Receptor 5 (Lgr5) and/or Lgr6. Preferably said antibody binds an endogenous human Leucine-rich-repeat-containing G-protein-coupled Receptor 5 (Lgr5) and/or Lgr6.

An antibody according to the present invention is preferably capable of binding and thereby detecting endogenous human Lgr5 and/or Lgr6 in a cell or cells, including living cells, tissue or living tissue, organ or living organ, organoid or living organoid. Preferred antibodies recognize an extracellular part of Lgr5 and/or Lgr6 such as, but not limited to, the extracellular N-terminal domain of Lgr5 and/or Lgr6. The antibodies according to the invention provide valuable tools for isolating human adult epithelial stem cells from diverse tissues, allowing in vitro culturing of organoids that can be used for treatment of damaged or diseased tissue. Moreover, these antibodies will be of great importance in pre-clinical experiments when studying effects of anticancer stem cell therapy, including the use of anti-Lgr5 and/or Lgr6 antibodies, in animals on xeno-transplanted human tumors.

The term endogenous refers to Lgr5 and/or Lgr6 receptor proteins that are expressed from the cellular genes that occur naturally in the cells. Endogenous may be replaced by the word native. In this case, it preferably means that a Lgr5 and/or Lgr6 receptor is natively expressed in a cell. By opposition to a Lgr5 and/or Lgr6 receptor that is expressed or over-expressed by means of recombinant DNA technology in a cell.

The term organoids (see example 12) refers to organ or tissue-like structures that are grown in vitro starting from isolated stem cells or isolated tissue fragments comprising stem cells. Although several antibodies against Lgr5 and/or Lgr6 are known in the art, none of these antibodies is capable of recognizing endogenous human Lgr5 and/or Lgr6. For example, Becker et al. (Becker et al, 2008. Scientific World J 8: 1168-1176) describe immuno staining of Lgr5 in a subpopulation of epithelial cells of normal human colon and small intestinal mucosa. The staining that is observed, however, is not limited to the base of the crypts, where Lgr5 positive stem cells are known to reside (Barker et al, 2007. Nature 449:1003-7). In fact, a high level of staining was observed also in non-epithelial cells, suggesting that the obtained staining pattern is not specific for stem cells. Moreover, colonic adenomas showed staining at the tip of the villi, a position where stem cells, including cancer stem cells, are known not to be present. Further reported antibodies against Lgr5 were only capable of recognizing Lgr5 after over-expression in HEK293 cells (WO2009005809). In addition, several Lgr5 antibodies are commercially available. Most of them are rabbit polyclonal antibodies and none of them seems to be suitable to be used in FACS analysis. For example, rabbit polyclonals from US BIO (product number: G8600-53) or from Abgent (product number: AP2754f or b) could be cited.

Lgr5 and Lgr6 are members of the leucine-rich repeat-containing G-Protein Coupled Receptors (GPCRs), which form a subgroup of the GPCR superfamily. Lgr5 and Lgr6 encode a 7-transmembrane protein with an N-terminal signal peptide and comprising leucine-rich repeats. No ligands have been identified that are capable of binding to Lgr5 and/or Lgr6, although it has been suggested that R-spondins might be ligands for Lgr receptors (WO2009005809). Other potential ligands for Lgr5 and/or Lgr6 comprise members of the insulin peptide family, such as Insl5 or relaxin3, or a cysteine-knot protein such as Noggin, Gremlinl or -2, Dan, or Cerberus. The nucleotide and amino acid sequences of these ligands are known and the skilled person is thus for example capable to produce said ligand protein. A human cDNA sequence of Lgr5 is represented by SEQ ID NO: 59. A corresponding amino acid sequence is represented by SEQ ID NO: 60. A human cDNA sequence of Lgr6 is represented by SEQ ID NO: 61. A corresponding amino acid sequence is represented by SEQ ID NO:62.

Lgr5 is specifically expressed in cycling Crypt Base Columnar (CBC) cells, which are small cells that are interspersed between the Paneth cells (Barker et al., 2007. Nature 449: 1003-1007). It was shown by lineage tracing that the Lgr5⁺ CBC cells constitute multipotent stem cells which generate all cell types of the epithelium. Also Lgr6, but not Lgr4, is a unique marker for adult stem cells. While Lgr5 is expressed in stem cells of brain, kidney, liver, retina, stomach, intestine, pancreas, breast, hair follicle, ovary, adrenal medulla, and skin, Lgr6 is expressed in stem cells of brain, lung, breast, hair follicle, and skin.

A further preferred antibody according to the invention is an antibody that binds a non-human Lgr5 and/or Lgr6 such as mouse Lgr5 and/or Lgr6, in addition to human Lgr5 and/or Lgr6. For example, antibodies that recognize both human and mouse Lgr5 and/or Lgr6 can be used in pre-clinical experiments to study effects of anticancer stem cell therapy in mice on xeno -transplanted human tumors. A more preferred antibody binding and thereby recognizing both human and mouse Lgr5 is 9G5. An hybridoma cell producing 9G5 has been deposited at the Belgian Coordinated Collections of Microorganisms (BCCM / LMBP) institute on

28/07/2009 with following accession number LMBP 6966CB. A more preferred antibody binding and thereby recognizing both human and mouse Lgr6 is Id8 or 3d8.

An antibody that specifically recognizes endogenous human Lgr5 and/or Lgr6 allows monitoring of the effect of the anti-cancer stem cell therapy on the xeno-tranplanted human tumors only. Antibodies that recognize mouse and human Lgr5 and/or Lgr6 allow monitoring of both cancer stem cells and mouse adult tissue stem cells. Especially when the anti-cancer stem cell therapy comprises Lgr5 and/or Lgr6 antibodies, the use of human- specific Lgr5 and/or Lgr6 antibodies provides targeting of only human cancer stem cells, because these antibodies do not interact with mouse adult epithelial stem cells. In contrast, antibodies that recognize both human and mouse Lgr5 and/or Lgr6 will target both human cancer stem cells and mouse adult stem cells. The latter condition more closely resembles anti-cancer stem cell therapy on human using antibodies that recognize human Lgr5 and/or Lgr6, because these will target both cancer stem cells and adult tissue stem cells.

An antibody of the invention can be a polyclonal or a monoclonal antibody. A preferred antibody according to the invention is a monoclonal anti-Lgr5 and/or Lgr6 antibody. In a more preferred embodiment, the invention provides rat monoclonal antibodies selected from 1D9, 2F10, 4Dl 1, 6C10, 9B3, 3A4, 5A7, 6G2, 9G5, 2B8, 3B9, 5C8, 7Bl 1, lOCl, 4D6, 5E9, 8F2, that recognize human Lgr 5, and/or rat antibodies selected from 6d8, 2f4, Id8 and 3d8, that recognize human Lgr6. In addition, antibody 9G5 recognizes both human and mouse Lgr5, while antibodies Id8 and 3d8 also recognize mouse Lgr6 and hLgr5, in addition to human Lgr6.

Antibodies Id8 and 3d8 are attractive since they can be used when it is needed to block or activate both Lgr5 and Lgr6 at the same time in a given cell population. Preferred cell population is cancer cells and/or stem cells. More preferred cells are stem cells and/or cancer stem cells which are believed to express both Lgr5 and Lgr6. The antibody 9G5 is a preferred antibody since as explained above, it is advantageous to bind and thereby recognize both human and mouse Lgr5.

These antibodies are fully characterized herein with respect to the identification of their complementarity-determining regions (CDRs), the amino acid sequences of both their light and heavy chain variable regions, the identification of their surface amino acids, and means for their expression in recombinant form.

The region of Lgr5 recognized by each of the Lgr5 antibody has been characterized (see for more details example 11, table 9). The 12 identified Lgr5 antibodies could be clustered into three distinct groups; each group binding a distinct region of Lgr5. A first group comprises 1D9. An hybridoma producing 1D9 has been deposited at the institute BCCM / LMBP.on 28/07/2009 with following accession number LMBP 696 ICB. A second group comprises 4Dl 1 and 7Bl 1. An hybridoma producing 4Dl 1 has been deposited at the institute BCCM / LMBP on 28/07/2009 with following accession number LMBP 6962CB. A third group comprises remaining identified antibodies (6C10, 3A4, 5A7, 9G5, 2B8, 3B9, 5C8, 8F2 and lOCl). Hybridomas producing 6C10, 5A7, 9G5 or 8F2 have been deposited at the institute BCCM / LMBP on 28/07/2009 with following accession numbers LMBP 6964CB, LMBP 6963CB, LMBP 6966CB and LMBP 6965CB.In a preferred embodiment, for any use of a Lgr5 antibody, it is advantageous to use two or three Lgr5 antibodies as identified herein wherein each Lgr5 antibody has been clustered in a separate group. This combined use of more than one Lgr5 antibody, each belonging to a distinct cluster as identified herein will provide a more sensitive and therefore more efficient detection of Lgr5 since a stronger signal for detection of Lgr5 is expected to be provided. Therefore it is expected that a method for isolating stem cells, a use of several Lgr5 antibodies for targeting cancer stem cells, a method for determining cancer stem cell content of a tumor or a body fluid each as later identified herein will each be more sensitive when using more than one Lgr5 antibody each belonging to a distinct cluster. For example it is preferred to use:

- 1D9 together with 4Dl 1, or

- lD9 with 7Bl l, or

-1D9 with an antibody selected from 6C10, 3A4, 5A7, 9G5, 2B8, 3B9, 5C8, 8F2 and

IOCI or -4Dl 1 with an antibody selected from 6C10, 3A4, 5A7, 9G5, 2B8, 3B9, 5C8, 8F2 and 1 OCl or

-7Bl 1 with an antibody selected from 6C10, 3A4, 5A7, 9G5, 2B8, 3B9, 5C8, 8F2 and 1 OCl or

-1D9 with 4Dl 1 and an antibody selected from 6C10, 3A4, 5A7, 9G5, 2B8, 3B9, 5C8, 8F2 and IOCI or

-1D9 with 7Bl 1 and an antibody selected from 6C10, 3A4, 5A7, 9G5, 2B8, 3B9,

5C8, 8F2 and IOCI.

It is to be noted that each of the Lgr5 antibodies as identified herein is able to specifically bind human Lgr5 in a native setting, i.e. organoid (see example 12). This is the first time Lgr5 antibody has been succesfully been reported to bind endogenous human Lgr5. For example, WO 2009/005809 only disclosed staining in a cell line over-expressing Lgr5. This suggests that antibodies identified herein are specific and sensitive for Lgr5 and are potentially valuable and powerful diagnostic and therapeutic tools.

In addition, it suggests that antibodies identified herein are specific and sensitive enough to be used for in vivo stainings and not only used for staining of cells over- expressing Lgr5 and/or Lgr6 as was the case in WO 2009/005809. Any of the use or method as identified herein with a Lgr5 and/or Lgr6 antibody may be an in vitro or ex vivo use or method. Alternatively, any of the use or method as identified herein may be an in vivo use or method. In vitro may mean that said use or method is carried out in a sample, in a cell free system or in a cell. Said cell is preferably isolated from the organism wherein it originates. In vivo may mean that said use or method is carried out in an organoid or in a tissue or in an animal, preferably a mammal, more preferably a human.

A conventional antibody is comprised of two identical heavy chains and two identical light chains that are joined by disulfide bonds. Each heavy and light chain contains a constant region and a variable region. Each variable region contains three CDRs which are primarily responsible for binding an epitope of an antigen. They are referred to as CDRl, CDR2, and CDR3, numbered sequentially from the N-terminus, of which the CDR3 region comprises the most variable region and normally provides a substantial part of the contact residues to a target. The more highly conserved portions of the variable regions are called the "framework regions".

The term antibody is used herein in the broadest sense and specifically covers monoclonal antibodies (including full length monoclonal antibodies) of any isotype such as IgG, IgM, IgA, IgD and IgE, polyclonal antibodies, multispecifϊc antibodies, chimeric antibodies, and antibody fragments. An antibody reactive with a specific antigen can be generated by recombinant methods such as selection of libraries of recombinant antibodies in phage or similar vectors, or by immunizing an animal with the antigen or an antigen-encoding nucleic acid. The term antibody includes functional equivalents of an antibody. Methods of producing functional equivalents are known to the person skilled in the art and comprise an antibody that comprises non-essential point mutations. Further functional equivalents are obtained by producing multiple monoclonal antibodies directed against Lgr5 and/or Lgr6 followed by testing whether said obtained monoclonal antibodies are able to compete with any of the antibodies disclosed in the present invention. These kinds of equivalents are obtained by routine experiments and hence are also included herein.

A monoclonal antibody is an antibody obtained from a population of substantially homogeneous antibodies, i.e. the antibodies forming this population are essentially identical except for possible naturally occurring mutations which might be present in minor amounts. These antibodies are directed against a single epitope.

An epitope is the site on the antigen to which an antibody binds. If the antigen is a polymer, such as a protein or polysaccharide, the epitope can be formed by contiguous residues or by non-contiguous residues brought into close proximity by the folding of an antigenic polymer. In proteins, epitopes formed by contiguous amino acids are typically retained on exposure to denaturing solvents, whereas epitopes formed by noncontiguous amino acids are typically lost under said exposure.

A preferred antibody according to the invention comprises one or more CDRs having an amino acid sequence selected from Figure 16 or 17. A most preferred antibody comprises at least a CDR 3 region having an amino acid sequence selected from Table 4 or 5.

In one embodiment, an antibody according to the invention comprises a single domain antibody, a F(ab')2, Fab, Fab', Facb, or single chain Fv (scFv) fragment. An Fc fragment, which for example activates complement and may bind to Fc receptors, can be present but is not required for an antibody and variants or derivatives thereof. Said antibody fragment comprises at least one CDR3 region having an amino acid sequence selected from Table 4 or 5; more preferred at least a heavy chain CDR3 region and a light chain CDR3 region, more preferred three, four, five or six CDR regions having an amino acid sequence selected from Figure 16 or 17. A most preferred fragment comprises 3 heavy chain CDR regions and 3 light chain CDR regions from a heavy chain and corresponding light chain antibody sequence selected from Figure 16 or 17.

A scFvs fragment is an epitope-binding fragment that contains at least one fragment of an antibody heavy chain variable region (VH) linked to at least one fragment of an antibody light chain variable region (VL). The linker may be a short, flexible peptide selected to assure that the proper three-dimensional folding of the VL and VH regions occurs once they are linked so as to maintain the target molecule binding- specificity of the whole antibody from which the single-chain antibody fragment is derived. The carboxyl terminus of the VL or VH sequence may be covalently linked by a linker to the amino acid terminus of a complementary VL or VH sequence.

A further preferred antibody according to the invention is a monoclonal antibody. Yet a further preferred antibody or derivative thereof is a chimeric antibody, a nanobody, and/or a bispecifϊc antibody.

A chimeric antibody comprises a binding portion, for example the variable region or part thereof of the heavy and light chains, of a non-human antibody, while the remainder portion, for example the constant region of the heavy and light chains, is of a human antibody. A nanobody is a single domain antibody that occurs naturally in camelids. In contrast to standard antibodies, nanobodies are relatively simple proteins comprising only a heavy chain- like variable region. A preferred nanobody of the invention comprises at least a CDR3 region selected from one of the heavy chain amino acid sequences depicted in Table 4 or 5. Bispecifϊc antibodies are artificially engineered monoclonal antibodies that consist of two distinct binding sites and are capable of binding two different epitopes such as, for example, two different Lgr5 epitopes or an Lgr5 and Lgr6 epitope.

