WO2007062073A2 - Models of metastatic tumors - Google Patents

Abstract

Cells derived from spontaneous brain tumors of inbred VM mice are described. The tumor cells (e.g., the VM-M2 and VM-M3 cell lines) are highly metastatic, with the capacity to metastasize to multiple organ systems (e.g., lymph node, skeletal muscle, heart, spinal cord, liver, spleen, kidney, lung, and brain) when introduced outside the brain. Also described are tumor cells (e.g., the VM-NMl cell line) that form non-metastatic tumors. The invention includes these cells and cell lines, as well as methods of using the same to identify candidate chemotherapeutic and antimetastatic agents.

Description

MODELS OF METASTATIC TUMORS

Related Applications

This application claims priority under 35 U.S. C. §119(e) from U.S. provisional application serial number 60/739,891, filed November 23, 2005 and U.S. provisional application serial number 60/788,139, filed March 31, 2006, the entire content of each of which is incorporated by reference herein.

Government Support

The work described herein was conducted in part with funds from Grant No. 5000183 from the National Cancer Institute and Grant Nos. CA102135 and HD39722 from the National Institutes of Health. The U.S. Government may therefore have certain rights in the invention.

Background

Cancer metastasis is a complex process that involves intrinsic properties of the metastatic tumor cells as well as interactions between these cells and the host immune system (Nicolson (1984) Exp. Cell Res. 150:3-22; Kanda et α/. (1992) Int. J. Cancer 52:682-687). Metastatic tumor cells are distinguished from nonmetastatic tumor cells by their ability to migrate from a primary tumor site to a distant location where they adhere and grow. Alterations in the cell surface glycocalyx may underlie the metastatic potential of some tumor cells (Kim et al. (1975) Proc. Natl. Acad. Sci. USA 72:1012-1016; Hakomori (1996) Cancer Res. 56:5309-5318).

Several animal models of metastasis involve transplantation of tumor cells or tissue to rodents. In many of these systems, human or mouse tumor cells or tissue are transplanted, and the development of metastatic tumors at sites distant from the transplantation site are observed. When the transplant and the host are of different genetic backgrounds (for example, different species), immunocompromised hosts are used to avoid an immune response to the foreign cells or tissue. The disadvantage of these systems is that they cannot be used to examine the role of the immune system in tumor and metastasis progression. . . Current in vivo models of metastasis are reviewed, for example, in Khanna and Hunter (2005) Carcinogenesis 26:513-523. Summary of the Claims

This invention is based, hi part, on the discovery of spontaneous brain tumors in inbred VM mice and the establishment of three cell lines derived from those tumors. Two of the cell lines, the VM-M2 and VM-M3 tumor cell lines, are highly metastatic. Cells of the VM-M2 and VM-M3 cell lines can metastasize to multiple organs and organ systems {e.g., the liver, the spleen, the urinary system, including the kidney, the respiratory system, including the lung, and the nervous system, including the brain, etc.) after administration to a laboratory animal {e.g., when administered subcutaneously to a rodent, such as a VM mouse). In contrast, the VM-NMl tumor cell line does not observably metastasize when administered or grown subcutaneously. The invention includes these cell lines and cells or cell lines derived therefrom (any of which may be isolated or purified), as well as methods of identifying candidate chemotherapeutic agents and antimetastatic agents.

The invention further includes cells obtained from nervous tissue {e.g. brain tumors) that have the physical characteristics and functional abilities of the cells described herein and cell lines derived therefrom. For example, the invention encompasses cells obtained from brain tumors {e.g., spontaneous brain tumors) in the VM mouse, other types of mice, other rodents, or other animals, including nonhuman primates {e.g., chimpanzees) and humans so long as those cells are isolated and purified and so long as they have the physical characteristics {e.g., ganglioside content and/or gene expression profiles) and functional abilities {e.g., a loss of cell cycle control and/or an ability to metastasize) described herein.

Although the invention is not so limited, the results to date indicate that the present cell lines and methods have certain features that are advantageous in modeling metastasis. As noted above, in many of the currently available models, tumor cells are implanted into immunocompromised animals. The cells described herein, however, were obtained from spontaneous brain tumors in the VM mouse strain and can metastasize in an immune competent VM mouse. This allows analysis of metastasis in the context of an intact immune system, which is the context in which most naturally occurring cancers arise. Cells of the invention can metastasize following administration through various routes, including administration by either intravenous or subcutaneous injection. The latter method may constitute a better model of the early steps of metastasis, which includes invasion and intravasation. Cells of the invention have invaded all of the areas of the brain that have been examined when the cells were implanted orthotopically, and the cells also invade organs of multiple organ systems (including, but not limited to, liver, spleen, kidney, lung, and brain, etc.) when implanted subcutaneously outside the brain. In addition, the rate at which meMsϊasϊi occur S ϊiϋ' relatively fast and reproducible, allowing for efficient animal testing of candidate chemotherapeutic agents (e.g., anti-metastatic agents, cytotoxic agents, cytostatic agents, cytokine agents, antiproliferative agents, immunotoxin agents, gene therapy agents, angiostatic agents, cell targeting agents, etc.). Cells of the invention may be used to prepare models (e.g., cell models, animal models, etc.) of numerous types of metastatic cancers including but not limited to, metastases from brain cancer, ovarian cancer, lung cancer, intestinal cancer, colon cancer, prostate cancer, etc. A cell of the invention may include a detectable label that may be expressed by or attached to the cell.

Cells of the VM-NMl cell line do not observably metastasize, and cells of that line can be used as a convenient metastasis-negative control. The combined properties of cells described herein make them suitable for both developing therapies for tumor metastasis and for basic metastasis research (e.g., identification of molecular events that precipitate, enhance, or inhibit one or more of the steps in the process of metastasis). Molecular events can be identified by, for example, exposing a metastatic cell to an agent (e.g., a small molecule, a peptide, an antibody, or an siRNA or shRNA) that exacerbates or inhibits the expression or activity of a cellular target (e.g., a gene or the protein it encodes, which may be an enzyme such as a protease, kinase, or phosphatase, a second messenger, a ligand or receptor (on the cell surface or within the nucleus), or a transcription factor. Such assays may be performed in vitro and/or in vivo. One would then determine whether the agent affects the metastatic potential of the cell. Agents that inhibit metastasis may be useful as anti- metastatic agents per se or they may be useful insofar as they have served to identify a cellular target for therapeutic intervention. Agents that promote metastasis may have served to identify cellular targets that can then be inhibited to reduce the metastatic potential of a tumor cell. As is usual in patent specifications, reference may be made herein to a single cell. Of course, pluralities of cells are within the scope of the invention (e.g., a plurality of VM- M2, VM-M3, or MV-NMl cells collected in a vial or grown in a tissue culture vessel).

As described further below, cells in the VM-M2, VM-M3, and VM-NMl cell lines have been characterized by analyzing ganglioside content. There may be significant differences between the amount of specific gangliosides in metastatic versus non-metastatic cells of the invention that have been established from a spontaneous brain tumor of a mouse or those such as the VM-M2 or VM-M3 cell lines (metastatic) or VM-NMl cell line (non- metastatic).

Accordingly, the invention encompasses an isolated metastatic cell of a mouse (e.g., a VM mouse) brain tumor that contains about, less than, or less than about 1% GM3 as a

total ganglioside content. In some embodiments, a metastatic cell or plurality of metastatic cells of the invention, will contain about, less than, or less than about 1% GM3 as a percentage of the total ganglioside content of the cell or plurality, respectively, (e.g., about, less than, or less than about 0.8%, 0.5%, 0.2%, 0.1%, 0.05%, 0.01% GM3). In some embodiments, the amount of GM3 maybe undetectable in standard assays (e.g., assays by high performance thin-layer chromatography (HPTLC) performed as described herein) of metastatic cells from spontaneous brain tumors (e.g., VM-M2 or VM-M3 cells).

The amount of other gangliosides, such as GM2, can also be assessed. A metastatic cell or cells of a brain tumor (e.g., a spontaneous brain tumor of a mouse or those established as the VM-M2 or VM-M3 cell lines) may express about, less than, or less than about 5% GM2 as a percentage of the total ganglioside content of the cell or cells (the content being determined by a standard method known in the art and/or as described herein). For example, the invention features an isolated cell obtained from a brain tumor (e.g., a spontaneous brain tumor of a mouse such as a VM mouse) that contains about, less than, or less than about 5%, 4%, 3%, 2%, or 1% GM2 as a percentage of the total ganglioside content of the cell as detected using standard assays.

The amount of the ganglioside GD3 may also be assessed. A metastatic cell or cells of a brain tumor (e.g., a spontaneous brain tumor of a mouse or those established as the VM- M2 or VM-M3 cell lines) may express zero GD3, and a non-metastatic cell or cell line of the invention (e.g., a cell established from a VM-NMl cell) may have approximately 5% GD3 as a percentage of the total ganglioside content of the cell as detected by a standard method known in the art and/or as described herein.

The amount of another ganglioside, GAl (asialo-GMl) may also be determined in the assessment of cells and/or cell lines of the invention. Using standard assays, GAl expression may be detected in metastatic cells from a spontaneous brain tumor of a mouse such as a VM mouse, and/or cell lines such as VM-M2 and VM-M3. There may be no detectable expression of GAl in VM-NMl cells and other non-metastatic cells from a spontaneous brain tumor of a mouse such as a VM mouse. In addition, none of the cell lines, VM-M2, MV-M3, and VM-NMl may have detectable expression of the gangliosides GTIb or GQIb, which are characteristic of neurons.

Other markers can also be used to identify cells (e.g., metastatic cells) within the scope of the invention. For example, cells of the invention may express markers that are associated with or normally expressed by microglia (or microgliomas) or macrophages, including one or more of the markers that serve to identify a cell as a glial (e.g., microglial) ^[ celf"orϊϊ9(S)f)ϊia|e? ϊtf '" example, cells of the invention may express CD68, a highly glycosylated transmembrane protein; AIFl (allograft inflammatory factor 1 (AIFl, also known as Ibal)), a calcium-binding protein; macrophage receptor 1 (Macl); F4/80, a cell surface glycoprotein; CD45, a protein tyrosine phosphatase; CXCR4, a chemokine receptor; or CDl Ib. Accordingly, the invention features an isolated cell of a mouse brain tumor (e.g., a VM mouse) that (1) expresses one or more of the genes encoding CD68, Ibal, Macl, CD45, CDl Ib, CXCR4, and F4/80 and (2) spontaneously metastasizes to one or more distant organ sites (e.g., one or more of the lymph node, skeletal muscle, heart, spinal cord, brain, liver, spleen, kidney, or lung) following subcutaneous administration of a plurality of the cells to a mouse. These isolated cells and cells of cell lines derived from them may have, but do not necessarily have, one or more of the additional characteristics described herein (e.g., the ganglioside profile described herein).

Other characteristics of cells of the invention (e.g., metastatic or non-metastatic cells from a spontaneous brain tumor of a mouse such as a VM mouse) may include adhesion characteristics, which may be elevated in metastatic cells of the invention, rendering the cells resistant to trypsin treatment. In contrast, non-metastatic cells of the invention, such as a VM-NMl cell may have normal adhesion properties and be readily susceptible to trypsin treatment. Phagocytosis is another characteristic that may be used to distinguish metastatic or non-metastatic cells from a spontaneous brain tumor as described herein. VM-M2, and VM- M3 and other metastatic cells may demonstrate a significantly higher level of phagocytosis than non-metastatic cells such as VM-NMl cells or other non-metastatic cells from a spontaneous brain tumor as described herein.

As used herein the term "derived from," in reference to a cell and/or cell line of the invention, means any cell that is produced or obtained from a cell or cell line of the invention, including, but not limited to cells that are replicates of a cell or cell line of the invention. A replicate may be a cell that descends from and is substantially unmodified from a cell of the invention. A cell or cell line derived from a cell of the invention may also be a cell that is produced from or may be a descendant of a cell produced from a cell of the invention. Some derived cells and/or descendant cells are modified cells or cell lines that have new properties. Examples of modifications include but are not limited to: cell fusions (e.g., to produce hybridomas), genetic engineering (e.g., to produce transgenic cells, mutant cells, etc.), labeling, etc. It will be understood by those of ordinary skill in the art that the invention encompasses a cell or cell line derived from a MV-M2, VM-M3, or VM-NMl cell or cell line as described herein. In some embodiments, genetically engineered cells are cells that are

^a detectable label. Such cells may be used to track and/or visualize metastasis by cells of the invention. Detectably labeled cells may be detected in vitro, or in vivo. In some embodiments detectably labeled cells may be detected using realtime imaging methods. hi specific embodiments, a cell of the invention may be a cell of the cell line designated VM-M2, which was deposited with the American Type Culture Collection (ATCC; Manassas, VA) as Accession No. PTA-7204 on November 1, 2005; a cell of the cell line designated VM-M3, which was deposited with the ATCC as Accession No. PTA-7205 on November 1, 2005; or a cell of the cell line designated VM-NMl, which was deposited with the ATCC as Accession No. PTA-7203 on November 1, 2005. These deposits were made in accordance with the Budapest Treaty.