The invention is not limited to the antibodies as identified herein. In a preferred embodiment, an antibody is a human antibody. Human antibody can be made following techniques well-known in the art, and described by G. Kohler and C. Milstein (Nature, 1975: 256: 495-497). As used herein, the term "human antibody" means an antibody in which the variable and constant domain sequences are derived from human sequences. A human antibody provides a substantial advantage in a use of the present invention, as it is expected to minimize the immunogenic and allergic responses that are associated with use of non-human antibodies in a human patient. An antibody can be raised by immunizing rodents (e. g. mice, rats, hamsters and guinea pigs) with either native human Lgr5 and/or Lgr6 as expressed on cells or purified from human cells or tissues, or recombinant Lgr5 and/or Lgr6 or its fragments, expressed in a eukaryotic or prokaryotic system. Other animals can be used for immunization, e. g. non-human primates, transgenic mice expressing human immunoglobulins and severe combined immunodeficient (SCID) mice transplanted with human B lymphocytes. Hybridomas can be generated by conventional procedures by fusing B lymphocytes from the immunized animals with myeloma cells (e. g. Sp2/0 and NSO), as described by G. Kohler and C. Milstein, Nature, 1975: 256: 495-497. In addition, an antibody can be generated by screening of recombinant single-chain Fv or Fab libraries from human B lymphocytes in phage-display systems. For treating a human subject, an antibody would preferably be a chimeric, deimmunised, humanized, human-like, resurfaced or humanized monoclonal or human antibodies.

Such antibodies can reduce immunogenicity and thus avoid human anti-mouse antibody (HAMA) response. It is preferable that the antibody be IgG4, IgG2, or other genetically mutated IgG or IgM which does not augment antibody-dependent cellular cytotoxicity (S. M. Canfield and S. L. Morrison, J. Exp. Med., 1991 : 173: 1483-1491) and complement mediated cytolysis (Y. Xu et al, J. Biol. Chem., 1994: 269: 3468- 3474; V. L. Pulito et al., J. Immunol, 1996; 156: 2840-2850).

A chimeric antibody may be produced by recombinant processes well known in the art, and has an animal variable region and a human constant region.

A humanized antibody usually has a greater degree of human peptide sequences than do chimeric antibodies. In a humanized antibody, only the complementarity determining regions (CDRs), which are responsible for antigen binding and specificity are animal derived and have an amino acid sequence corresponding to the animal antibody, and substantially all of the remaining portions of the molecule (except, in some cases, small portions of the framework regions within the variable region) are human derived and correspond in amino acid sequence to a human antibody (see L. Riechmann et al., Nature, 1988; 332:323-327; G. Winter, United States Patent No. C. Queen et al., U. S. patent number 5,530, 101). Methods for humanizing non-human antibodies are known in the art. As is known to the skilled person, antibodies such as rat antibodies can be humanized by grafting their CDRs onto the variable light (VL) and variable heavy (VH) frameworks of human Ig molecules, while retaining those rat framework residues deemed essential for specificity and affinity (Jones, P.T., et. al., (1986). Nature, 321, 522). Overall, CDR grafted antibodies consist of more than 80% human amino acid sequences (Queen, C. et al. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 10029; Carter, P. et al. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 4285). Despite these efforts, CDR-grafted, humanized antibodies were shown to still evoke an antibody response against the grafted V region (Hwang and Foote, (2005). Methods 36: 3-10).

A deimmunised antibody is an antibody in which the T and B cell epitopes have been eliminated, as described in International Patent Application PCT/GB98/01473. They have reduced immunogenicity when applied in vivo. To further decrease the content of non- human sequences in therapeutic mAbs, humanization methods based on different paradigms such as resurfacing (Padlan, E. A., et. al., (1991). MoI. Immunol., 28, 489), superhumanization (Tan, P., D. A., et. al., (2002) J. Immunol, 169, 1119), human string content optimization (Lazar, G. A., et. al., (2007). MoI. Immunol., 44, 1986) and humaneering have been developed (Almagro, J. C, et. al., (2008). Frontiers in Bioscience 13, 1619). As in CDR grafting approaches, these methods rely on analyses of the antibody structure and sequence comparison of the non- human and human mAbs in order to evaluate the impact of the humanization process into immunogenicity of the final product.

A preferred de-immunized, human-like, resurfaced, or humanized monoclonal antibody comprises an amino acid homology to an amino acid sequence as identified in Figure 16 or 17 of at least about 80%, at least about 90%, such as about 91%, more preferred about 92%, more preferred about 93%, more preferred about 94%, more preferred about 95%, more preferred about 96%, more preferred about 97%, more preferred about 98%, or more preferred about 99% sequence homology, as determined, for example, by the FASTA search method in accordance with Pearson and Lipman, 1988, Proc. Natl. Acad. ScL USA, 85: 2444-2448. A more preferred de- immunized, human-like, resurfaced, or humanized monoclonal antibody comprises an amino acid homology to an amino acid sequence as identified in Figure 16 or 17 of at least 80%, at least 90%, such as 91%, more preferred 92%, more preferred 93%, more preferred 94%, more preferred 95%, more preferred 96%, more preferred 97%, more preferred about 98%, or more preferred 99% sequence homology, as determined, for example, by the FASTA search method in accordance with Pearson and Lipman, 1988, Proc. Natl. Acad. ScL USA, 85: 2444-2448.

De-immunization is another approach developed to reduce the immunogenicity of chimeric or rat antibodies. It involves the identification of linear T-cell epitopes in the antibody of interest, using bioinformatics, and their subsequent replacement by site-directed mutagenesis to human or non- immunogenic sequences (WO09852976A1). Although de-immunized antibodies exhibited reduced immunogenicity in primates, compared with their chimeric counterparts, some loss of binding affinity was observed (Jain, M., et. al., (2007). Trends in Biotechnol. 25, 307).

A human antibody can be made by several different ways, including by use of human immunoglobulin expression libraries (Stratagene Corp., La Jolla, California) to produce fragments of human antibodies VH, VL, Fv, Fd, Fab, or (Fab')2, and using these fragments to construct whole human antibodies using techniques similar to those for producing chimeric antibodies. Alternatively, these fragments may be used on their own as antibody. Human antibodies can also be produced in transgenic mice with a human immunoglobulin genome. Such mice are available from Abgenix. Inc., Fremont, California, and Medarex, Inc., Annandale, New Jersey.

One can also create single peptide chain binding molecule in which the heavy and light chain Fv regions are connected. Single chain antibodies ("ScFv") and the method of their construction are described in U. S. Patent No. 4,946,778. Alternatively, Fab can be constructed and expressed by similar means (M. J. Evans et al.,

J.Immunol. Meth., 1995; 184:123-138). All of the wholly and partially human antibodies are less immunogenic than wholly murine MAbs, and the fragments and single chain antibodies are also less immunogenic. All these types of antibodies are therefore less likely to evoke an immune or allergic response. Consequently, they are better suited for in vivo administration in a human subject than wholly animal antibodies, especially when repeated or long-term administration is necessary. In addition, the smaller size of the antibody fragment may help improve tissue bioavailability, which may be critical for better dose accumulation in acute disease indications, such as tumor treatment. Based on the molecular structures of the variable regions of a Lgr5 and/or Lgr6 antibody, one could use molecular modeling and rational molecular design to generate and screen molecules which mimic the molecular structures of the binding region of the antibodies and activate CTLs. These small molecules can be peptides, peptidomimetics, oligonucleotides, or other organic compounds. The mimicking molecules can be used for treatment of cancers. Alternatively, one could use large- scale screening procedures commonly used in the field to isolate suitable molecules from libraries of compounds. Antibody characteristics, such as on-rates (ka), off-rates (kd) and affinities (KD) can be determined in competitive binding assays using known platforms such as Octet™ (ForteBio), ProteOn™ (Bio-Rad), and Biacore™ (GE Healthcare). By testing whether antibodies block one another's binding to their antigen in a pair wise fashion, a blocking profile for each antibody relative to the others in the panel is established. A preferred antibody according to the invention binds Lgr5 and/or Lgr6 with a ko of 3 x 10^"8 M or lower, more preferred with a ko of 6 x 10^~8 M or lower, more preferred with a ko of 1,2 x 10^~9 M or lower, most preferred with a ko of 3 x 10^{" 9} M or lower.

In a further aspect, there is provided a nucleic acid encoding an antibody according to the invention. A nucleic acid as used in the invention is typically but not exclusively a ribonucleic acid (RNA) or a deoxyribonucleic acid (DNA). Alternative nucleic acids are available for a person skilled in the art, such as for instance peptide nucleic acids (PNA). A nucleic acid according to the invention may be present in a plasmid or expression construct. A nucleic acid is for instance comprised in a cell. When said nucleic acid is expressed in said cell, said cell produces a polypeptide and/or a binding body and/or an antibody according to the invention. Therefore, the invention in one embodiment provides a cell comprising a nucleic acid according to the invention. Said cell is preferably an animal cell, more preferably a mammal cell, more preferably a primate cell, most preferably a human cell. For the purposes of the invention a suitable cell is any cell capable of comprising and expressing a nucleic acid according to the invention. A preferred nucleic acid according to the invention comprises a nucleic acid sequence that encodes an amino acid sequence as identified in Figure 16 or 17. A most preferred nucleic acid according to the invention comprises at least one of the nucleotide sequences as identified in Figure 22 or Figure 23. Methods and means for generating cells that produce a polypeptide encoded by a nucleic acid according to the invention are known in the art. It will be clear to the skilled person that the nucleotide sequence is preferably adapted to the optimal codon usage of the cell or organism in which the polypeptide encoded by a nucleic acid is produced.

In a further aspect, there is provided a cell producing an antibody according to the invention. In a preferred embodiment, said cell is a hybridoma cell, a Chinese hamster ovary (CHO) cell, an NSO cell or a PER-C6™ cell. In a particularly preferred embodiment said cell is a CHO cell. Some preferred hybridoma cells producing 1D9, 4Dl 1, 6C10, 5A7, 9G5, 8F2 have been deposited at the institute BCCM / LMBP on 28/07/2009 and got the following accession numbers LMBP 696 ICB, LMBP 6962CB, LMBP 6964CB, LMBP 6963CB, LMBP 6966CB, LMBP 6965 CB. Further provided is a cell culture comprising a cell according to the invention. Various institutions and companies have developed cell lines for the large scale production of antibodies, for instance for clinical use. Non- limiting examples of such cell lines are CHO cells, NSO cells or PER.C6™ cells. Cell lines developed for industrial scale production of proteins and antibodies are herein further referred to as industrial cell lines. Thus in a preferred embodiment the invention provides the use of a cell line developed for large scale production of protein and/or antibody for the production of an antibody according to the invention. A most preferred cell or cell line is capable of producing multiple antibody according to the invention, such as for example 2 different antibodies or epitope-binding fragments thereof, or 3 different antibodies or epitope-binding fragments thereof, most preferably in the same cell.

In a further aspect, there is provided a method for producing an antibody comprising culturing a cell of the invention and harvesting said antibody from said culture. Preferably said cell is cultured in a serum free medium. Preferably said cell is adapted for suspension growth. Further provided is an antibody obtainable by a method for producing an antibody according to the invention. The antibody is preferably purified from the medium of the culture. Preferably said antibody is affinity purified. A cell of the invention is for instance a CHO cell, an NSO cell or another cell type known for its suitability for antibody production for clinical purposes. In a particularly preferred embodiment said cell is a human cell, preferably a cell that is transformed by an adenovirus El region or a functional equivalent thereof. A preferred example of such a cell line is the PER.C6™ cell line or equivalent thereof. In a particularly preferred embodiment said cell is a CHO cell or a variant thereof. This variant preferably makes use of a Glutamine synthetase (GS) vector system for expression of an antibody.

Expression vectors for expression of an antibody are known in the art and can be generated using materials and methods that are known in the art. A heterodimeric antibody can be generated by transfecting or infecting a cell or cell line with equimolar quantities of two expression vectors that each expresses one of the heterodimeric proteins. Alternatively, an expression vector that expresses both proteins can be used. Standard transfection of infection procedures can be used, such as calcium phosphate precipitation or lipofectin. Selection of a desired cell line may be carried out in accordance with standard procedures known for the particular selectable markers that are present on the expression vector.

Antibodies are glycoproteins containing between 3 and 12% carbohydrate. The carbohydrate units are transferred to acceptor sites on the antibody chains after the heavy and light chains have combined. The major carbohydrate units are attached to amino acid residues of the constant region of the antibody. The carbohydrate units may affect overall solubility and the rate of catabolism of the antibody. It is also known that carbohydrate is necessary for cellular secretion of some antibody chains. It has been demonstrated that glycosylation of the constant region plays a vital role in the effector functioning of an antibody; without this glycosylation in its correct configuration, the antibody may be able to bind to the antigen but may not be able to bind for example to macrophages, helper and suppressor cells or complement, to carry out its role of blocking or lysing the cell to which it is bound. It is preferred that glycosylation of the antibody by the producer cells maintains antigen binding capability and effector functionality.

Antibodies produced according to a method of the invention may further differ in acetylation, pegylation, phosphorylation, and/or amidation, compared to the antibody produced by the hybridoma cell.

In each of the use of a Lg5 and/or Lgr6 antibody as defined below, a Lgr4 antibody may replace a Lgr5 antibody or a Lgr4 antibody may be used in addition to a Lgr5 and/or Lgr6 antibody. Lgr4 antibodies are commercially available. For example, a monoclonal Lgr4 antibody from Antibodies on line is available (product number: 12H6).

In a further aspect, there is provided a method for isolating stem cells or enriching a cell suspension in stem cells comprising preparing a cell suspension from a tissue or organ sample, contacting said cell suspension with an Lgr5 and/or 6 binding antibody according to the invention, obtaining cells bound to said antibody, and isolating cells from said antibody. "Enriched", as in an enriched population of cells, can be defined phenotypically based upon an increased number of cells having a given marker such as Lgr5 and/or Lgr6 in a suspension of cells as compared with the number of cells having the marker in the unfractionated set of cells. A cell suspension may be a "sample". It may mean a specimen or culture obtained from any source, as well as biological and environmental samples. Biological samples can be obtained from animals (including humans) and encompass fluids, solids, tissues, and gases. Biological samples include blood products, such as plasma, serum and the like.

In an alternative embodiment, two or more antibodies directed to two or more different epitopes are used to obtain a substantially complete capture of the desired stem cells. Said two or more antibodies can also be directed against two different stem cell markers (i.e. one antibody against Lgr5 and one against Lgr6). Whenever use is made of two or three or even more antibodies, said antibodies may be from the same class, for example all being antibodies, or may be from different classes, such as for example an antibody directed to Lgr6 and an epitope-binding fragment for binding to Lgr5.

In a preferred embodiment, at least two different antibodies or epitope-binding fragments thereof capable of binding to Lgr5 and/or Lgr6 are contacted with a cell suspension.

After allowing the binding compounds to interact with the cell suspension (for a certain amount of time or under different conditions such as pH, temperature, salt etc.), subsequent identification of obtained bound complexes is performed. This is for example accomplished by using FACS analysis. Fluorescence-activated cell- sorting (FACS) is a specialised type of flow cytometry. It provides a method for sorting a heterogeneous mixture of biological cells into two or more containers, one cell at a time, based upon the specific light scattering and fluorescent characteristics of each cell. It is a useful scientific instrument as it provides fast, objective and quantitative recording of fluorescent signals from individual cells as well as physical separation of cells of particular interest. Moreover, it allows for sorting of living cells expressing Lgr5 and/or Lgr6. These living cells may then subsequently be used in other (culture) assays.