A cell (e.g., an isolated cell of a cell line) that originated with a cell described herein [e.g., a VM-M2, VM-M3, or VM-NMl cell or a derivative, progeny thereof (e.g., descendant thereof), or a cell having the characteristics described herein isolated from a brain tumor of another VM mouse or other source, including a human patient], is also within the scope of the invention. While these cells are described further below, it will be understood that they can be hybrid cells (i.e., cells in which either the cytoplasm or the nuclei, but not both, is that of a cell such as a VM-M2, VM-M3, or VM-NMl cell) or genetically modified cells.

The isolated cell or a plurality thereof also has a dysregulated cell cycle (resulting, for example, in unwanted or excessive proliferation) or a particular capacity for metastasis. For example, with respect to metastasis, following subcutaneous administration of a cell of the invention, or a plurality thereof, to a mammal (e.g., a mouse), the cell or cells will spontaneously metastasize to one or more distant organ sites including one or more of the lymph nodes, skeletal muscle, heart, brain, liver, spleen, kidney, lung, or spinal cord. Moreover, the metastasis can occur quickly (e.g., within 2-6 weeks). In addition, metastasis may occur with high fidelity when an animal is administered or implanted with a cell or cells of the invention, with up to 85%, 90%, 95%, 96%, 97%, 98%, 99%, or 100% of animals administered or implanted with metastatic cells of the invention exhibiting metastasis.

Compositions that include one or more of the cell types described herein are also within the scope of the present invention. For example, the invention features compositions that include one or more of these cell types and a tissue culture medium. A cryoprotective agent may be included to protect the cells when they are frozen and stored, hi addition, the compositions can include agents that may affect the ability of the cells to proliferate and/or ^u '^! rfieySliiite'gJώό'M of potential chemotherapeutic agents or molecules that may elicit an immune response).

Other compositions include kits, which can include a first container comprising a metastatic cell described herein (e.g., a cell of the VM-M2 cell line or the VM-M3 cell line), and, optionally, instructions for use. In some aspects of the invention, a kit may include a first container comprising a cell of the VM-M2 cell line and a second container comprising a cell of the VM-M3 cell line and, optionally, instructions for use. A kit of the invention, may also include container that includes a non-metastatic cell described herein (e.g., a cell of the VM-NMl cell line), and optionally, instructions for use.

Cells of the invention may be "isolated" by virtue of their existence outside of the living organism in which they originated. Thus, an isolated cell can be a cell within a dissected tumor or a cell that has been removed from a host or donor organism and maintained in culture or implanted into a recipient organism. A given cell or a collection of cells may also be described as "purified" by virtue of being separated from one or more other types of cells. For example, the invention features cells obtained from a spontaneous brain tumor and those cells (the cells of interest) are purified when they constitute 50% or more of a collection of cells (e.g., 60%, 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or more, including each percentage in between). For example, a tumor cell or a plurality thereof obtained from the brain of a mouse (e.g., a VM mouse) is purified when the cell or cells constitute 50% or more of the cells collected from the brain.

As noted, the cells described herein can be used to study metastasis in vivo. Accordingly, other aspects of the invention feature nonhuman subjects that include one or more of the cell types described herein (e.g., a VM-M2 cell, a VM-M3 cell, or a VM-NMl cell, whether as deposited with the ATCC, further modified, or later isolated from a brain tumor). The subject can be, for example, an animal used in preclinical studies, such as a rodent [e.g., a guinea pig, rat, or mouse (e.g., a VM mouse), a cat, dog, or nonhuman primate], and may or may not be immunocompromised.

Methods of the invention can be carried out to identify a candidate chemotherapeutic agent. In one embodiment, the method can include steps of (a) providing a nonhuman subject; (b) administering to the subject a plurality of cells of a type described herein and a test compound; and (c) measuring a parameter indicative of cancer progression (e.g., the size, location, and/or number of tumors in the subject after a given time). A test compound that improves the outcome is a candidate chemotherapeutic agent. For example, a test compound that reduces the size and/or number of tumors in the subject or inhibits metastasis is a ^•canαϊdite cneino&erapeutic agent. The outcome in an animal or group of animals exposed to the test compound can be compared to a reference standard or "control", which could be suitably designed by one of ordinary skill in the art. Examples of reference standards are described further below. As with the models described above, the subject of the screen can be an animal used in preclinical studies, such as a rodent (e.g., a guinea pig, rat or mouse (e.g., a VM mouse), a cat, a dog, or a nonhuman primate), and may or may not be immunocompromised. Cells of the invention can be administered to the subject by any route of administration (e.g., subcutaneously or intravenously) and may or may not be administered at the same time or by the same route as the test compound is administered.

In some embodiments of the invention, metastatic cells (e.g., VM-M2, VM-M3, or cells having their characteristics) can be used in conjunction with non-metastatic cells (e.g., VM-NMl or cells having their characteristics) to identify chemotherapeutic (e.g., antimetastatic and/or antiproliferative) agents. Assays to identify chemotherapeutic agents may be conducted with cells of the invention in vitro and/or in vivo. Those of ordinary skill in the art will understand how to use routine culture methods to assay potential chemotherapeutic agents in vitro. For example, a cell of the invention, (e.g. a VM-M2 or VM-M3 cell) may be contacted in vitro with a candidate agent and then examined for metastatic characteristics and a non-metastatic cell of the invention (e.g., a MV-NMl cell) may be similarly contacted with the agent and the effect of the cells compared, hi some embodiments, a cell so treated may be administered to a non-human animal for assessment of metastasis, etc. In some embodiments, cells of the invention may be assayed in in vivo assays. For example, one can carry out a method having the following steps: (a) providing a first nonhuman subject; (b) administering to the first subject a plurality of cells of a metastatic (e.g., VM-M2 or VM-M3) cell line and a test compound; (c) providing a second nonhuman subject; (d) administering to the second subject a plurality of cells of a non- metastatic (e.g., VM-NMl) cell line and the test compound; and (e) measuring a parameter indicative of cancer progression (e.g., measuring the size and/or number of tumors in the first subject and the second subject). A test compound that reduces the size and/or number of tumors in the first subject relative to the second subject is a candidate antimetastatic agent.

Those of ordinary skill in the art will recognize that in certain embodiments of the invention, metastatic cells (e.g., VM-M2, VM-M3, or cells having their characteristics) can be used without use of non-metastatic cells (e.g., VM-NMl or cells having their characteristics) to identify chemotherapeutic (e.g., antimetastatic) agents. Assays to identify chemotherapeutic agents may be conducted with cells of the invention in vitro and/or in vivo.

art will understand how to use routine culture methods to assay agents in vitro. For example, a cell of the invention may be contacted in vitro with a candidate agent and then examined for metastatic characteristics. In some embodiments, a cell so treated may be administered to a non-human animal for assessment of metastases, etc. Cells of the invention may also be assayed in in vivo assays. A non-limiting example of an in vivo assay, may include (a) providing a non-human subject; (b) administering to the subject a plurality of cells of a metastatic (e.g., VM-M2 or VM-M3) cell line and a test compound; (c) measuring a parameter indicative of cancer progression (e.g., measuring the size and/or number of tumors) in the non-human subject to a control level of the parameter, wherein a test compound that reduces the size and/or number of tumors in the subject relative to the control level is a candidate antimetastatic agent. Similar methods to assay candidate chemotherapeutic agents may also be used that include contacting a non-metastatic cell or cell line of the invention (e.g., VM-NMl) and may include administering one or more non- metastatic cells of the invention to a subject for an in vivo assay.

Methods of the invention can also be carried out to induce a tumor in a non-human subject (e.g., a mouse or other animal used for preclinical research (examples of which are provided herein)). In this embodiment, one provides the subject and administers thereto one or more of the cell types described herein. Standard methods (e.g., palpation or image analysis (e.g., analysis of x-rays, MRIs, and the like)) can be used to detect and analyze tumor induction. The subject can then be exposed to potential therapeutic agents or otherwise tested or examined to identify the cause or study the course of the resulting tumors.

The details of one or more embodiments of the invention are set forth in the accompanying drawing and the description below. Other features, objects, and advantages of the invention will be apparent from the description and the drawing, and from the claims.

Description Of Drawings Fig. 1 is a reproduction of a high performance thin-layer chromatogram of gangliosides in cultured VM tumor cells. Lane 1 carries a mouse brain ganglioside standard; lane 2 carries a GM3 standard; lanes 3-7 carry radiolabeled gangliosides synthesized by cultured cells (lane 3, astrocytes; lane 4, macrophages; lane 5, VM-NMl; lane 6, VM-M2; and lane 7, VM-M3).

Fig. 2 is a reproduction of a gel showing the results of semi-quantitative reverse transcription-polymerase chain reaction (RT-PCR) to detect expression of macrophage ^"the control /3-actin in the cell lines VM-NMl, VM-M2, and VM- M3. Expression in astrocytes (Ast) and macrophages (MO) is presented as a control.

Fig. 3 is a reproduction of a gel showing the results of semi-quantitative reverse transcription-polymerase chain reaction (RT-PCR) to detect expression of macrophage markers F4/80 and Macl and the control /?-actin (bactin) in the cell lines VM-NMl (NMl), VM-M2 (M2), and VM-M3 (M3). Expression in astrocytes (AC) and macrophages (MO) is presented as a control.

Fig. 4 is a reproduction of four thin-layer chromatograms of (Fig. 4A) neutral lipids, (Fig. 4B) acidic lipids, (Fig. 4C) gangliosides, and (Fig. 4D) GAl in VM tumors and control cell lines. The abbreviations used on the HPTLC plates for the neutral and acidic lipid plates were as follows: CE, cholesterol esters; TG, triglycerides; IS, internal standard; C, cholesterol; CM, ceramide; CB, cerebrosides (doublet); PE, phosphatidylethanolamine; PC, phosphatidylcholine; SPM, sphingomyelin; O, origin; SF, solvent front of the first developing solvent system; C, cholesterol; CL, cardiolipin; PA, phosphatidic acid; SuIf, sulfatides (doublet); PS, phosphatidylserme; and PI, phosphatidylinositol.

Fig. 5 is a reproduction of two gels showing results of semi-quantitative RT-PCR to detect gene expression in cells grown under identical culture conditions. A battery of genes characteristic of macrophages were expressed in the VM-M2, VM-M3, and the RAW 264.7 cells, but their expression was undetectable in the VM-NMl or the AC cells (Fig. 5A). The expression genes characteristic of neural cells was undetectable in the VM-M2, VM-M3 and RAW 264.7 cells (Fig. 5B). Genes characteristic of neural progenitor/stem cells (nestin and SAT IT) were expressed in the VM-NMl cells. The AC cells expressed nestin and the gene characteristic of mature astrocytes (GFAP). The NF200 gene (expressed in mature neurons) was undetectable in all the cell lines. B designates control brain tissue that involved embryonic brain for nestin and SAT II and adult brain for GFAP and NF200. /3-actin was used a control.

Detailed Description

The invention is based, in part, on the generation of three cell lines from spontaneous brain tumors of inbred VM mice. Cells within two of these lines, the VM-M2 and VM-M3 tumor cell lines, are highly metastatic. Cells in these lines metastasize to organs in multiple organ systems (e.g., liver, spleen, kidney, lung, and brain, etc.) when grown subcutaneously outside the brain. Metastatic cells of the invention may be used as a model or to prepare animal models of metastatic cancer arising from various types of cancer, including, but not Y / E - ^CI» IΠLRΓ ^^{* 1}MI-Fi "S 1 "9 limited^' to^' biliary tract cancer, brain cancer (including glioblastomas and medulloblastomas), breast cancer, cervical cancer, choriocarcinoma, colon cancer, endometrial cancer, esophageal cancer, gastric cancer, intraepithelial neoplasms, including Bowen's disease and Paget's disease, liver cancer, lung cancer, lymphomas, including Hodgkin's disease and lymphocytic lymphomas, neuroblastomas, oral cancer, including squamous cell carcinoma, ovarian cancer, including those arising from epithelial cells, stromal cells, germ cells and mesenchymal cells, pancreatic cancer, prostate cancer, rectal cancer, renal cancer including adenocarcinoma and Wilms tumor, sarcomas, including leiomyosarcoma, rhabdomyosarcoma, liposarcoma, fibrosarcoma and osteosarcoma, skin cancer, including melanoma, Kaposi's sarcoma, basocellular cancer and squamous cell cancer, testicular cancer, including germinal tumors (seminomas, and non-seminomas such as teratomas and choriocarcinomas), stromal tumors and germ cell tumors, thyroid cancer, including thyroid adenocarcinoma and medullary carcinoma; lymphoid disorders including but not limited to acute lymphocytic leukemia and chronic lymphoproliferative disorders (e.g., lymphomas, myelomas, and chronic lymphoid leukemias); lymphomas, which may include, but are not limited to Hodgkin's disease and non-Hodgkin's lymphoma; chronic lymphoid leukemias, which may include T cell chronic lymphoid leukemias and B cell chronic lymphoid leukemias; myeloid disorders, which may include, but are not limited to chronic myeloid disorders such as, for instance, chronic myeloproliferative disorders, myelodysplastic syndrome and acute myeloid leukemia. In contrast, cells of the VM-NMl tumor cell line do not observably metastasize when grown subcutaneously. The invention includes these cell lines and cells with similar characteristics as described herein, as well as methods of using them in animal models of cancer and/or metastasis and methods of using them to aid in identifying candidate chemotherapeutic agents (e.g., antimetastatic agents and/or antiproliferative agents).