Alternatively, Lgr5 expression may be detected by immunohistochemistry. Examples of technique that can be routinely used are ELISA (enzyme- linked immunosorbant assay), "sandwich" immunoassays, immunoradiometric assays, gel diffusion precipitation reactions, immunodiffusion assays, in situ immunoassays (e g , using colloidal gold, enzyme or radioisotope labels, for example), Western blots, precipitation reactions, agglutination assays (e g , gel agglutination assays, hemagglutination assays, etc.), complement fixation assays, immunofluorescence assays, protein A assays, FACS (Fluorescence Activted Cell Sorting), and

Immunoelectrophoresis assaysln one embodiment, antibody binding is detected by detecting a label on the Lgr5 antibody. In another embodiment, the Lgr5 antibody is detected by detecting binding of a secondary antibody or reagent to the Lgr5 antibody. In a further embodiment, the secondary antibody is labeled. Many methods are known in the art for detecting binding in an immunoassay and are within the scope of the present invention.

In a preferred embodiment, for detecting Lgr5, any of the Lgr5 antibodies 9G5, 2B8, 3B9, 8F2, 6C10, lOCl, 2F10 and/or 3B9 is used in FACS. Each of these antibodies has already been extensively tested and is functional in FACS (see for example results shown in figures 13, 14 and 20 and table 1).

In another preferred embodiment, for detecting Lgr5, the Lgr5 antibody 1D9 is used in immunohistochemistry. This antibody is functional to be used in immunohistochemistry as shown in figure 27. In another preferred embodiment, for detecting Lgr5, the Lgr5 antibody 8F2 is used in immunofluorescence assay. This antibody is functional to be used in immunofluorescence assay as shown in figure 31.

In another preferred embodiment, for detecting Lgr5, any of the Lgr5 antibodies 1D9, 4Dl 1, 3B9 is used in ELISA. These antibodies are functional to be used in ELISA as shown in figure 25. In another preferred embodiment, for detecting Lgr5, any of the Lgr5 antibodies 3B9 and/or 2B8 is used in Western blot. These antibodies are functional to be used in Western blot as shown in figure 21.

For each of these preferred embodiments, the skilled person knows how to carry out each of these techniques. Preferably, each of these techniques is carried out as described in the examples.

In yet a further preferred embodiment, the invention provides an antibody according to the invention for use as a medicament. Said medicament is preferably for treating a subject suspected of having a risk of having a cancer or an individual at risk for cancer.

A preferred antibody according to the invention for use as a medicament is a human, humanized or deimmunised anti Lgr5 and/or Lgr6 antibody as described herein above. In a preferred embodiment said antibody is specific for human Lgr5 and/or human Lgr6. Preferably said antibody is a monoclonal antibody. In one embodiment, said antibody comprises at least one CDR sequence as depicted in Figure 16 or 17. Preferably, said antibody comprises a CDR3 sequence, a CDR2 sequence and a CDRl sequence of a light chain and/or a heavy chain selected from Figure 16 or 17.

In a preferred embodiment, the antibody according to the invention is a bifunctional agent and is linked to a toxic agent or linked to an enzyme capable of converting a prodrug to a toxic agent. Said toxic agent is covalently attached, directly or via a cleavable or non-cleavable linker, to the antibody. In more preferred embodiments, the toxic agent is a taxol, a maytansinoid, a tomaymycin derivative, a leptomycin derivative, CC- 1065 or a CC- 1065 analog. Other examples of drugs that can be used as a toxic agent include, but are not limited to, methotrexate, daunorubicin, doxorubicin, vincristine, vinblastine, melphalan, mitomycin C, chlorambucil, calicheamicin, tubulysin and tubulysin analogs, duocarmycin and duocarmycin analogs, dolastatin and dolastatin analogs. In a further embodiment, the toxic agent is a radiotoxin comprising, for example, strontium-90 or cesium- 137.

A medicament comprising said bifunctional agent is preferably provided together with general anti-cancer therapy. Examples of said general anti-cancer therapy are radiation, chemotherapy, antibody-based therapy or small molecule based treatments. Combined treatment leads to an approach of killing the minority cancer stem cell population as well as the bulk of the tumor. Said general anti-cancer therapy can be provided before, during, or after the provision of a medicament comprising said bifunctional agent. As used herein, the term "cancer" refers to or describe the physiological condition in mammals in which a population of cells are characterized by unregulated cell growth. Examples of cancer include, but are not limited to, carcinoma, lymphoma, blastoma, sarcoma, and leukemia. More particular examples of such cancers include squamous cell cancer, small-cell lung cancer, non-small cell lung cancer, adenocarcinoma of the lung, squamous carcinoma of the lung, cancer of the peritoneum, hepatocellular cancer, gastrointestinal cancer, pancreatic cancer, glioblastoma, cervical cancer, ovarian cancer, liver cancer, bladder cancer, hepatoma, breast cancer, colon cancer, colorectal cancer, endometrial or uterine carcinoma, salivary gland carcinoma, kidney cancer, liver cancer, prostate cancer, vulval cancer, thyroid cancer, hepatic carcinoma and various types of head and neck cancer.

The present invention therefore includes a method for inhibiting the growth of a cancers stem cell expressing Lgr5 and/or Lgr6. In preferred embodiments, the method for inhibiting the growth of the cancer stem cell expressing Lgr5 and/or Lgr6 takes place in vivo and results in the death of the cell, although in vitro and ex vivo applications are also included.

The invention further provides a pharmaceutical composition comprising an antibody according to the invention and a pharmaceutically acceptable carrier or excipient. Preferably, said pharmaceutical composition further comprised a pharmaceutically acceptable excipient, stabilizer, activator, carrier, permeator, propellant, desinfectant, diluent and/or preservative. A pharmaceutical composition may be in any desired form, e.g. a tablet, infusion fluid, capsule, syrup, etc. Formulation of medicaments, and the use of pharmaceutically acceptable excipients are known and customary in the art and for instance described in Remington; The Science and Practice of Pharmacy, 21^nd Edition 2005, University of Sciences in Philadelphia. Ways of administration are known and customary in the art are for instance described in Remington; The Science and Practice of Pharmacy, 21^st Edition 2005, University of Sciences in Philadelphia. A Lgr5 and/or Lgr6 antibody may be formulated to be suitable for intravenous or subcutaneous, or intramuscular administration, although other administration routes can be envisaged, such as mucosal administration or intradermal and/or intracutaneous administration, e.g. by injection. In a preferred embodiment, the invention provides the use of an antibody according to the invention for the manufacture of a medicament for the treatment of cancer. Said treatment preferably further comprises additional general anti-cancer therapy. Examples of said general anti-cancer therapy are radiation, chemotherapy, antibody- based therapy or small molecule based treatments. Combined treatment leads to an approach of killing the minority cancer stem cell population as well as the bulk of the tumor.

The invention also provides use of an Lgr5 and/or 6 antibody according to the invention in the preparation of a diagnostic for the diagnosis of cancer stem cell presence and/or content in a sample, wherein said antibody is conjugated to a substance that allows radioactive imaging, positron emission tomography (PET) scanning, magnetic resonance imaging (MRI) scanning, or X-ray/ computed tomography (CT) scanning. Examples of such substances are radioactive labels such as Indium- 11 , Technetium-99m, Iodine- 131 or Fluorine- 19.

The invention further provides a method for determining whether a body fluid comprises a cancer stem cell, comprising contacting said body fluid with an anti-Lgr5 and/or Lgr6 antibody according to the invention, removing unbound antibody, detecting any bound complex comprising an anti-Lgr5 and/or Lgr6 antibody, and determining the presence of a cancer stem cell based on the presence of detected antibody.

The invention also provides a method for determining cancer stem cell content of a tumor or a body fluid, comprising contacting said tumor or body fluid with an anti- Lgr5 and/or Lgr6 antibody according to the invention, removing unbound antibody and determining whether any bound antibody is present in said tumor or body fluid. In a preferred embodiment, said method is an in vitro method. Said antibody according to the invention is preferably labeled such that it can be identified. Suitable labels are for example a protein (fragment) such as the antibody Fc tail or

Staphylococcal protein A or Glutathion-S-transferase, a short antigenic peptide tag such as the Myc, FLAG or HA tag or an oligomeric Histidine-tag, an enzymatic tag such as Alkaline Phosphatase, a fluorescent protein tag (such as Green Fluorescent Protein). However, it is also possible to use a second compound that has affinity for the binding compound and labeling said second compound with a suitable label (i.e. an indirect analysis). Examples of body fluid are blood, urine, lymph fluid or tears. In a preferred embodiment, said method is an in vitro method.

The described diagnostic methods are also very useful for determining whether an anti cancer therapy leads to eradication of (at least part of the) cancer stem cells. If for example use is made of general anti cancer therapy, optionally combined with an anti-Lgr5 and/or Lgr6 antibody, the effect of said therapy on the cancer stem cell can be determined by determining the presence or absence of cells bearing Lgr5 and/or Lgr6 in the cancer.

In yet another embodiment, the invention provides a method for determining the effectivity of an anti cancer treatment, comprising treating cancer and determining whether cancer stem cells are present comprising contacting said cancer with an anti- Lgr5 and/or Lgr6 antibody according to the invention. This method can be performed in vitro as well as in vivo. It is preferred that the presence of cancer stem cells is determined before treatment and during or after treatment such that it can determined whether or not the applied treatment results in a changed (preferably decreased) amount of cancer stem cells.

The invention further provides a method for treating an individual in need thereof or an indivividual suspected of having cancer or an individual at risk for cancer comprising administering an effective amount of a herein described pharmaceutical composition to said individual and optionally further subjecting said individual to conventional cancer therapy such as radiation or chemotherapy.

As used herein, the term "subject" or "individuaP'or "patient" refers to any animal (e.g., a mammal), including, but not limited to, humans, non-human primates, rodents, and the like, which is to be the recipient of a particular treatment or in which a diagnosis is carried our. Typically, the terms "subject" and "patient" are used interchangeably herein in reference to a human subject.

As used herein, the term "subject suspected of having cancer" refers to a subject that presents one or more symptoms indicative of a cancer (e.g., a noticeable lump or mass) or is being screened for a cancer (e.g., during a routine physical). A subject suspected of having cancer can also have one or more risk factors. A subject suspected of having cancer has generally not been tested for cancer. However, a "subject suspected of having cancer" encompasses an individual who has received an initial diagnosis but for whom the stage of cancer is not known. The term further includes people who once had cancer (e.g., an individual in remission). As used herein, the term "subject at risk for cancer" refers to a subject with one or more risk factors for developing a specific cancer. Risk factors include, but are not limited to, gender, age, genetic predisposition, environmental exposure, previous incidents of cancer, preexisting non-cancer diseases, and lifestyle.

General definitions as to the production of an antibody as identified herein

A nucleic acid encoding an antibody as identified herein may be present on a nucleic acid construct, preferably being a vector. Preferably the vector is a replicative vector comprising an origin of replication (or autonomously replication sequence) that ensures multiplication of the vector in a suitable host for the vector. Alternatively the vector is capable of integrating into a host cell's genome, e.g. through homologous recombination or otherwise. A particularly preferred vector is an expression vector wherein a nucleotide sequence encoding an antibody, is operably linked to a promoter capable of directing expression of the coding sequence in a host cell for the vector.

As used herein, the term "promoter" refers to a nucleic acid fragment that functions to control the transcription of one or more genes, located upstream with respect to the direction of transcription of the transcription initiation site of the gene, and is structurally identified by the presence of a binding site for DNA-dependent RNA polymerase, transcription initiation sites and any other DNA sequences, including, but not limited to transcription factor binding sites, repressor and activator protein binding sites, and any other sequences of nucleotides known to one of skill in the art to act directly or indirectly to regulate the amount of transcription from the promoter. A "constitutive" promoter is a promoter that is active under most physiological and developmental conditions. An "inducible" promoter is a promoter that is regulated depending on physiological or developmental conditions. A "tissue specific" promoter is only active in specific types of differentiated cells/tissues, such as preferably a human and/or tumour and/or mammary cell or tissue derived thereof. An expression vector may allow an antibody as defined above to be prepared using recombinant techniques in which a nucleotide sequence encoding said antibody is expressed in a suitable cell, e.g. cultured cells or cells of a multicellular organism, such as described in Ausubel et al., "Current Protocols in Molecular Biology", Greene Publishing and Wiley-Interscience, New York (1987) and in Sambrook and Russell (2001, supra); both of which are incorporated herein by reference in their entirety. Also see, Kunkel (1985) Proc. Natl. Acad. Sci. 82:488 (describing site directed mutagenesis) and Roberts et al. (1987) Nature 328:731-734 or Wells, J.A., et al. (1985) Gene 34: 315 (describing cassette mutagenesis). Typically, a nucleic acid encoding said antibody is used in an expression vector.

The phrase "expression vector" generally refers to nucleotide sequences that are capable of effecting expression of a gene in hosts compatible with such sequences. These expression vectors typically include at least suitable promoter sequences and optionally, transcription termination signals. Additional factors necessary or helpful in effecting expression can also be used as described herein. A nucleic acid or DNA encoding said antibody is incorporated into a DNA construct capable of introduction into and expression in an in vitro cell culture. Specifically, DNA constructs are suitable for replication in a cultured mammalian, plant, insect, e.g., Sf9, yeast, fungi or other eukaryotic cell lines. DNA constructs prepared for introduction into a particular host typically include a replication system recognized by the host, the intended DNA segment encoding the desired antibody, and transcriptional and translational initiation and termination regulatory sequences operably linked to the antibody-encoding segment. A DNA segment is "operably linked" when it is placed into a functional relationship with another DNA segment. For example, a promoter or enhancer is operably linked to a coding sequence if it stimulates the transcription of the sequence. DNA for a signal sequence is operably linked to DNA encoding a polypeptide if it is expressed as a pre protein that participates in the secretion of said polypeptide. Generally, DNA sequences that are operably linked are contiguous, and, in the case of a signal sequence, both contiguous and in reading phase. However, enhancers need not be contiguous with the coding sequences whose transcription they control. Linking is accomplished by ligation at convenient restriction sites or at adapters or linkers inserted in lieu thereof. The selection of an appropriate promoter sequence generally depends upon the host cell selected for the expression of the DNA segment. Examples of suitable promoter sequences include prokaryotic, and eukaryotic promoters well known in the art (see, e.g. Sambrook and Russell, 2001, supra). The transcriptional regulatory sequences typically include a heterologous enhancer or promoter that is recognised by the host. The selection of an appropriate promoter depends upon the host, but promoters such as the trp, lac and phage promoters, tRNA promoters and glycolytic enzyme promoters are known and available (see, e.g. Sambrook and Russell, 2001, supra). Expression vectors include the replication system and transcriptional and translational regulatory sequences together with the insertion site for the polypeptide encoding segment can be employed. Examples of workable combinations of cell lines and expression vectors are described in Sambrook and Russell (2001, supra) and in Metzger et al. (1988) Nature 334: 31-36. For example, suitable expression vectors can be expressed in, yeast, e.g. S.cerevisiae, e.g., insect cells, e.g., Sf9 cells, mammalian cells, e.g., CHO cells and bacterial cells, e.g., E. coli. The host cells may thus be prokaryotic or eukarotic host cells. A host cell may be a host cell that is suitable for culture in liquid or on solid media. A host cell is preferably used in a method for producing an antibody of the invention as defined above. Said method may comprise the step of culturing a host cell under conditions conducive to the expression of said antibody. Optionally the method may comprise recovery of said antibody. An antibody may e.g. be recovered from the culture medium by standard protein purification techniques, including a variety of chromatography methods known in the art per se.