Cells

Cells with high metastatic potential have been discovered and isolated from VM mouse brain tumors. The VM-M2 and VM-M3 cells, when implanted subcutaneously into a VM mouse, spontaneously metastasize to one or more distant organ sites (e.g., brain, liver, spleen, kidney, or lung, etc.). Metastasis is efficient and rapid, resulting in detectable metastases in nearly all implanted mice within 2-6 weeks. The cells have been characterized based on their lipid content, which included assessment of gangliosides, neutral lipids and acidic lipids, and according to the expression of certain genes. With respect to ganglioside cϊbnljpόfitϊbπ, the fcβlls express low levels of the gangliosides GM3 and GM2. The amount of GM3 can be about 1% or less of the total ganglioside content, and the amount of GM2 can be about 5% or less of the total ganglioside content.

Gangliosides are sialic acid-containing glycosphingolipids that are ubiquitously distributed on the plasma membranes of vertebrate cells. Gangliosides are synthesized in the Golgi apparatus by the sequential transfer of carbohydrate residues onto a ceramide lipid anchor. A combinatorial biosynthetic pathway results in a large diversity of oligosaccharide structures on gangliosides. GM3 ganglioside is a key structure because it contains the simplest ganglioside oligosaccharide in the pathway (it is a trisaccharide composed of glucose, galactose, and sialic acid) and serves as a precursor for most of the more complex ganglioside species.

Gangliosides and glycosphingolipids can be analyzed by the methods described in El- Abbadi et al. (2001) Brit. J. Cancer 85:285-92. Briefly, tumor or cell samples are frozen at — 2O⁰C and lyophilized to remove water. Gangliosides are isolated and purified as described in Seyfried et al. (1978) J. Neurochem. 31:21-27 and Seyfried et al (1987) Cancer Res. 47:3538-42. The purified gangliosides are then examined using high performance thin-layer chromatography (HPTLC) (Whatman HPK silica gel) as described in Ando et al. (1978) Anal. Biochem. 89:437-50. Briefly, the samples are spotted on the plate and developed with chloroform:methanol:water (55:45:10) (v/v/v) containing 0.02% CaCl₂-2H₂O. After development, ganglioside bands are visualized by spraying with resorcinol. The ganglioside content of cell lines can be analyzed by growing the cells in medium containing radiolabeled galactose (e.g., ¹⁴C galactose), purifying the gangliosides from the cells, and performing HPTLC as described above, followed by autoradiography of the HPTLC plate (El-Abbadi et al. (2001) Brit. J. Cancer 85:285-92). Ganglioside content can be quantified by densitometric analysis or by Phosphorimager3. One of ordinary skill in the art could readily subject any given population of cells to this procedure to determine the ganglioside content.

With respect to gene expression, the results indicate that the metastatic cells have characteristics of microglia or macrophages. The cells express high levels of CD68, Macl, F4/80, Iba-1 proteins, CD45, CDl Ib, CXCR4, and asialo GMl, a marker for activated macrophages (Ecsedy et al. (1998) J. Lipid Res. 39:2218-2227). Expression of these proteins and related mRNAs (i.e., an mRNA encoding a given protein) can be determined by any means known in the art (e.g., immunoblotting, enzyme-linked immunosorbent assay (ELISA), radioimmunoassay (RIA), immunoprecipitation, peptide arrays, Northern blotting, reverse transcription-polymerase chain reaction (RT-PCR), primer extension, serial analysis of gene" expression tSϊtGE)¹, or nucleotide microarrays). See, e.g., Current Protocols in Molecular Biology, Ausubel et al, eds., John Wiley & Sons, Inc. (2005). Expression of asialo GMl can be determined by the methods for analyzing gangliosides and glycosphingolipids described above. '

VM-NMl cells have also now been discovered and isolated from a VM mouse brain tumor. Unlike the VM-M2 and VM-M3 cells, there is no indication that these cells metastasize when grown subcutaneously. This property makes VM-NM 1 cells useful as a negative control for metastasis relative to the metastatic VM brain tumor cells described herein and useful for study of cancer and tumor growth and treatment.

Cells that were initially isolated and that have since been established as cell lines and deposited with the ATCC as well as later-isolated cells having the characteristics of those described herein can be further modified in a variety of ways. For example, they can be used to generate hybrid cells, which can be produced by nuclear transfer or cellular fusion. Where nuclear transfer is employed, the nucleus of a first cell is removed and replaced with the nucleus of a second cell (Mullins et al. (2004) J. Physiol. 554:4-12). Hybrid cells include cells from which either the cytoplasm or the nucleus is that of a cell described herein. For example, a hybrid cell can include the nucleus of a cell described herein {e.g., a VM-M2, VM-M3, or VM-NMl cell) and the cytoplasm of a second, heterologous cell (i.e., a cell other than VM-M2 where the nucleus was donated by a VM-M2 cell; a cell other than VM-M3 where the nucleus was donated by a VM-M3 cell; or a cell other than VM-NMl where the nucleus was donated by a VM-M3 cell). Similarly, a hybrid cell can also include the cytoplasm of a cell described herein (e.g., a VM-M2 cell, a VM-M3 cell, or a VM-NMl cell) and the nucleus of a second, heterologous cell.

Cells described herein can also be modified by cellular fusion. For example, a cell described herein can be fused with another cell type to produce a hybridoma. These hybridomas include nuclear fusion events when the nucleus, together with cytoplasm, fuse with nucleus of a second cell to form a "fusion hybrid".

Other modifications can produce mutant or transgenic cells. For example, cells described herein (e.g., VM-M2 or VM-M3 cells) can be mutagenized by exposure to a chemical mutagen (e.g., ethylmethane sulfonate (EMS)) or ionizing radiation. Transgenic cells can be produced by introducing a nucleic acid sequence or "transgene" into one of the cell types described herein. The transgene can be introduced by any method known in the art (e.g., by transfection or transformation method) and can direct the disruption or expression of a specific gene or biologically active nucleotide sequence within the cell. For example, the ^['"' tfansgine cin^' direCΪ the expression of an antisense oligonucleotide or a sequence that mediates RNAi (e.g., an siRNA or shRNA). The expression can be transient or non-transient. In some instances, the transgene will be selected based on its ability to facilitate or inhibit a cellular function, hi that case, the transgene can be used or screened as a chemotherapeutic (e.g., antimetastatic) agent according to the methods described herein.

Cells of the invention can also be modified by the inclusion of a detectable label (or "tag" or "marker"). For example, the cells can be modified to include a transgene that directs the expression of a cellular marker (e.g., a detectable label), such as an enzyme (e.g., horseradish peroxidase, chloramphenicol acetyltransferase, or /3-galactosidase), a fluorescent protein (e.g., green fluorescent protein (GFP), red fluorescent protein (RFP), yellow fluorescent protein, cyan fluorescent protein, blue fluorescent protein, enhanced-green fluorescent protein (EGFP), etc., or a luminescent protein,(e.g., luciferase, including, but not limited to firefly luciferase, Renilla luciferase, or a Genji-botaru luciferase), etc. Detectable labels and suitable detection methods may also be used to characterize tumors and metastases of cells of the invention following administration and/or implantation into a subject. For example, cells of the invention maybe detectably labeled and their location and metastatic spread may be determined using standard methods such as histology and imaging. Imaging methods may include, but are not limited to in vitro imaging, in vivo imaging. In some embodiments, in vivo imaging may be real-time imaging. In some embodiments, a cell of the invention may be a genetically modified cell that expresses a detectable label.

Tumor Induction

Cells and methods described herein can be used to induce tumors (e.g., metastatic tumors or non-metastatic tumors) in nonhuman subjects. To induce tumors, cells are administered to a non-human subject, in which they will subsequently divide to form tumors (e.g., metastatic tumors) at various locations in the subject.

Cells of the invention can be implanted into an immunocompromised animal (e.g., an immunocompromised mouse (e.g., a nude mouse, a severe combined immunodeficiency (SCID) mouse, or an SCID-Beige mouse, etc.) (Clarke (1996) Breast Cancer Res. Treat. 39:69-86). One may select an immunocompromised animal to allow for tumor growth when the mouse cells described herein are transplanted into an animal of another species.

Cells of the invention can also be administered to a VM mouse. As noted, cells of the VM-M2, VM-M3, and VM-NMl cell lines were derived from mice of the inbred VM strain, and can therefore grow and form tumors in VM mice with intact immune systems. The ^"F /"Jl 1 ""1( 11^"It IKi ^•'" H ^κit "1 ^"I ¹Rl ' a'bility to produce tunrof s^'that metastasize reproducibly in mice with healthy immune systems will allow for the investigation of the role of the immune system in metastasis and cancer growth.

The cells can be implanted into a specific location in an animal (e.g., intracranially, intraperitoneally, within an organ other than the brain, or subcutaneously). Cells of the invention can be implanted directly from culture, or they can be implanted as a segment of a tumor dissected from a mouse. Exemplary methods of implantation of isolated cells and tumor segments to the flank and brain of mice are given in Examples 1 and 2 herein. The VM-M2 and VM-M3 cells form metastatic tumors when implanted in the flank of VM mice. When implanted orthotopically, the VM-M2 and VM-M3 cells invade all areas of the brain with possible metastasis to non-neural tissues. The VM-NMl cells form tumors in brain and flank, when implanted into the brain and flank, respectively, but exhibit no detectable metastasis.

Alternatively, cells described herein can be introduced directly to the systemic circulation of the animal. This method can also be used to observe the formation of tumors at various sites within the subject. Typically, the tumor cells are injected into a large vein of the animal (e.g., the tail vein of a mouse).

To characterize the tumors and metastases (e.g., following implantation or administration of the cells described herein, complete necropsies can be performed on the tumor-bearing animals. Hematoxylin and eosin (HE) staining can be used to assess the histological appearance of primary and metastatic lesions. The expression of protein markers (e.g., markers of neuroectodermal, neuronal, glial, microglial, and lymphoid cells) can be evaluated in lesions. An exemplary neuroectodermal marker is S-IOO (Kahn et al. (1983) Am. J. Clin. Pathol. 79:341-347); exemplary neuronal markers are neuron-specific enolase (NSE) (Vinores et al. (1984) Arch. Pathol. Lab. Med. 108:536-540) and neurofibrillary protein (NFP) (Schlaepfer (1987) J. Neuropathol. Exp. Neurol. 46:117-129); an exemplary glial marker is glial fibrillary acidic protein (GFAP) (Eng (1985) J Neuroimmunol. 8:203- 214); exemplary microglial markers are CD68 and Ibal (Drage et al. (2002) J. Neurocytology 31: 681-92); exemplary lymphoid markers are leukocyte common antigen (LCA) (Powers et al. (1992) J Neuropathol. Exp. Neurol. 51:630-643), Mac-ll (Sanchez Madrid et al. (1983) J. Exp. Med. 158:586-602), B cell surface glycoprotein (B 220) (Coffhian and Weissman (1981) Nature 289:681-683), and intercellular adhesion molecule (ICAM-I) (Said et al. (1979) Cancer 44:504-528). ' C T S/cre Uen Sing CI MJBet /ho «4ds*5 an JdL 1 Te. «s3t C _ompound ,s

Cells of the invention are useful in methods of identifying candidate chemotherapeutic agents (e.g., anti-metastatic agents, etc.) by examining the effect of a test compound has on the development of tumors (e.g., metastatic tumors) in nonhuman subjects following administration of one or more types of cells (or pluralities thereof). Examples of types of chemotherapeutic agents that may be tested include, but are not limited to: anti- metastatic agents, cytotoxic agents, cytostatic agents, cytokine agents, anti-proliferative agents, immunotoxin agents, gene therapy agents, angiostatic agents, cell targeting agents, etc.). A test compound can be administered before cells of the invention are administered; at the same or about the same time as cells of the invention are administered, or after cells of the invention are administered. Cells of the invention and test compound(s) can be administered by the same or different routes. The effect on tumor development can be assessed by determining whether the test compound reduces the size, location, and/or number of tumors in the subject. If so, or if another clinically beneficial result is obtained, the test compound is a candidate chemotherapeutic agent (e.g., a candidate anti-metastatic agent, etc.). Other clinically beneficial results include: (a) inhibition or arrest of primary tumor growth, (b) inhibition of any metastatic dissemination or spread to distant organs, and (c) extension of survival of nonhuman host.