In this document and in its claims, the verb "to comprise" and its conjugations is used in its non-limiting sense to mean that items following the word are included, but items not specifically mentioned are not excluded. In addition the verb "to consist" may be replaced by "to consist essentially of meaning that an antibody as defined herein may comprise additional component(s) than the ones specifically identified, said additional component(s) not altering the unique characteristic of the invention. In addition, reference to an element by the indefinite article "a" or "an" does not exclude the possibility that more than one of the element is present, unless the context clearly requires that there be one and only one of the elements. The indefinite article "a" or "an" thus usually means "at least one". The word "approximately" or "about" when used in association with a numerical value (approximately 10, about 10) preferably means that the value may be the given value of 10 more or less 1% of the value.

All patent and literature references cited in the present specification are hereby incorporated by reference in their entirety.

The following examples are offered for illustrative purposes only, and are not intended to limit the scope of the present invention in any way.

Legends to the figures

Figure 1. Restricted expression of a GPR49-LacZ reporter gene in adult mice. Expression of GPR491acZ in selected adult mouse tissues. LGR5 is restricted to rare cell populations in the (A) adrenal gland (black circle), (B) the eye (inner nuclear layer of the retina, see arrows), (C) brain (glomeruli of the olfactory bulb, see arrows and several other poorly defined regions) and (D) liver (cells surrounding the portal triads, see arrow)

Figure 2. Lineage tracing in the stomach Lgr5-EGFP-CreERT2 mice were crossed with Rosa26R reporter mice and Cre enzyme activity induced in the LGR5+ve cells by IP injection of Tamoxifen. LacZ reporter gene activity is initially restricted to the LGR5⁺ cells (see circles) (A), but rapidly expands to include the entire epithelium in the stomach over time (see circle) (B). This "lineage tracing" is maintained over long periods of time. This demonstrates that all epithelial cells are derived from the LGR5⁺ population in this tissue, proving that they are stem cells.

Figure 3. Lineage tracing in the mammary gland Lgr5-EGFP-CreERT2 mice were crossed with Rosa26R reporter mice and Cre enzyme activity induced in the LGR5+ve cells by IP injection of Tamoxifen. LacZ reporter gene activity is initially restricted to the LGR5 cells (see circle) (A), but expands to include the myoepithelium of newly- formed milk glands in lactating females (as indicated by dark grey cells lining the epithelium) (B), indicating that LGR5 is specifically marking myoepithelial stem cells in this organ.

Figure 4. Lineage tracing in the adrenal gland. Lgr5-EGFP-CreERT2 mice were crossed with Rosa26R reporter mice and Cre enzyme activity induced in the LGR5+ve cells by IP injection of Tamoxifen. LacZ reporter gene activity is initially restricted to the LGR5 cells (see circles) (A), but expands to include the medulla of the adrenal gland (LGR5⁺ cells are dark grey) (B), indicating that LGR5 is specifically marking adrenal medulla stem cells.

Figure 5. Lgr6 is expressed in cells of the upper bulge area of the mouse hair follicle and in basal cells of the epidermis. Skin sections of appr. 26 days old Lgr6-EGFP- Ires-CreERT2 mice (early anagen) were obtained and stained for nuclear DNA (Topro, light grey cells) and EGFP (bright white cells) visualized using confocal microscopy (A-C). During early anagen Lgr6 is expressed in the upper bulge (A, C) and the basal epidermis (A, B). Examples of Topro and GFP positive cells are indicated by circles and arrows.

Figure 6. The progeny of Lgr6+ cells contribute to all structures of the hair follicles (HF), interfollicular epidermis (IFE) and sebaceous glands (SG). To trace the progeny of Lgr6+cells Lgr6-EGFP-Ires-CreERT2/ROSA26-LacZ mice were injected with tamoxifen (TM) at P20 when HFs are in telogen (A). At P23 a first staining in the IFE and HFs was detected (see circles) (B). Analysis of LacZ staining progeny at P38 (1st anagen, C, D) and P52 (2nd telogen, E, F) revealed contribution to all parts of the HFs, IFEs and SGs (LGR6⁺ cells are dark grey).

Figure 7. The progeny of Lgr6+ cells contribute to the myoepithelium of the lung. To trace the progeny of Lgr6+cells Lgr6-EGFP-Ires-CreERT2/ROSA26-LacZ mice were injected with tamoxifen (TM) at P20. Analysis of LacZ staining progeny at P38 (A, 1Ox, 2Ox and 4Ox magnification from left to right) and P52 (B, 1Ox, 2Ox and 4Ox magnification from left to right) revealed contribution to the myoepithelium underlying the bronchioles of the lung (examples of LGR6⁺ cells are indicated by circles and arrows).

Figure 8. Low-dose oral induction with β-NF does not induce Cre-mediated deletion in stem cells of AHCre mice. Intestinal whole-mounts stained for β-galactosidase from AhCre+ Rosa26R+ mice, a: No activation of the Rosa-lacZ reporter gene is observed in intestines from non- induced AhCre+ Rosa26R+ mice, b: Readily visible expression of lacZ throughout the intestine 2 days after a single gavage of lmg/kg β- napthoflavone, indicating efficient Cre-mediated activation of the lacZ reporter. No lacZ expression is visible at the crypt base (lower panel) demonstrating the absence of Cre-mediated recombination at the crypt base, c: No lacZ-positive crypt/villus units are visible on whole-mount intestines 100 days post-induction, indicating that this dosing regime very rarely causes recombination within the intestinal stem cells. Figure 9. Transformation of non-stem cells through loss of APC does not efficiently drive adenoma formation over extended time-periods, a-c: β-catenin IHC performed on intestinal sections from AhCre+ Rosa26R+ Apcfl/fl 3 days following a single gavage of 1.0mg/kg β-napthoflavone. Clusters of transformed cells with nuclear β- catenin were frequently observed on the villus (a) and upper regions of the crypt (b). β-catenin^hlgh clusters were only very rarely observed at the crypt base (c). These clusters are highlighted with black arrows, d: Quantification of the location of the β- catenin^hlgh cell clusters on intestinal sections from AhCre+ Rosa26R+ Apcfl/fl 4 days following a single gavage of 1.0mg/kg β-napthoflavone. Box-plots showing numbers of foci observed at the crypt base, the upper crypt and the villus in 1600 crypt-villus units. Significantly more clusters were seen at the upper regions of the crypt than any other region (p=0.04, Mann Whitney, n=3). Nuclear β-catenin foci were observed only very rarely at the crypt base. e:α- β-catenin IHC performed on intestinal section from AhCre+ Rosa26R+ Apcfl/fl 24 days following a single gavage of 1.0mg/kg β- napthoflavone. Here, nuclear β-catenin is seen in a small lesion 24 days after ere induction (see circle). f,g: β-catenin IHC performed on intestinal section from AhCre+ Rosa26R+ Apcfl/fl 167 days following a single gavage of 1.0mg/kg β- napthoflavone showing a microadenoma (f) and small adenoma (g) with nuclear β- catenin. h: Quantification of adenoma formation over extended time-periods in AhCre+ Rosa26R+ Apcfl/fl following a single gavage of 1.Omg/kg β-napthoflavone. Lesion size was scored on intestinal whole-mounts from AhCre+ Rosa26R+ Apcfl/fl mice that had been stained for lacZ to help visualise the small lesions (at least 3 mice were used for each time-point). No adenomas were seen in mice up to and including day 24 and there was only the very rare microadenoma in mice at day 24. The occasional adenoma was observed in AhCre+ Rosa26R+ Apcfl/fl at 100 days (plus), however the majority of lesion remained microscopic showing that most lesions were not progressing to adenoma despite a long latency period.

Figure 10. Lgr5+ve intestinal stem cells transformed following loss of APC persist and fuel the rapid formation of β-catenin^hlgh microadenomas, a-i: The consequences of Lgr5+ve intestinal stem cell transformation and their subsequent fate was tracked over an eight day period using β-catenin and GFP as markers of transformed cells and Lgr5+ve stem cells respectively, a-c: Accumulation of the Wnt effector, β- catenin is first observed in scattered Lgr5+ve stem cells 3 days after Cre induction in Lgr5-EGFP-Ires-CreERT2/APCfl/fl intestines. Representative examples of β- catenin^hlgh Lgr5+ve stem cells are circled, d-f : Five days post-induction the transformed Lgr5-GFP+ve stem cells remain (e,f: black arrows) and are associated with clusters of transformed (β-catenin^hlgh) cells within the TA compartment, g-h: Eight days post-induction the clusters of transformed cells have expanded to fill the TA compartment (h: black circle). The transformed Lgr5-GFP+ve stem cells at the crypt base persist (h,i: black arrows), but their transformed progeny within the TA compartment are Lgr5-GFP^neg (h,i: black circles).

Figure 11. Selective transformation of Lgr5⁺ stem cells following loss of APC efficiently drives adenoma formation throughout the small intestine, a-h: The appearance and development of intestinal adenomas and the expression of the Lgr5- GFP stem cell marker within these adenomas was tracked over a 36 day period using GFP (f) and β-catenin (all others) IHC. a-b: Multiple small adenomas are readily visible throughout the intestine 14 days after Lgr5+ve stem cell transformation (gathering of dark grey cells), c-f: Multiple macroscopic adenomas (> 100) are present after 24 days. Lgr5-GFP expression in adenomas is restricted to rare scattered cells (f; circled). g,h: At 36 days, a large proportion of the intestine is filled with macroscopic adenomas.

Figure 12. Presence of Lgr5+ stem cells in intestinal adenomas. Intestinal adenomas express high levels of β-catenin (dark grey staining) as a result of chronic activation of the Wnt pathway (A). In contrast to other Wnt target genes which are highly expressed throughout the adenoma (not shown), expression of the intestinal stem cell marker Lgr5-GFP is restricted to scattered cells with characteristic stem cell morphology: slender, comma-shaped cells; An example is indicated with black arrow (B). We speculate that these Lgr5+ve cells within the adenoma are stem cells dedicated to maintaining the growth of the adenoma (so-called cancer stem cells).

Figure 13. FACS analyses of LGR5 expression in L8 cells, which are clonal derivatives of LS174T cells, which express dominant negative Tcf4 (DNTcf4) upon Doxycycline (DOX). DNTcf4 turns off constitutive active Wnt pathway. After 48hrs of DOX induction, a reduction in hLgr5 protein levels is observed. Rat IgG is used as negative isotype control. 9G5 is a rat monoclonal derived antibody directed against hLgr5

Figure 14. Endogenous hLgr5 staining of a human colon cancer cell line (L8) using several Lgr5 -specific monoclonal antibodies. L8 cells are a clonal derivative of the parental LS174T cell- line. Following Doxycycline (DOX) induction the L8 cells express a dominant-negative form of Tcf-4 (DNT cf4). DNTcf4 efficiently blocks the constitutive Wnt pathway activity in these cells and consequently switches off Tcf target genes. After 48hrs of DOX induction a major reduction in hLgr5 protein levels is observed. Rat IgG is used as negative isotype control.

Figure 15. Light chain + heavy chain sequences analyzed using KABAT method. CDR regions are in bold and underlined.

Figure 16. Alignment of sequences of the variable region of antibodies specific for Lgr5. Alignment was performed using the CLUST AL W2 algorithm http://www.ebi.ac.uk/Tools/clustalw2/index.html). A. V_H sequence alignment of 11 different Lgr5 specific antibody clones. B. V_L sequence alignment of 11 different Lgr5 specific antibody clones. The extension of the name indicates the subclone of which the sequence was obtained.

Figure 17. Alignment of sequences of the variable region of antibodies specific for Lgr6. Alignment was performed using the CLUST AL W2 algorithm http://www.ebi.ac.uk/Tools/clustalw2/index.html). A. V_H sequence alignment of 4 different Lgr6 specific antibody clones. B. V_L sequence alignment of 4 different Lgr6 specific antibody clones. The extension of the name indicates the subclone of which the sequence was obtained.

Figure 18. Alignment of nucleotide sequences of the V_H region of antibodies specific for Lgr5 for the different groups containing antibodies that likely recognize the same epitope on Lgr5 (see Table 4). A. Alignment of V_H region nucleotide sequences of Lgr5 specific antibodies clustered in group 1 (Table 4). B. Alignment of V_H region nucleotide sequences of Lgr5 specific antibodies clustered in group 2 (Table 4). C. Alignment of V_H region nucleotide sequences of Lgr5 specific antibodies clustered in group 3 (Table 4). The extension of the name indicates the subclone of which the sequence was obtained.

Figure 19. Alignment of nucleotide sequences of the V_H region of antibodies specific for Lgr6 clustered in group 4 that likely recognize the same epitope on Lgr6 (see Table 5). The extension of the name indicates the subclone of which the sequence was obtained.

Figure 20. Expression of endogenous hLgr5 measured by FACS. Lgr5 specific antibody 3B9 was used to stain epithelial cells derived from wt crypt or Villin hLgr5 crypts.

Figure 21 A. 3B9 was able to detect hLgr5 and not mLGR5 after transfection of mLGR5 or hLGR5 using western blot. B. Lgr5 specific antibody clone 2B8 was able to detect hLgr5 in cells transfected with hLgr5.

Figure 22 A. Alignment of nucleotide sequences of the V_H region of antibodies specific for Lgr5. B. Alignment of nucleotide sequences of the V_L region of antibodies specific for Lgr5. Alignment was performed using the CLUSTAL W2 algorithm (http://www.ebi.ac.uk/Tools/clustalw2/index.htm).

Figure 23 A. Alignment of nucleotide sequences of the V_H region of antibodies specific for Lgr6. B. Alignment of nucleotide sequences of the V_L region of antibodies specific for Lgr6. Alignment was performed using the CLUSTAL W2 algorithm (http://www.ebi.ac.uk/Tools/clustalw2/index.html).

Figure 24. Schematic representation of the LGR5 protein. LGR5 is a G-coupled, seven-transmembrane receptor encompassing a signal peptide (aa: 1-21), N-Terminal Region (N-term, aa: 22-70), 17 Leucine-Rich Repeats (LRR, aa: 71-473), C-Flanking cysteine rich linker region (CRL, aa: 474-561), Transmembrane region (TM, aa: 562- 825) and C-terminal region (C-term, 826-907).

Figure 25. ELISA showing the binding of LGR5 antibodies present in hybridoma supernatants (diluted 1 :1 in PB S/ 10% FCS) to the LGR5-Fc fusion protein. Negative control 1 (Neg con 1) is not coated with LGR5-Fc fusion and has been incubated with rabbit anti hulgG/HRP. Negative control 2 (Neg con 2) is not coated with LGR5-Fc and has been incubated with Goat anti-Rat IgG/HRP. Negative control 3 (Neg con 3) is coated with LGR5-Fc and has been incubated with Goat Anti-Rat IgG/HRP. The positive control is coated with LGR5-Fc and has been incubated with Rabbit anti hulgG/HRP. The hybridoma supernatants have been detected with goat anti-Rat IgG/HRP.

Figure 26. Schematic representation of the various LGR5 (deletion) constructs and the extracellular LGR5-Fc fusion. The major protein domains expressed by the various constructs are indicated. The leader peptide/signal peptide is not indicated since this will be removed when the protein is properly expressed.

Figure 27. Immunochemical staining of fixed COS cells transfected with a human flag tagged full length LGR5. When cells are incubated without a first antibody no staining is observed (A). Anti-flag (B) and hybridoma supernatant 1D9 (C) as first antibody result in the same specific staining.

Figure 28. Overlay plot of the LGR5-Fc injection (resulting in a shift of the refractive index) followed by the injection of monoclonal LGR5 antibody 5A7 (upper line) and an irrelevant antibody (lower line) using label free detection in Surface Plasmon Resonance (SPR) imaging. The addition of 5A7 results in an extra shift indicating a specific interaction between the 5A7 antibody and the extracellular LGR5 domain, while the irrelevant antibody does not show an effect. Data has been normalized against a reference curve.