The following provides examples of test compounds and is not meant to be limiting. Those of ordinary skill in the art will recognize that there are numerous additional types of suitable test compounds that may be tested using the methods, cells, and/or animal models of the invention. Test compounds can be small molecules (e.g., compounds that are members of a small molecule chemical library). The compounds can be small organic or inorganic molecules of molecular weight below about 3,000 Daltons. The small molecules can be, e.g., from at least about 100 Da to about 3,000 Da (e.g., between about 100 to about 3,000 Da, about 100 to about 2,500 Da, about 100 to about 2,000 Da, about 100 to about 1,750 Da, about 100 to about 1,500 Da, about 100 to about 1,250 Da, about 100 to about 1,000 Da, about 100 to about 750 Da, about 100 to about 500 Da, about 200 to about 1500, about 500 to about 1000, about 300 to about 1000 Da, or about 100 to about 250 Da). Test compounds can also be microorganisms, such as bacteria (e.g., Escherichia coli, Salmonella typhimurium, Mycobacterium avium, or Bordetella pertussis), fungi, andprotists (e.g., Leishmania amazonensis), which may or may not be genetically modified. See, e.g., U.S. Patents No. 6,190,657 and 6,685,935 and U.S. Patent Applications No. 2005/0036987 and 2005/0026866. The small molecules can be natural products, synthetic products, or members of a combinatorial chemistry library. A set of diverse molecules can be used to cover a variety of functions such as charge, aromaticity, hydrogen bonding, flexibility, size, length of side chain, hydrophobicity, and rigidity. Combinatorial techniques suitable for synthesizing small molecules are known in the art (e.g., as exemplified by Obrecht and Villalgrodo, Solid- Supported Combinatorial and Parallel Synthesis of Small-Molecular-Weight Compound Libraries, Pergamon-Elsevier Science Limited (1998)), and include those such as the "split and pool" or "parallel" synthesis techniques, solid-phase and solution-phase techniques, and encoding techniques (see, for example, Czarnik, A. W., Curr. Opin. Chem. Biol. (1997) 1:60). In addition, a number of small molecule libraries are publicly or commercially available (e.g., through Sigma- Aldrich, TimTec (Newark, DE), Stanford School of Medicine High- Throughput Bioscience Center (HTBC), and CheniBridge Corporation (San Diego, CA).

Compound libraries screened using the new methods can comprise a variety of types of test compounds. A given library can comprise a set of structurally related or unrelated test compounds. In some embodiments, the test compounds are peptide or peptidomimetic molecules. Li some embodiments, test compounds include, but are not limited to, peptide analogs including peptides comprising non-narurally occurring amino acids, e.g., /3-amino acids or /3-substituted /3-amino acids ("/3³ -amino acids"), phosphorous analogs of amino acids, such as α-aminophosphonic acids and α-aminophosphinic acids, or amino acids having non- peptide linkages, or other small organic molecules. In some embodiments, the test compounds are jS-peptide molecules; peptidomimetics (e.g., peptoid oligomers, e.g., peptoid amide or ester analogues, /?-peptides, D-peptides, L-peptides, oligourea or oligocarbamate); peptides (e.g., tripeptides, tetrapeptides, pentapeptides, hexapeptides, heptapeptides, octapeptides, nonapeptides, decapeptides, or larger, e.g., 20-mers or more); cyclic peptides; other non-natural or unnatural peptide-like structures; and inorganic molecules (e.g., heterocyclic ring molecules). Test compounds can also be nucleic acids.

The test compounds and libraries thereof can be obtained by systematically altering the structure of a first "hit" compound that has a chemotherapeutic (e.g., anti-metastatic) effect, and correlating that structure to a resulting biological activity (e.g., a structure-activity relationship study).

Such libraries can be obtained using any of the numerous approaches in combinatorial library methods known in the art, including: biological libraries; peptoid libraries (libraries of molecules having the functionalities of peptides, but with a novel, non-peptide backbone which are resistant to enzymatic degradation but which nevertheless remain bioactive; see, e.g.,

37:2678-85 (1994)); spatially addressable parallel solid phase or solution phase libraries; synthetic library methods requiring deconvolution; the "one- bead one-compound" library method; and synthetic library methods using affinity chromatography selection (Lam, Anticancer Drug Des. 12:145 (1997)). Examples of methods for the synthesis of molecular libraries can be found in the art, for example in: DeWitt et al, Proc. Natl. Acad. Sci. USA, 90:6909 (1993); Erb et al, Proc. Natl. Acad. ScL USA, 91:11422 (1994); Zuckermannet ah, J. Med. Chem., 37:2678 (1994); Cho et al., Science, 261:1303 (1993); Carrell et al, Angew. Chem. Int. Ed. Engl, 33:2059 (1994); Carell et al, Angew. Chem. Int. Ed. Engl, 33:2061 (1994); and in Gallop et al, J. Med. Chem., 37:1233 (1994). Libraries of compounds can be presented in solution {e.g., Houghten (1992) Biotechniques, 13:412-421), or on beads (Lam (1991) Nature, 354:82-84), chips (Fodor (1993) Nature, 364:555-556), bacteria (Ladner, USP 5,223,409), spores (Ladner, U.S. Patent No. 5,223,409), plasmids (Cull et al (1992) Proc. Natl Acad. Sci. USA, 89:1865-1869) or on phage (Scott and Smith (1990) Science, 249:386-390; Devlin (1990) Science, 249:404-406; Cwirla et al (1990) Proc. Natl. Acad. Sci. USA, 87:6378-6382; Felici (1991) J. MoI Biol, 222:301-310; Ladner, supra.).