Figure 29. In order to cluster the antibodies in independent binding groups, competition analysis was carried out using label free detection in Surface Plasmon Resonance (SPR) imaging. Hereby the effect of change in R is measured for two antibodies successively loaded on the LGR5-Fc-coated SPR sensor. Overlay plot of the LGR5-Fc injection followed by the injection of antibody 5C8 and subsequently 4Dl 1 (upper line) demonstrated an increase in R after each addition, suggesting binding of both antibodies to different regions of the extracellular domain of the LGR5 protein. The addition of 7Bl 1 (lower line) followed by 4Dl 1 does not show a second increase in R, suggesting that these two antibodies bind to the same region of the LGR5 protein. Data has been normalized against a reference curve. Figure 30. Overlay plot of the LGR5-Fc injection followed by subsequent antibody injections shows that the subsequent injection of 3 antibodies from three different clusters (6C10, 4Dl 1 and 1D9) result in an additive signal in the SRP assay. Data has been normalized against a reference curve.

Figure 31. Human colon organoid culture stained with DAPI (A), bright field (B) or with anti LGR5 hybridoma 8F2 (C) visualizes endogenous LGR5 positive cells in the budding regions of the human organoid culture.

Examples

Example 1

Lgr5 tissue expression and evidence for Lgr5+ stem cells in these tissues

Materials and Methods

Northern blotting and induced Wnt pathway inhibition in LS174T clone L8: As in van de Wetering, M. et al. The beta-catenin/TCF-4 complex imposes a crypt progenitor phenotype on colorectal cancer cells. (Cell 111, 241-50 (2002)). The probe spanned the entire reading frame of mouse Gpr49. Crypt and villus epithelial preparations for RNA isolation were generated from 0.5cm lengths of intestine by 4 successive rounds of incubation in pre-warmed 3OmM EDTA at 37°C for 10 minutes, followed by vigorous shaking (1Ox) in ice-cold PBS. Fractions 1 and 4, comprising predominantly villi and crypts respectively were used for RNA isolation. Mice: GPR49-LacZ mice were generated by homologous recombination in ES cells targeting an Ires-LacZ cassette to the 5' end of the last exon, essentially removing the region containing all TM regions and creating a null allele (Lexicon). GPR49-EGFP- Ires-CreERT2 mice were generated by homologous recombination in ES cells targeting an EGFP-Ires-CreERT2 cassette to the ATG of GPR49. Rosa26-lacZ Cre reporter mice were obtained from Jackson Labs.

Tamoxifen induction: Mice of at least 8 weeks of age were injected once intraperitoneally with 200 μl of Tamoxifen in sunflower oil at 10 mg/ml. BrdU injection: Mice were injected intraperitoneally at four hour intervals with 200 μl of a BrdU solution in PBS at 5 mg/ml. Immuno Electron Microscopy: Intestines were dissected and perfuse-fϊxed in 4% PFA in 0.2 M PHEM-buffer, embedded in gelatin, cryosectioned with a Leica FCS cryoultratome and immuno labelled against GFP with polyclonal rabbit anti-GFP antibody. Samples were trimmed using a diamond Cryotrim 90 knife at -100 ⁰C (Diatome, Switzerland) and ultrathin sections of 70 nm were cut at -120 ⁰C using a Cryoimmuno knife (Diatome, Switzerland). For the low magnification EM images the 15 nm protein A-gold particles (UMCU, Utrecht, The Netherlands) were briefly silver enhanced with R-GENT SE-EM (Aurion, The Netherlands) according to the manufacturers instructions. Non-specific binding to Paneth cell granules was diminished by applying Blocking solution (Aurion, The Netherlands) prior to the primary antibody.

Tissue sample preparation for immunohistochemistry, in-situ hybridization and LacZ expression analysis: All performed as previously described in Muncan, V. et al. Rapid loss of intestinal crypts upon conditional deletion of the Wnt/Tcf-4 target gene c-Myc. MoI Cell Biol 26, 8418-26 (2006). In-situ probes comprising a lkb N- terminal fragment of mGPR49 were generated from sequence-verified Image Clone 30873333. Ki67 antibodies were purchased from Monosan (The Netherlands), Phospho-histone H3 from Campro Scientific (The Netherlands), anti-synaptophysin from Dako, anti BrdU from Roche. Polyclonal rabbit anti-GFP was provided by Edwin Cuppen, Hubrecht Institute.

Generation of suspension of human (tumor) tissue cells.

Using a razor blade, mince freshly isolated human (tumor) tissue as much as possible. Do this in serum-free media. Draw minced tumor into a 25ml pipette. Place the solution into a 50ml conical tube. Incubate at 37C for 30-60 min after adding collagenase IV (200units/ml) (Sigma). The final concentration should be 200units/ml. Pipette up and down a few times every lOmin (appro x). Pass the solution through a filter (45 micrometer pore size; Becton Dickinson). Wash the filter with 4-5 ml of serum- free medium. Centrifuge the solution @ 1500rpm for lOmin (4°C) Resuspend the pellet in hypotonic ammonium chloride (approx. 5 ml). Leave 10 min at room temperature (this will lyse red blood cells). Then add equal volume of serum- free media and centrifuge again. Resuspend pellet in serum free medium. If clumpy then pass through another filter. Count with trypan blue to see the percent dead cells.

Cancer stem cell assay by xenografting in immunodeficient mice The mice are sublethally irradiated with 320 Rad. The experimental procedure involves injecting human (colon) cancer cell suspensions under the renal capsule of NOD/SCID mice. The mice are handled using sterile techniques and anaesthetized using inhalational anaesthesia: isoflurane. The mice are placed on a heating pad during the procedure.

A clipper is used to shave the abdomen, which is then prepped sequentially with: (1) iodine based solution and (2) 70% ethanol solution. The area is then dabbed with a gauze. The mouse is placed on its side (left side up). A 1 cm (approximately) flank incision is made with scissors, just below the costal margin on the left side. Deliver the kidney into the wound. The cell suspension to be assayed for cancer stem cell activity is mixed 1 :1 (medium : Matrigel) on ice. Utilizing a tuberculin syringe, inject 25 microliter of the cell suspension under the renal capsule. Deliver the kidney back into the abdomen. If cancer stem cell activity is present in the cell suspension, a tumor will grow out in the subsequent weeks/months which is analysed by histology and should resemble the original human tumor.

Results and Discussion We studied the Lgr5 expression in multiple other tissues in the mice carrying lacZ integrated into the last exon of the Gpr49 gene, removing all transmembrane regions of the encoded Gpr49 protein. We determined that, analogous to colon and small intestine, Lgr5+ cells were detected in brain, retina, liver and adrenal gland (Fig 1). In adult mice, LGR5 is restricted to rare cell populations in the brain (glomeruli of the olfactory bulb and several other poorly defined regions), the eye (inner nuclear layer of the retina), liver (cells surrounding the portal triads) and adrenal gland.

Lineage tracing in the stomach, mammary gland and adrenal gland proves that Lgr5 is marking stem cell populations in these tissues. We used the LGR5KI/Rosa26-lacZ mice 16 (See example 1 for experimental strategy) to study the presence of Lgr5+ stem cells in multiple other tissues. Injection of Tamoxifen activates the CreERT2 fusion enzyme in Gpr49-expressing cells. Cre-mediated excision of the roadblock sequence in the Rosa26-LacZ reporter should then irreversibly mark the Gpr49+ve cells. Moreover, while potential progeny of these cells will no longer express GFP, the activated LacZ reporter should act as a permanent genetic mark, which will be passed on to any descendents of the LGR5+ve cells, allowing us to track their appearance and fate in- vivo.

Lineage tracing was initiated in young LGR5KI/Rosa26-lacZ mice and the stomach epithelium analyzed for LacZ activity after 6 months. LGR5-lacZ positive cells are initially restricted to the base of the glands (Fig. 2a). After 6 months, multiple entirely lacZ-positive glands are visible throughout the stomach (Fig. 2b), demonstrating that the LGR5+ve cells are capable of generating all cell-types on the glandular epithelium over long periods of time. Similar lineage tracing experiments were performed and the mammary gland epithelium analyzed for LacZ activity over a 3 month period. LGR5-lacZ positive cells are initially restricted to rare basal epithelial cells on virgin glands (Fig. 3a). Following pregnancy, LacZ-positive cells are visible in the basal epithelium of the newly-formed milk glands (Fig. 3b). This demonstrates that LGR5+ cells in the mammary gland are myoepithelial stem cells.

Lineage tracing in the adrenal glands analyzed for LacZ activity over a 3 month period. LGR5-lacZ positive cells are initially restricted to the periphery of the adrenal gland 5 days after induction (Fig. 4a). After 3 months the majority of the adrenal medulla is LacZ positive (Fig 4b). This remains positive over a 14 month period (not shown). This demonstrates that the LGR5+ cells are the stem cells of the adrenal medulla.

Example 2 Lgr6 tissue expression and Lgr6 expression in related stem cells

Material and Methods

Transgenic mice and treatments.

Lgr6-EGFP-Ires-CreERT2 mice were generated by homologous recombination in embryonic stem cells targeting an EGFP-Ires-CreERT2 cassette to the ATG of Lgr6. Rosa26-LacZ reporter mice were obtained from the Jackson laboratory. Mice were fed ad libitum. The Cre recombinase was activated in Lgr6-EGFP-Ires-

CreERT2/Rosa26-LacZ mice by injecting 200 μl of tamoxifen (10 mg/ml dissolved in sunflower oil) intraperitoneally.

Confocal analysis of EGFP expression

For confocal imaging the skin samples were fixed in formalin for 15 minutes at RT and embedded in 4% low melting agarose. Longitudinal sections between 100 and 200 μm thick were prepared using a vibratome. Sections were then permeabilized in PBS supplemented with 1% BSA + 1% DMSO + 0.1% TritonX, stained for 30 minutes with TO-PRO 1 : 1000 dilution (Molecular Probes) and embedded using Vectashield (Vector Labs). Sections were imaged with a Sp2 confocal microscope (Leica) and processed using Volocity and Photoshop CS2 software. Detection of beta-galactosidase activity

Freshly obtained tissues were fixed for 2 hours in 1% Formaldehyde/0.2% glutaraldehyde/0.02% NP40 in PBSO solution at 4°C on a rolling platform. Samples were washed 3 times for 20 min with rinse buffer (2 mM MgC12/0.02% NP40/PBS0) and stained for 36-48 h in a solution consisting of 1 mg/ml X-gal, 5 mM ferrothiocyanide, 5 mM ferrithiocyanide, 0.1% sodium deoxycholate in rinse buffer. The substrate was removed and the samples washed twice in PBSO for 20 min at room temperature on a rolling platform. The tissues were then fixed overnight in 4% PFA in PBSO at 4°C in the dark on a rolling platform. The PFA was removed and the tissues washed twice in PBSO for 20 min at room temperature. The samples were embedded in paraffin, sectioned (4 μm) and counterstained with neutral red.

Results and Discussion

To characterize the expression of Lgr6 in the skin we utilized a knock- in mouse, where the Lgr6 promoter controls the expression of EGFP and the CreERT2 fusion protein, termed Lgr6-EGFP-Ires-CreERT2. At P25 when the hair follicles (HFs) are in the growing (anagen) phase, the GFP -positive cells were localized to cells of the upper bulge/isthmus area of the HF (Fig 5A, C) and basal cells of the inter follicular epidermis (IFE, Fig 5A, B). This expression pattern suggests that Lgr6 expression marks a SC/early progenitor cell population of the hair follicle and the epidermis.

To address the question whether the Lgr6+ cells of the anagen HF and IFE represent functional stem cells 20 day-old Lgr6-EGFP-Ires-CreERT2/Rosa26-LacZ mice were injected with tamoxifen. At P20 Lgr6 is expressed in the upper bulge/isthmus area of the HF and basal cells of the IFE (data not shown). Three days post tamoxifen injection a scattered pattern of labeled cells could be seen in the HFs and the IFE (Fig. 6B). At 18 days post-injection the progeny of Lgr6+ cells could be seen in the anagen HFs (Fig. 6C, D) as well as in the IFE and the sebaceous glands (SG) (Fig 6C, D). In the next telogen labeled cells were found in the bulge and isthmus of the HFs (Fig. 6E, F) and the IFE and SGs (Fig. 6E, F). This observation strongly suggests that Lgr6+ cells located in the bulge/isthmus area of the HF and the basal IFE exhibit stem cell properties. In particular, Lgr6+ cells can contribute to all the appendages of the skin, i.e. the growing HFs, the IFE and the SG. It seems rather unique that adult stem cells can be identified on the basis of expression of a single gene, in this case Lgr6. This phenomenon may not be restricted to the skin, because we observe highly restricted expression of Lgr6 in a variety of other tissues. To address the question whether the Lgr6+ cells represent functional stem cells in any other tissues 20 day-old Lgr6-EGFP-Ires-CreERT2/Rosa26-LacZ mice were injected with tamoxifen. LacZ staining was performed on 18 and 32 days post tamoxifen injection to assess for lineage tracing in a variety of tissues.

Interestingly, LacZ positive cells were present in the myoepithelium underlying the bronchioles of the lung at both timepoints (Fig. 7). Thus, Lgr6+ cells contribute to the myoepithelium of the lung strongly suggesting that Lgr6+ cells located in the lung exhibit stem cell properties as well.

Example 3 The role of Lgr5+ cancer stem cells in adenoma

Material and Methods

Transgenic mice and treatments.

Ah-cre/Apcflox/flox mice were generated by interbreeding mice carrying a floxed Ape allele (Apc580S/flox) (Shibata et al. Science 1997. 278: 120-123) and the Ah- cre allele (Ireland et al. Dev. Dyn. 2005. 233: 1332-1336). Lgr5-EGFP-IRES- creERT2/APCflox/flox mice were generated by interbreeding mice carrying a floxed Ape allele, (Apc580S/fiox) and the Lgr5-EGFP-IRES-creERT2 allele (Barker et al. Nature 2007. 449: 1003-1007).

Cre was acivated in Ah-cre/Apcflox/flox mice through oral β-NF induction. Mice aged 6-8 weeks were treated with a single oral gavage of lmg/kg β-naphthoflavone (β-NF) in corn oil. The Cre recombinase was activated in Lgr5-EGFP-IRES- creERT2/APCflox/flox mice by injecting 200μl of tamoxifen in sunflower oil at lOmg/ml.

Tissue fixation and immunohistochemistry

Intestinal tissue was fixed and processed into paraffin blocks according to standard procedures, β-catenin, EGFP, c-Myc, CD44 and Ki67 immunohistochemistry was performed as previously described (van de Wetering et al. 2002 Cell 111 : 241-250). LacZ analysis was done as described in example 1.

Results and discussion The anatomy of the intestinal crypt is uniquely suited to study adult stem cells in their niche. The epithelium of the murine small intestine renews every five days (Barker et al. Gen Dev 2008. 22: 1856-1864; Potten. Bull Cancer 62, 419-30, 1975). Vigorous proliferation occurs within the crypt compartment. We have recently identified slender, undifferentiated cells expressing the Lgr5 gene located at crypt bottoms as the stem cells of the small intestine and colon. Each small intestinal crypt contains approximately 6 independent, long-lived stem cells that are intermingled with Paneth cells in the small intestine and with goblet cells in the colon. Counter-intuitively, these cells are not quiescent, but complete a cell cycle every day (Barker et al. Nature 449, 1003-7, 2007). Leblond and colleagues have originally named these cells morphologically Crypt Base Columnar (CBC) cells (Cheng & Leblond. Am J Anat 141, 537-61 (1974); Cheng & Leblond. Am J Anat 141, 461-79 (1974)). Their daughter cells constitute the readily distinguishable transit amplifying (TA) crypt compartment. TA cells divide every 12-16 hours, generating some 300 cells per crypt every day (Marshman et al. Bioessays 24, 91-8 (2002)). Newly- formed TA cells reside within crypts for approximately 48-72 hours, undergoing up to 6 rounds of cell division while migrating upwards (Marshman et al. Bioessays 24, 91-8 (2002)). When the committed TA cells reach the crypt-villus junction, they rapidly and irreversibly differentiate. The proliferation is balanced by apoptosis at the other end of the epithelial conveyor belt, the tip of the villus. Only Paneth cells escape this flow; they have a residence time of 3-6 weeks at the crypt base (Bjerknes & Cheng. Am J Anat 160, 65-75 (1981); Bjerknes & Cheng. J Anat 160, 51-63 (1981); Ireland et al. Dev Dyn 233, 1332-6 (2005)).

Initiating mutation in intestinal malignancies in mouse and man target components of the Wnt pathway, most frequently the negative Wnt regulator APC (Jones et al. Proc Natl Acad Sci U S A 105, 4283-8 (2008); Kinzler & Vogelstein. Cell 87, 159-70 (1996)). This results in the constitutive activation of a Wnt target gene program that drives the formation of benign adenomas or polyps (Korinek et al. Science 275, 1784-7 (1997); Morin et al. Science 275, 1787-90 (1997); van de Wetering et al. Cell 111, 241-50 (2002); Van der Flier et al. Gastroenterology 132, 628-32 (2007)). However, it remains unclear which cell type sustains the cancer- initiating mutation.

The Cytochrome P450-promoter-driven AH-Cre mouse allows conditional deletion of floxed alleles in the intestinal epithelium following administration of the inducing agent, β-Napthoflavone (β-NF). Importantly, the AH-Cre allele is highly active in all cell types of the epithelium, including the stem cells (Sansom et al. Genes Dev 18, 1385-90 (2004)). We have previously employed a floxed allele of APC (Shibata et al. Science 278, 120-3 (1997)) in combination with the AH-Cre mouse line to demonstrate that acute loss of APC throughout the adult intestinal epithelium following IP injection of β-NF leads to an immediate quantitative transformation of the epithelium (Sansom et al. Genes Dev 18, 1385-90 (2004), a process almost entirely dependent on the downstream Wnt target gene c-Myc (Sansom et al. Nature 446, 676-9 (2007)). High-dose oral β-NF induces more stochastic deletion of APC, resulting in rapid adenoma formation throughout the intestine within 3 weeks (Sansom et al. J Biol Chem 280, 28463-7 (2005)). Both these high-dose induction protocols effect deletion in all compartments of the epithelium, including the stem cells at the crypt base.

Having validated the AHCre/APCflox/flox mouse as an inducible model of intestinal cancer, we sought to dissect the mechanism of adenoma formation by identifying its cell-of-origin. We reasoned that oral administration of low-dose β-NF would restrict its range of action to cells on the villi and the upper regions of the crypts. Careful titration of the required dosage revealed that following oral administration of lmg/kg β-NF, the efficiency of Cre activation in the stem cells at the crypt base was extremely low, as measured by the negligible frequency of long-term lineage tracing initiated in AHCre/R26R mice receiving this dose. This dose was still very efficient in inducing Cre activity in the TA compartment and villus epithelium, as detected using the Rosa26-LacZ mouse (Soriano. Nat Genet 21, 70-1 (1999)) as a Cre reporter (Fig. 8a, b). In a typical experiment over 70% of villi contained blue cells 2 days after induction, but at day 7 blue staining could no longer be detected. In line with this, no crypt/villus ribbons were detected at day 100 post-oral induction (Fig 8c).

Using this dosing regime on AHCre/APCflox/flox/R26R mice, multiple β- cateninhigh foci/lesions rapidly became visible throughout the upper crypt and villus epithelium. Representative pictures taken at day 3 post-induction are given in Fig 9. Mutant APC foci evident by high β-catenin levels occurred predominantly at crypt- villus junctions, but were also seen on the villi (Fig 8d). Very infrequently these lesions were also seen near the crypt base.

The majority of the APC-deficient cells present on the villus epithelium were lost after 4-5 days, presumably by shedding. The remaining APC-deficient lacZ-positive lesions/foci present within the crypts failed to expand over a 24 day period. A typical example of such a lesion is given in Figure 9e. No macroscopic adenomas were visible at this stage. Strikingly, these small lesions persisted over a 180 day period (Fig 9g), and only very rarely progressed to small adenomas, which did not expand beyond 2-3 villi (Fig 9f, h). This was in stark contrast to the high frequency formation of large adenomas initiated in the AHcre/APCflox/flox mice following high-dose β-NF induction. This suggested that the vast majority of adenomas in the latter mice resulted from loss of APC in stem cells.

In order to formally prove that transformation of intestinal stem cells is the major route to adenoma formation, we employed our Lgr5-EGFP-ires-CreERT2 knock-in mice as a stem cell -specific Cre line to inducibly delete the floxed APC. To this end, Lgr5-EGFP-ires-CreERT2 x APCflox/flox mice were generated. In these mice, the stem cell-specific Cre enzyme was activated with a single IP injection of Tamoxifen (Fig 10a). Subsequent phenotypic changes in the intestine were tracked over a 2 month period. Accumulation of the Wnt-effector protein β-catenin was first observed in isolated CBC cells at the crypt base after 3 days (Fig.10a). These transformed cells were GFP -positive, confirming the targeted deletion of APC in the intestinal stem cells (Fig.10b). After 5 days, multiple crypts throughout the intestine were observed to harbor transformed (i.e. β-catenin high) stem cells in association with highly proliferative clusters/pockets of β-catenin high cells within the transit-amplifying (TA) compartment (Fig.10c, d). This indicated that the Wnt-transformed stem cells remain viable and rapidly generate an expanding population of transformed progeny higher up the crypts. Eight days after inducing APC deletion in the stem cells, the "pockets" of transformed cells had continued to expand within the crypts and outpockets/evaginations of the crypt epithelium and small microadenomas within the associated villus stroma became evident (Fig 1Oe). Cells with accumulated β-catenin were never present on the villus epithelium in these mice, demonstrating that the expanding transformed population was restricted to the intestinal crypts. These observations are strikingly reminiscent of a model of adenoma formation, in which Wnt-transformed cells expressing high levels of the Wnt target gene EphB2 and -B3 expand within the crypt until they come into contact with the Ephrin-positive villus epithelium. The resulting repulsive forces consequently dictate that the microadenoma can only continue to expand by invading the stroma of the neighbouring villus where it is shielded from the Ephrin-positive villus epithelium.

The "outpockets" and microadenomas present in the 8 day induced

Lgr5KI/ APCflox/flox mice continued their aggressive expansion, as evidenced by the presence of multiple large adenomas throughout the intestine 36 days after initiating stem cell transformation (fig 1Of). To further investigate the hierarchy that exists between the APC-deficient stem cells and their transformed progeny, we examined expression of the stem cell marker protein Lgr5-EGFP during the various stages of adenoma formation in our model. In non-transformed stem cells, Lgr5-EGFP expression was restricted to the Crypt Base Columnar (CBC) cells (Fig 1 Ia). Expression of this stem cell marker was maintained following the initial transformation of the stem cells after 3 days (Fig 1 Ib) and was also clearly evident in the "pockets" of recently expanded transformed progeny within the crypts after 8 days (Fig 1 Ic), indicating that at least some aspect of "sternness" was conferred to these cells. However, there was a marked down- regulation of Lgr5-EGFP expression on the larger adenomas present in the intestines of 36-day induced mice, despite uniformly high β-catenin levels throughout the tumor (Fig l id). Lgr5-EGFP expression was limited to a few scattered cells within the tumor mass (Fig. 12). These GFP-positive cells retained the slender, wedge- shaped morphology characteristic of the CBC intestinal stem cells. It is therefore tempting to speculate that the Lgr5 expression in larger adenomas is marking a rare population of stem cells responsible for fueling their continued growth. Taken together, these data demonstrate that transformation of stem cells through loss of Ape is an extremely efficient route towards initiating intestinal adenoma formation. The kinetics of this process suggest that no further mutations are required once both Ape alleles are lost in intestinal epithelium, which is in accordance with the tissue-tropism of Ape's tumor suppressor activity.

Example 4 Generation and use of antibodies directed against LGR5 and LGR6

Materials and Methods

Monoclonal rat antibodies were generated by Genovac (Freiburg, Germany) by intramuscular injection of rats with an expression plasmid expressing either human Lgr5 or Lgr6. Rat B cells were fused with mouse myeloma cells. The resulting hybridomas were screened on HEK293 cells that were transfected with human or Mouse Lgr5 or Lgr6 expression plasmids.

L8 (DNTcf4-LS174T) cells were cultured with and without Doxycycline for 48hrs. L8 cells are clonal derivatives of LS174T cells. Upon Doxycycline (DOX) induction the L8 cells express a dominant negative form of T-cell Factor 4 (DNTcf4; see Roose et al, 1999, Science 285: 1923-1926). DNTcf4 turns off constitutive active Wnt pathway. Rat IgG was used as negative isotype control. After 48hrs cells are washed with ice cold PBS and brought into suspension using 5mM EDTA. All the following steps are done at 4°C. Cells were blocked for 30 min in PBS containing 2% BSA. Primary (1st) and Secondary (2nd) antibody reagent were incubated subsequently for lhr, and washed with ice cold PBS/2%BSA. For the primary antibody staining we used undiluted hybridoma supernatant, 2nd antibody staining was done using Qdot® 655 goat F(ab')2 anti-rat IgG conjugates (H+L) (Molecular Probes/Invitrogen). Prior to analysis propidium iodine was added to exclude dead cells in the analysis.

Results

The specificity of some of the isolated antibodies is shown in Tables 1 and 2, as tested by FACS analysis. 9G5 is a rat monoclonal antibody directed against hLgr5. The analysis of endogenous Lgr5 expression was determined in L8 cells. L8 cells are clonal derivatives of LS174T cells. Upon Doxycycline (DOX) induction, L8 cells express dominant negative Tcf4 (DNTcf4). DNTcf4 turns off constitutive active Wnt pathway. This is reflected in Fig. 13, showing FACS staining of L8 cells with 9G5 antibody or IgG control antibody. After 48hrs of DOX induction, indeed a reduction in endogenous hLgr5 protein levels was observed, as also becomes clear from the reduction in the fluorescent means of the peak for the L8 cells treated with doxycycline. This experiment was also performed with LGR5-specifϊc antibodies 2F10, IOCI and 6C10. As shown in Figure 14, similar results are obtained with any of these LGR5- specifϊc antibodies.

Example 5 Sequence determination of light chain and heavy chains, including CDR regions, of LGR5-specifc/LGR6-specific antibodies

Materials and methods:

Hybridoma sequence

The hybridomas were produced as described in the materials and methods section of

Example 4.

Hybridoma sequences were determined from Lgr5-specific and/or Lgr6-specific clonal hybridoma cell lines NR 2F10 (see Table 1) and 6d8 and 2f4 (see Table 2).

Total RNA was isolated using Trizol reagent and cDNA generated using superscript reverse transcriptase (Promega). cDNA was amplified using PCR primers designed to amplify the IgG antibody Fv-DNA sequences in a 'touch-down' PCR. PCR fragments were cloned into either P JET 1.2 (Fermentas) or PGEM-T (Promega) cloning vectors and subsequently sequenced using vector- specific primers on an ABI sequencer. Primer sequences are provided in Table 3.

In this experiment, mouse-specific oligos are used.

Results

The light chain sequence of LGR5 -specific antibody NR 2F10 and the heavy chain sequences of LGR6-specifϊc antibodies 6d8 and 2f4 are depicted in Figure 15. The CDR regions are indicated in bold and in italics. The CDR sequences were determined according to Kabat (Kabat et al., "Sequences of Proteins of Immunological Interest," U.S. Dept. of Health and Human Services, National

Institute of Health, 1987). Antibodies or functional equivalents thereof comprising at least one of these CDR sequences constitute a high affinity binding compound with a high specificity for their target proteins LGR5 and/or LGR6. Example 6 Sequence determination of light chain and heavy chains, including CDR regions, of all other LGR5-specifc/LGR6-specific antibodies

Materials and Methods mRNA was extracted from the hybridoma cell pellets. Total RNA was extracted from the pellets using Fusion Antibodies Ltd in-house RNA extraction protocol. cDNA was created from the RNA by reverse-transcription with an oligo(dT) primer. PCR reactions were performed using variable domain primers to amplify both the VH and VL region of the monoclonal antibody DNA. Succesfull VH and VL PCRs were cloned into the Invitrogen sequencing vector pCR2.1 and transformed into TOPlO for positive transformants. Selected colonies were picked and analyzed through sequencing. Results for the Lgr5 specific antibodies are shown in Figure 16. Results for the Lgr6 specific antibodies are shown in Figure 17. For all antibodies the CDRl, 2 and 3 are indicated as sequences in the boxed region.

Example 7 Alignment of the Lgr5 and Lgr6 antibody sequences.

Materials and Methods

To determine whether the sequenced antibodies are derived from a common parental clone, the VDJ gene usage was analysed as follows: The heavy chain variable domain for each of the sequenced antibodies was BLASTed using the IMGT server to determine the separate Variable (V), Joining (J) and Diversity (D) genes that comprise the immunoglobulin variable heavy domain. The results of this BLAST analysis are shown in Table 4 for the Lgr5 antibodies and in Table 5 for the Lgr6 antibodies. Those antibodies having the same V, J and D genes were then grouped and aligned separately to confirm their similarity (Figure 18 for Lgr5 antibodies en Figure 19 for Lgr6 antibodies).

Results Lgr5 specific antibodies (Table 4)

All antibodies that contain an identical VDJ recombination likely derive from the same identical parent clone; the mutations probably reflect the result of somatic hypermutation and they probably recognize the same identical epitope. Therefore, antibodies within group 1 , 2 and 3 contain related antibodies recognizing the same epitope. The antibodies in Group 1,2 and 3 may have derived from the same initial VDJ recombination event but have likely diverged through subsequent somatic mutation in vivo prior to monoclonal hybridoma fusion. They probably bind the same antigen epitope but are likely to have differing affinities and avidities.

The remainder of the antibodies do not show significant similarities in their gene usage and CDR3 region. This suggests that these antibodies most likely recognize different epitopes on the extracellular region of Lgr5.

It is interesting to see that the only antibody that recognizes both human and mouse Lgr5 (9C5) does not fall within any of the groups 1-3. Also the antibodies that detect Lgr5 most efficiently (3B9 and 8F2) both fall within group 1. These observations supports the conclusion that this antibody recognizes an epitope different from the epitopes recognized by the other antibodies that did cluster in one of the groups (in case of 9C5) and that these antibodies are derived from one parental clone and recognize the same epitope (in case of 3B9 and 8F2). Lgr6 specific antibodies (Table 5)

All antibodies that contain an identical VDJ recombination likely derive from the same identical parent clone; the mutations probably reflect the result of somatic hypermutation and they probably recognize the same identical epitope. Therefore, antibodies within group 4 contain related antibodies recognizing the same epitope . The sequences of both the VH and VL CDR3 for both antibody clones are identical and have near identical sequences for the complete VH region, suggesting that they are indeed derived from the same parental clone and recognize the same epitope on Lgr6. The remainder of the antibodies do not show significant similarities in their gene usage and CDR3 region. This suggests that these antibodies most likely recognize different epitopes on the extracellular region of Lgr6.

The fact that 2F4 and 6D8 do not recognize mouse Lgr6 (see Table 2) support the conclusion that these antibodies differ significantly from the antibodies that are clustered in group 4 and recognize different epitopes on Lgr6.

Example 8 hLGR5 specific antibody 3B9 recognizes hLgr5

Materials and Methods

Crypts are isolated from wild type mice and mice transgenic for hLGR5 driven by the Villin promoter (Villin hLGR5 mice). The villin promoter is active in all epithelial derived cells of the gastrointestinal lineage. Crypt cells were isolated according to protocol described in Nature 457, 608-611. Dissociated crypts cells were stained with hLGR5 specific antibody clone 3b9. Appropriate controls included cells stained with the secondary antibody only.

Results

Until now, no Lgr5 or Lgr6 specific antibodies are commercially available that are able to bind to Lgr5 or 6 on living cells. In example 4 we showed that all of the generated Lgr5 specific antibodies were able to bind to cancer cells that express Lgr5 endogenous Iy (LS 174) and a cell line overexpressing human Lgr5. The antibodies were tested in FACS analysis, staining living cells. Here we show that endogenous hLgr5 could also be detected using one of the Lgr5 specific antibodies as listed in Table 1. Figure 20 shows that endogenous hLgr5 can be detected using the Lgr5 specific antibody clone 3B9. A significant increase in fluorescent signal (from 79 as measured for Lgr5 on wt crypts compared to 713 on Villin hLgr5 crypts) was detected. Non specific binding was excluded using appropriate negative controls

Example 9 Detection of LGR5 by LGR5 specific antibodies using molecular biology

Materials and Methods

A selection of the Lgr5 antibodies were tested in western blot experiments. Cells were transfected with hLGR5, mLGR5 or control plasmid. After 48hrs cells were lysed with SDS/β-mercapthoethanol containing buffer. Samples were run on a 10% SDS page gel, each well contains 40 μg of cell lysate. SDS page gel was transferred unto PVDF membrane and stained overnight with hLgr5 specific antibodies 3B9, 2B8 or Flag antibody. After washing away the primary antibody, a secondary antibody with conjugated HRP incubation allowed for ECL development of the blot.

Results

Figure 21 shows that both 3B9 and 2B8 were able to detect hLGR5 in a cell lysate of cells transfected with hLGR5. Importantly, 3B9 did not detect mLGR5 in a cell lysate of cells transfected with mLGR5, confirming the FACS data represented in table 1.

Example 10 Identification of binding of LGR5-specific antibodies to LGR5 deletion mutants

Material and Methods

Generation of LGR5-Fc fusion protein

A human LGR5-Fc construct was generated using standard recombinant DNA techniques. A PCR-produced DNA fragment encoding the extracellular domain (aa: 1-546) of the LGR5 receptor was cloned in frame into the Age-I and Xho-I restriction sites of the pFuse-hlgGl-Fcl expression vector (Invivogen). The cDNA sequence of the LGR5-Fc fusion is represented by SEQ ID NO: 63. The corresponding amino acid sequence of the LGR5-Fc fusion protein is represented by SEQ ID NO:64. Proper insertion of PCR product was confirmed by restriction analysis, sequence verification and western blot (data not shown). To obtain LGR5-Fc-conditioned medium, 15.10⁶ HEK cells were transfected with 20μg of DNA using Fugene (Roche) as a transfection reagent, according to manufacturer's instructions. After 20 hrs, the FCS-supplemented RPMI medium (Gibco) was replaced by Optimem (Gibco). After another 48 hrs, the conditioned medium was harvested, cells removed by centrifugation, and samples were subjected to a dot blot for quantification of the fusion protein. Dot blotting

For dot blotting, 25 μl of a 2-fold dilution series in Optimem (Gibco) of LGR5-Fc conditioned medium and a calibration curve of purified human IgGl (Sigma) ranging from 10 - 0.15 mg/ml were spotted on pre-wetted (standard SDS-PAGE transfer buffer containing 20% methanol) immobilon-P (Millipore), saturated with PBS/10% (w/v) milk proteins, incubated with Goat anti HulgG/HRP (Pierce)(l :1000) in PBS/10% (w/v) milk proteins, extensively washed with PBS and imaged using ECL (Amersham).

LGR5 ELISA

To assess the binding of LGR5 antibody to the LGR5-Fc fusion protein, we developed an ELISA method. 96-well flat bottom tissue culture plates (Greiner) were incubated overnight with 200 μl of LGR5-Fc conditioned medium, diluted with Optimem to a final concentration of 1 μg/ml of the LGR5-Fc protein. After removal of the conditioned medium and a brief wash with Optimem, wells were incubated for 2 hrs with 150 μl of hybridoma supernatants containing LGR5 antibodies diluted 1 :1 with PBS/10% FCS. After washing twice with PBS/10%FCS, binding of the LGR5 antibodies was detected by incubating at room temperature for 1 hr with 1 :2000 diluted Goat anti Rat IgG/HRP (Pierce) in PBS/10% FCS. After washing the wells extensively, 150 μl of the HRP substrate TMB (BD Biosciences) was added. The reaction was stopped by adding 150 μl of 2M H₂SO₄ and measured by monitoring the absorbance at 450 nm.

LGR5 deletion mutants

We produced shortened, C-terminus FLAG-tagged versions of the human LGR5 receptor through standard recombinant DNA techniques (Figure 24 and 26). In total, 5 deletion constructs were generated. In the shortest version, only the transmembrane domain (aa: 562-825), C-flanking cysteine rich linker region (aa: 474-561) and leader peptide (aa: 1-21) are represented. The other four plasmids encode in addition the N- terminal region (aa: 22-70) and respectively, LRRl-4 (aa: 71-166), LRRl-8 (aa: 71- 258), LRR1-12 (aa: 71-356) or LRR1-4+13-17 (aa: 71-166 + 357-473). The 5 constructs were verified using restriction analysis, sequence analysis and correct recombinant protein production was tested using anti-FLAG Western blotting (data not shown).

Immunocytochemical staining of transfected COS cells COS cells were transfected with the different LGR5 deletion constructs using 3 μl of Fugene (Roche) per μg DNA and replated into 48-well plates after 24 hrs. 48 Hrs after transfection, cells were incubated in PBS containing 4% paraformaldehyde and 0.2% Triton X-IOO for 5 minutes. After washing cells with PBS/10%FCS three times, cells were incubated in hybridoma supernatants containing LGR5 antibodies or antibodies (in PBS/10% FCS) against the FLAG-tag (M2, Sigma) or rabbit anti-hulgG/HRP (Pierce). Detection of rat LGR5 antibodies was performed using Goat anti Rat IgG/HRP (Pierce), detection of the mouse M2 antibody was performed using rabbit anti Mouse IgG/HRP (Pierce). The HRP substrate AEC (BD Biosciences) was used to visualize binding of the various antibodies.

Results and discussion

LGR5 is a G-coupled, seven-transmembrane receptor encompassing a signal peptide (aa: 1-21), N-Terminal Region (N-term, aa: 22-70), 17 Leucine-Rich Repeats (LRR, aa: 71-473), C-Flanking cysteine rich linker region (CRL, aa: 474-561), Transmembrane region (TM, aa: 562-825) and C-terminal region (C-term, 826-907) (Hsu et al, 1998 MoI Endocrinol. Dec;12(12):1830-45). Figure 24 shows a schematic representation of the different domains present in the LGR5 protein. Because the LGR5 antibodies were obtained with DNA vaccination using a recombinant vector, encoding also the TM region, some of the antibodies could be reactive with any of the three exo-loops that are typical for this class of seven-membrane spanning proteins. To more specifically assess the precise binding regions of the anti LGR5 monoclonal antibodies (as generated in example 4), we performed an assay to establish whether the antibodies bind the extracellular domain of the G-protein coupled receptor. To test this, we generated a fusion protein of the extracellular LGR5 domain (lacking the TM region) to a human Fc. In an ELISA approach, we tested the binding of the anti-LGR5 antibodies to the LGR5-Fc fusion protein. Figure 25 shows that the hybridioma supernatants from clone 1D9, 4Dl 1, 6C10, 3A4, 5A7, 9G5, 2B8, 3B9, 5C8, 7Bl 1 and 8F2 tested in this ELISA all resulted into specific binding with the extracellular part of the LGR5 receptor. In an attempt to more precisely map antibody reactivity towards the LGR5 receptor we generated 5 deletion constructs. Figure 26 gives a schematic representation of the different deletion constructs generated for the extracellular domain of the LGR5 receptor. We performed binding experiments with the LGR5 antibodies on permeabilized COS cells expressing full length and truncated forms of the receptor. As expected, all antibodies recognized the full length LGR5 receptor in this immunocytochemical assay. Figure 27 shows a representative example of the COS cell assay, transfected with a full length flag-tagged LGR5 construct. As a negative control, staining with a secondary antibody only was performed, showing no signal (Figure 27A). However, incubation with anti-Flag (Figure 27B) or the anti-LGR5 monoclonal antibody 1D9 (Figure 27C) results in a specific signal in transfected COS cells. When deletion constructs of the LGR5 receptor were expressed by the COS cells, only antibody 1D9 appeared to be able to bind all the truncated forms of the receptor. None of the other antibodies showed significant binding to any of the truncated LGR5 versions. Table 6 summarizes the results from the COS cell assays. Since antibody 1D9 works both in the ELISA experiment and in the COS cell assay, we conclude that this antibody reacts within the CRL domain (aa: 474-546) of the LGR5 protein, because the CRL domain is the only domain present in all deletion constructs and the LGR5-Fc fusion construct. The other antibodies recognize epitopes within the LRR region. We propose that the deletion of part of the LRR repeats results in conformational changes which cause the destruction of epitopes that are recognized by the remaining antibodies. Therefore, none of the LGR5 antibodies, except for antibodies produced by hybridoma 1D9, recognize the deletion mutants. In contrast, they do recognize the full length LGR5 and the LGR5-Fc fusion protein.

Example 11 Classification of Lgr5 antibodies through Surface Plasmon Resonance.

Material and Methods

LGR5-Fc fusion protein

The LGR5-Fc fusion protein was generated as described in example 10.

Surface Plasmon Resonance Label free detection was carried out to assess the interaction of anti LGR5 antibody with the extracellular part of the LGR5 receptor (LGR5-Fc) using Surface Plasmon Resonance (IBIS-iSPR). The LGR5-Fc fusion protein was immobilized at the sensor surface by capturing to a covalent coupled rabbit anti-Fc (Pierce). For the immobilization of the anti-huFc on the surface of an S-Easy6Spot G-AE sensor (IBIS Technologies, lot H408-117), the IBIS easy2spot immobilization protocol and standard operating procedures were applied. At 6 positions (each 2 mm square) a 1.5 ul droplet of the Rabbit anti-huFc (1 :10) in MES buffer (2-(N- morpholino)ethanesulfonic acid, 50 mM) pH 5.5 was added and the anti-Fc antibodies was allowed to immobilize covalently to the sensor surface. As a reference two different spots were obtained. One spot was coupled with clean MES buffer only (so no protein will be immobilized) and the other spot was made using HSA (Human Serum Albumin) 200 ug/ml in Sodium Acetate buffer pH 4.5. After installing the sensor chip in the instrument, 12 regions of interests (RoI) were positioned on the anti-huFc modified surface (two per treated area). The standard injection procedure included a glycerol injection for sensor calibration purposes. Spot to spot calibration was not necessary because of sequential injections of different compounds and comparison of the binding behaviour of several analysis cycles (baseline, (association, dissociation) nx, regeneration) of a single spot/RoI. In an overlay plot the repeated injections could be compared adequately with each other. To prevent mixing of the sample with buffer, 25 μl air plugs were applied after each sample injection. First, pure LGR5-Fc fusion protein in Optimem medium (-20-40 microgram/ml) was injected to allow capturing the fusion protein to the anti-huFc antibody during 1800 seconds. In this time scale, the LGR5-Fc fusion protein was allowed to almost saturate the surface. After washing the surface with buffer (dissociation phase), an LGR5 antibody was injected and allowed exposure to the sensor surface during 600 seconds. The sensor surface was regenerated with an acid step (Glycine-HCl pH 2) to remove the LGR5-Fc and/or antibodies, after which it could be loaded again, up to 20 times.

Results and discussion

To be able to functionally group the anti-LGR5 antibodies (as generated in example 4) in clusters recognizing different binding domains, we tested interaction of the monoclonal LGR5 antibodies to the extracellular domain of the LGR5 receptor using label free detection in Surface Plasmon Resonance imaging. With this assay a change in mass (e.g. binding of a LGR5 antibody) on a surface (e.g. LGR5-Fc) can be measured because of an increase (shift) of the refractive index (m°) against the baseline indicated as R. Because the LGR5-Fc was also covalently coupled to an anti Fc antibody on the sensor surface, loading of this fusion protein resulted in a SPR shift (LGR5-Fc R). Subsequent loading of an LGR5 antibody resulted again in a SPR shift (Ab R). To be able to compare the antibodies with each other, we made use of %R (Ab R/LGR5-FC R *100%). The tested antibodies 1D9, 4Dl 1, 6C10, 3A4, 5A7, 9G5, 2B8, 3B9, 5C8, 7Bl 1, 8F2 and IOCI all showed a typical binding curve with the LGR5-Fc receptor. For example, in figure 28 the binding of the LGR5-Fc fusion is shown followed by the binding of monoclonal antibody 5A7 resulting in a clear second SPR shift. Addition of an irrelevant antibody to the fusion protein did not result in an effect. Table 7 shows the %R of the 12 tested antibodies with respect to the LGR5-Fc-coated SPR sensor. Antibody 1D9 showed a relative response of about 33%, whereas all other antibodies show relative binding capacities ranging between 9.5 and 16.2%.

In order to determine whether the LGR5 antibodies are able to recognize the same epitope containing regions and to cluster the antibodies based on their binding of the same region of the extracellular domain of the receptor, competition analysis was carried out. In this inhibition mapping method, after capturing the LGR5-Fc, a single antibody is injected followed by a second antibody. If the two antibodies are able to bind to an epitope in the same region or the same epitope, sequential injection of these antibodies into the LGR5-Fc-coated SPR sensor will result in an inability of the second antibody to bind and will therefore not result in a change in R. If an additional increase in R is detected after injection of the second LGR5 antibody, this indicates that the antibodies bind different regions, and epitopes, of the LGR5-Fc fusion protein and will therefore fall into different clusters. As an example figure 29 shows that the sequential injection of antibody 5C8 followed by 4Dl 1 (upper line) on an LGR5-Fc- coated SPR sensor results in an additive shift upon both antibody injections, suggesting that 4Dl 1 and 5C8 bind to different regions within the extracellular domain of the LGR5 receptor. In contrast, sequential addition of 4Dl 1 when first 7Bl 1 (lower line) was bound does not show an extra shift, indicating that binding of 7Bl 1 to LGR5-Fc prevents subsequent binding of 4Dl 1 to LGR5-Fc. This indicates that both antibodies bind the same region within the extracellular domain of LGR5. These competition experiments were performed for all LGR5 antibodies. Table 8 summarizes the results from these competition experiments and shows that the LGR5 antibodies can be divided in three binding clusters. The first group only contains antibody 1D9. A second group was identified, consisting of antibody 4Dl 1 and 7Bl 1. AU other antibodies (6C10, 3A4, 5A7, 9G5, 2B8, 3B9, 5C8, 8F2 and lOCl) are clustered in a third group. Figure 30 shows that the combination of 3 antibodies from the three different clusters (6C10, 4Dl 1 and 1D9) results, as expected, in an additive signal in the SRP assay and confirms the conclusion from the region mapping experiment (table 8).

Conclusion

The 12 tested antibodies all show a typical binding curve with the LGR5-Fc receptor in Surface Plasmon Resonance imaging. The antibodies were clustered in 3 independent groups based on competition analysis. SPR data showed that antibody 1D9 binds to a unique region confirming the deletion assay data from example 10. Since the %R of this antibody is always about twice the value of the other antibodies, this suggests that the epitope recognized by this antibody is represented twice in the extracellular domain of LGR5. Antibody 4Dl 1 and 7Bl 1 were placed in a second region cluster and the remaining antibodies (6C10, 3A4, 5A7, 9G5, 2B8, 3B9, 5C8, 8F2 and lOCl) all fell in a third cluster. This does not mean that all these antibodies bind the same epitope, e.g. allosteric hinderance by one antibody may prevent the binding of another antibody. Based on the alignment of the sequencing data (example 7) of the CDR regions of the antibodies, the third cluster of the SPR epitope competition method can be divided into at least 4 subgroups that recognize 4 different epitopes. These are as follows: cluster 3-1 (IBIS-Seq data), containing the clones 3B9, 5C8, 8F2 and lOCl. Cluster 3-2, containing the clones 3A4 and 5A7. Cluster 3-5 contains clone 6C10 and cluster 3-6 only contains clone 9G5 which is the only antibody that also recognizes mouse Lgr5 next to human LGR5. The second cluster of the SPR epitope competition contains antibodies 4Dl 1 and 7Bl 1. Alignment of the sequencing data also clustered the antibodies 4Dl 1 and 7Bl 1 in the same subgroup (example 7). The properties of the different antibodies are summarized in table 9.

Example 12 Visualization of endogenous LGR5 in human organoid cultures

Material and Methods

Human organoid cultures

Human colon crypts were isolated from resected normal colonic specimen and cultured as organoid structures for 7 days using the established organoid culture system (Sato et al, 2009 Nature May 14;459(7244):262-5). Since this protocol was optimized for mouse derived organoid cultures, we made a small change by the addition of Wnt3a conditioned medium, in order to ensure optimal growth of the human colon organoids. To obtain this conditioned medium, Wnt3a is expressed in a cell line by transfecting a suitable expression construct encoding said ligand. Said cell line is cultured and the culture medium comprising the secreted ligand is harvested at suitable time intervals. For example, cells start the production of Wnt3a at the moment they reach confluency and stop growing. Culture medium from cells that were not transfected or infected with said empty expression construct was used as a negative control. The conditioned medium was harvested and tested, for example in an assay wherein luciferase expression in controlled by TCF responsive elements to quantitate the presence of a Wnt agonist such as Wnt3a (Korinek et al., 1997. Science 275:1784-1787).

Staining of endogenous LGR5

Human colon organoid structures were fixed with 4% paraformaldehyde. After 10 minutes, the structures were washed and blocked using blocking reagent (Power

Block (BioGenex)). Primary antibody (hybridoma supernatant from anti LGR5, clone 8F2) was incubated overnight at 4°C. Structures were washed with PBS/2% BSA and then treated with PBS/0.3% H₂O₂ for 10 minutes. H₂O₂ was washed (PBS/ 2% BSA) away prior to 2^nd antibody incubation. The 2^nd antibody, anti rat-HRP (Simple Stain Rat Max PO (Nichirei, Japan))., was incubated for 6hr at 4°C. Organoids were washed (PBS/ 2% BSA) prior to subsequent TSA amplification: Tyramide Signal Amplification (TSA) amplification system (Invitrogen, Tyramide signal amplification kit, catalog nummer MP 20911). TSA amplification mixture (ImI amplification buffer + 2ul 0.3%H2O2 + 4ul labeled Alexa Fluor 488 Tyramide) was added to the structures for 10 minutes at room temperature. DAPI staining was used for visualization of nuclei. After washing, staining was visualized using confocol microscopy.

Results and discussion

Human colon organoids were grown for 7 days using our established culture conditions. Similar to mouse organoids (Sato et al., 2009 Nature May 14;459(7244):262-5), human organoids also contain budding structures comprising dividing cells (data not shown). We employed Tyramide Signal Amplification to detect LGR5 expression in human colon organoid cultures using rat anti-hLGR5 antibody subclone 8F2 as an example. Figure 31 shows LGR5 specific staining in the budding regions of the human organoid cultures. This is exactly the regions where LGR5 positive cells are also detected in mouse organoids (Sato et al., 2009 Nature May 14;459(7244):262-5). Figure 3 IA shows the DAPI staining to visualize nuclei. Figure 3 IB shows bright field image and Figure 31C shows the LGR5 staining with 8F2 supernatant of the same region of a human organoid. The generated anti LGR5 monoclonal antibodies, as generated in example 4, can be used to visualize endogenous LGR5 expression.

Table 1. Specificity of Lgr5 antibodies. All antibodies recognize endogenous Lgr5 in LS 174 cells. 9G5 recognize both mouse and human Lgr5. The colon cancer cell lines; DLDl and SW480, LIMl 863 do not show specific staining for Lgr5. These antibodies were tested negative for cross reactivity against mouse Lgr4, 6 and human Lgr4, and 6.

Table 2 Specificity of Lgr6 antibodies. Antibodies Id8 and 3d8 recognize mouse Lgr6 and hLgr5 in addition to human Lgr6. The colon cancer cell lines; LS 174, DLDl and SW480 do not show specific staining for Lgr6. These antibodies were tested negative for cross reactivity against mouse Lgr4, 5 and human Lgr4.

Table 3 IgG antibody Fc-DNA sequence PCR primers:

Kappa L-chain reverse primers; 25 individually synthesized oligos, pooled, representing 50 variants:

MVK-I GACATTGTTCTCACCCAGTCTCC

MVK-2 GACATTGTGCTSACCCAGTCTCC

MVK-3 GACATTGTGATGACTCAGTCTCC

MVK-4 GACATTGTGCTMACTCAGTCTCC

MVK-5 GACATTGTGYTRACACAGTCTCC

MVK-6 GACATTGTRATGACACAGTCTCC

MVK-7 GACATTMAGATRACCCAGTCTCC

MVK-8 GACATTGCAGATGAMCCAGTCTCC

MVK-9 GACATTCAGATGACDCAGTCTCC

MVK-IO GACATTCAGATGACACAGACTAC

MVK-11 GACATTCAGATGATTCAGTCTCC

MVK-12 GACATTGTTCTCAWCCAGTCTCC

MVK-13 GACATTGTTCTCTCCCAGTCTCC

MVK-14 GACATTGWGCTSACCCAATCTCC

MVK-15 GACATTSTGATGACCCARTCTC

MVK-16 GACATTKTGATGACCCARACTCC

MVK-17 GACATTGTGATGACTCAGGCTAC

MVK-18 GACATTGTGATGACBCAGGCTGC

MVK-19 GACATTGTGATAACYCAGGATG

MVK-20 GACATTGTGATGACCCAGTTTGC

MVK-21 GACATTGTGATGACACAACCTGC

MVK-22 GACATTGTGATGACCCAGATTCC

MVK-23 GACATTTTGCTGACTCAGTCTCC

MVK-24 GACATTGTAATGACCCAATCTCC

MVK-25 GACATTGTGATGACCCACACTCC Kappa L-chain forward primer: mck-1 ACACTCATTCCTGTTGAAGCTCTTGAC H-chain variable region reverse primers, 25 individually synthesized oligos, pooled, representing 88 variants:

MVH- 1 GCCGGCCATGGCCGAGGTRMAGCTTCAGGAGTCAGGAC MVH-2 GCCGGCCATGGCCGAGGTSCAGCTKCAGCAGTCAGGAC MVH-3 GCCGGCCATGGCCCAGGTGCAGCTGAAGSASTCAGG MVH-4 GCCGGCCATGGCCGAGGTGCAGCTTCAGGAGTCSGGAC MVH-5 GCCGGCCATGGCCGARGTCCAGCTGCAACAGTCYGGAC MVH-6 GCCGGCCATGGCCCAGGTCCAGCTKCAGCAATCTGG MVH-7 GCCGGCCATGGCCCAGSTBCAGCTGCAGCAATCTGG

MVH-8 GCCGGCCATGGCCCAGGTYCAGCTGCAGCAGTCTGGRC MVH-9 GCCGGCCATGGCCCAGGTYCAGCTYCAGCAGTCTGG MVH- 10 GCCGGCCATGGCCGAGGTCCARCTGCAACAATCTGGACC MVH- 11 GCCGGCCATGGCCCAGGTCCACGTGAAGCAGTCTGGG MVH-12 GCCGGCCATGGCCGAGGTGAASSTGGTGGAATCTG MVH- 13 GCCGGCCATGGCCGAVGTGAAGYTGGTGGAGTCTG MVH- 14 GCCGGCCATGGCCGAGGTGCAGSKGGTGGAGTCTGGGG MVH- 15 GCCGGCCATGGCCGAKGTGCAMCTGGTGGAGTCTGGG MVH- 16 GCCGGCCATGGCCGAGGTGAAGCTGATGGARTCTGG MVH- 17 GCCGGCCATGGCCGAGGTGCARCTTGTTGAGTCTGGTG MVH- 18 GCCGGCCATGGCCGARGTRAAGCTTCTAGAGTCTGGA MVH- 19 GCCGGCCATGGCCGAAGTGAARSTTGAGGAGTCTGG MVH-20 GCCGGCCATGGCCGAAGTGATGCTGGTGGAGTCTGGG MVH-21 GCCGGCCATGGCCCAGGTTACTCTRAAAGWGTSTGGCC MVH-22 GCCGGCCATGGCCCAGGTCCAACTVCAGCARCCTGG MVH-23 GCCGGCCATGGCCCAGGTYCARCTGCAGCAGTCTG MVH-24 GCCGGCCATGGCCGATGTGAACTTGGAAGTGTCTGG MVH-25 GCCGGCCATGGCCGAGGTGAAGGTCATCGAGTCTGG H-chain forward primers: MJH-REVl &2 GGGGGTGTCGTTTTGGCTGAGGAGACGGTGACCGTGG MJH-REV2INT GGGGGTGTCGTTTTGGCTGAGGAGACGGTGACAGTGG MJH-REV3

GGGGGTGTCGTTTTGGCTGAGGAGACGGTGACCAGAG MJH-REV4

GGGGGTGTCGTTTTGGCTGAGGAGACGGTGACCGAGG Variable position key: R (A/G); M (A/C); Y (T/C); W (A/T); S (G/C); K (G/T); H (A/T/C); B (G/C/T); V (G/A/C); D (G/A/T); N (G/A/T/C)

Table 5: Heavy chain variable domain gene analysis of Lgr6 specific hybridomas

Table 6 Summarizing of COS cell transfection assay with full length and

LGR5 deletion constructs.

Table 7 Interaction of the 12 anti LGR5 antibodies with the LGR5-Fc fusion protein as tested by Surface Plasmon Resonance. With this assay a change in mass (e.g. binding of a LGR5 antibody) on a surface (e.g. LGR5-Fc) can be measured because of an increase (shift) of the refractive index (m°) against the baseline indicated as R. Because the LGR5-Fc was also covalently coupled to an anti Fc, loading of this fusion protein resulted in a SPR shift (LGR5-Fc R). Subsequent loading of an LGR5 antibody resulted again in a SPR shift (Ab R). Because the interaction of LGR5-Fc (LGR5-Fc R) differs between various measurements and decreased in time as a result of regeneration steps, the binding of the antibody (Ab R) to LGR5-Fc decreased too. To be able to compare the antibodies with each other we made use of %R (100% * Ab R/LGR5-FC R).

Table 8 In order to cluster the antibodies in independent binding groups, competition analysis was carried out. This means that a sequence of experiments has been carried out adding two antibodies sequentially as indicated in the table. Hereby the effect of change in R is measured for two antibodies successively loaded on the LGR5-Fc-coated SPR sensor. IfR changes upon addition of the second antibody a "1" is given in the table indicating that these two antibodies bind to a different region. IfR does not change upon addition of the second antibody, a zero is given in the table, indicating that these two antibodies bind the same region of LGR5.

Table 9 Summary of results obtained with LGR5 antibodies in different assays as described in the examples. Seq VL and Seq VH indicate the presence of sequencing data (example 7) as shown in figure 16. Cluster according to Seq Alignment VH, stands for the clustering of the hybridomas according to their VH/CDR homology (Figure 18). Cluster according to SPR, is the clustering of antibodies into 3 groups according to the competition analysis in the Surface Plasmon Resonance imaging (Table 8). The COS cell assay is the immunocytochemical staining of deletion mutant form of the receptor expressing only transmembrane region and the C-Flanking

10 cysteine rich linker region (Table 6). ELISA is the LGR5-Fc ELISA (Fig 25). FACS assays are the transfection assays of 293T cells transfected with full length hLGR5 or mLgr5 (Table 1).

Cluster according to TM/CRL

Primary Seq Cluster plasmid clone Definitive Seq VL Seq VH Alignment according COS cell hLGR5 mLGR5

LGR5 subclone Fig 16 Fig 16 VH Fig 18 to SPR assay ELISA FACS FACS

15 Table 10: Summary of results obtained with LGR6 antibodies in different assays as described in the examples. Seq VL and Seq VH indicate the presence of sequencing data (example 7) as shown in figure 17 Cluster according to Seq Alignment VH, stands for the clustering of the hybridomas according to their VH/CDR homology (Figure 19). FACS assays are the transfection assays of 293T cells transfected with full length hLGR5, hLGR6 or mLgrβ (Table 2).

Claims

Claims

1. An antibody that binds endogenous human Leucine-rich-repeat-containing G- protein-coupled Receptor 5 (Lgr5) and/or Lgr6.

2. An antibody according to claim 1, wherein said antibody in addition binds mouse Lgr5 and/or Lgr6.

3. An antibody according to claim 1 or claim 2, wherein the antibody is a monoclonal antibody.

4. An antibody according to claims 1-3, comprising one or more complementarity- determining region having an amino acid sequence selected from Figure 16 or 17.

5. An antibody according to any of the previous claims, wherein the antibody is a single domain antibody, a F(ab')2, Fab, Fab', Facb, or single chain Fv (scFv) fragment.

6. An antibody according to any of the previous claims, wherein the antibody is de- immunized, human-like, resurfaced, or humanized.

7. An antibody according to any of the previous claims, wherein the antibody binds Lgr5 and/or Lgr6 with a kD of 3 x 10 8 M or lower.

8. A cell producing an antibody according to any of claims 1-7.

9. A method for producing an antibody comprising culturing a cell as defined in claim 8 and harvesting said antibody from said culture.

10. A method for isolating stem cells comprising

- preparing a cell suspension from a tissue or organ sample;

- contacting said cell suspension with an anti-Lgr5 and/or Lgr6 antibody as defined in any of claims 1-7; - obtaining cells bound to said antibody;

- isolating cells from said antibody.

11. An antibody according to any of claims 1-7, for use as a medicament.

12. An antibody according to any of claims 1-7 which is linked to a toxic agent.

13. A pharmaceutical composition comprising an antibody or epitope-binding fragment thereof as defined in any of claims 1-7, or the antibody of claim 12, and a pharmaceutically acceptable carrier or excipient.

14. Use of the antibody as defined in claim 12 for targeting cancer stem cells.

15. A method for determining cancer stem cell content of a tumor or a body fluid, the method comprising - contacting said tumor or body fluid with an anti-Lgr5 and/or Lgr6 antibody as defined in any of claims 1-7;

- removing unbound antibody;

- detecting any bound complex comprising an anti-Lgr5 and/or Lgr6 antibody;

- determining the cancer stem cell content of the tumor or body fluid based on the presence of detected antibody.