Small molecules identified as having a chemotherapeutic or anti-metastatic effect can be selected and systematically altered, e.g., using rational design, to optimize binding affinity, avidity, specificity, or other parameters. Such optimization can also be screened for using the methods described herein. Thus, one can screen a first library of small molecules using the methods described herein, identify one or more compounds that are "hits," (by virtue of, for example, their ability to reduce the size and/or number of tumors, e.g., at the original site of surgical implantation and at metastasis sites), and subject those hits to systematic structural alteration to create a second library of compounds structurally related to the hit. The second library can then be screened using the methods described herein.

A variety of techniques useful for determining the structures of compounds are known and can be used in the methods described herein {e.g., NMR, mass spectrometry, gas chromatography equipped with electron capture detectors, fluorescence, and absorption spectroscopy).

Assays of chemotherapeutic activity of test compounds maybe conducted in vitro and/or in vivo using cells of the invention. For example, a test compound may be administered to a nonhuman subject to which has been administered {e.g., implanted or injected with) a plurality of the cells described herein, e.g., a number of cells sufficient to induce the formation of one or more tumors {e.g., metastatic tumors). The nonhuman subject ' can be; e.g., a. rodent (e.g., a mouse, e.g., a VM mouse). The test compound can be administered to the subject by any regimen known in the art. For example, the test compound can be administered prior to, concomitant with, and/or following the administration of cells of the invention. A test compound can also be administered regularly throughout the course of the method, for example, one, two, three, four, or more times a day, weekly, bi-weekly, or monthly, beginning before or after cells of the invention have been administered. In other embodiments, the test compound is administered continuously to the subject (e.g., intravenously). The dose of the test compound to be administered can depend on multiple factors, including the type of compound, weight of the subject, frequency of administration, etc. Determination of dosages is routine for one of ordinary skill in the art. Typical dosages are 0.01-200 mg/kg (e.g., 0.1-20 or 1-10 mg/kg).

The size and/or number of tumors in the subject can be determined following administration of the tumor-forming cells and the test compound. The size and/or number of tumors can be determined non-invasively by any means known in the art. For example, tumor cells that are fluorescently labeled (e.g., by expressing a fluorescent protein such as GFP) can be monitored by various tumor-imaging techniques or instruments, e.g., noninvasive fluorescence methods such as two-photon microscopy. The size of a tumor implanted subcutaneously can be monitored and measured underneath the skin. Typically, the size and/or number of tumors in the subject is determined by necropsy following a set period of time to allow the cells to develop tumors. Methods of necropsy to identify and characterize tumors are described herein.

To determine whether a compound affects tumor formation or metastasis, the size and/or number of tumors in the subject can be compared to a reference standard (e.g., a control value). A reference standard can be a control subject which has been subject to the same regimen of administration of administration of tumor-forming cells and test compound, except that the test compound is omitted or administered in an inactive form. Alternately, a compound believed to be inert in the system can be administered. The reference standard can also be a numerical figure or figures representing the size and/or number of tumors expected in an untreated subject. This numerical figure(s) can be determined by observation of a representative sample of untreated subjects. A reference standard may also be the test animal before administration of the compound.

In one aspect of the invention, a metastatic cell line (e.g., VM-M2 or VM-M3) is administered to a subject along with a test compound, as described above. A second, non- metastatic cell line is administered to the same subject (e.g., at a second site) or to a second

The effect of the test compound on metastasis, separate from general antitumor effects, can be characterized by observing the effect on the size and/or number of tumors derived from the metastatic cells compared to the size and/or number of tumors derived from the non-metastatic cells. A test compound that reduces the size and/or number of tumors derived from the metastatic cells compared to the non- metastatic cells is a candidate anti-metastatic agent.

In some aspects of the invention, a metastatic cell line (e.g., VM-M2 or VM-M3) is administered to a subject and a test compound is also administered to the subject (as described above) and the effect of the test compound on tumor growth and metastases can be characterized by observing the effect on the size and growth and dispersion of tumors derived from the metastatic cell. A test compound that reduces the size, dispersion, and/or growth of a tumor derived form the metastatic cell is a candidate agent for treating cancer. m some aspects of the invention, a non-metastatic cell line (e.g., VM-NMl) is administered to a subject and a test compound is also administered to the subject (as described above) and the effect of the test compound on tumor growth and metastases can be characterized by observing the effect on the size and growth of tumors derived from the non- metastatic cell. A test compound that reduces the size or growth of a tumor derived form the non-metastatic cell is a candidate agent for treating cancer. hi some aspects of the invention, a cell or cells from a metastatic cell line (e.g., VM- M2 or VM-M3) or a non-metastatic cell line (e.g., MV-NMl) may be contacted in vitro with a test compound in an assay to determine whether the compound has an effect on cell metastasis or cell growth. In some embodiments, such a cell may be assessed in vitro to determine whether the compound can alter metastatic characteristics of the cell(s) or if it can alter tumor growth. In some embodiments, high-throughput screening methods may be used for in vitro assays using cells of the invention. In some aspects of the invention, a cell or cells of the invention that have been contacted in vitro with a test compound may be administered to a test subject to determine whether the amount, speed, or extent of metastasis of the cell(s) and/or tumor growth have been modified by contact with the test compound. For such an assay, the level, amount, speed, or extent, etc., of metastasis and/or tumor growth may be compared to a control level, e.g., a level in a subject to which a similar amount of the same type of cells have been administered, but wherein the cells were not contacted with the test compound. If the amount, speed, extent, etc. of metastasis and/or tumor growth is reduced in the test subject compared to the control amount, speed, or extent, etc. of metastasis IT an -'"d 1/o 1f S "tu IDmfo IBr g /ro 4w"tBh, J^'iϊ . L is a '9h indication that the test compound may be a chemotherapeutic agent for the treatment of cancer and/or metastasis in cancer.

Examples Example 1 : Establishment of the VM-M2 Cell Line

AVM mouse bearing a spontaneous brain tumor was sacrificed, and the tumor removed with all connective tissue. The tumor was minced in ice cold, sterile phosphate buffered salts (PBS) into ~1 mm³ pieces. The tumor pieces were implanted into five additional VM mice. Each mouse was anesthetized with a peritoneal injection of tribromomethanol, and one of the 1 mm³ pieces was inserted into the trocar through a whole drilled in the skull above the right cerebral hemisphere. The remaining tumor pieces were frozen for further study.

The five mice with implanted tumor pieces all developed brain tumors. These mice were sacrificed, and the tumors minced and implanted into VM mice as above. Once these mice developed brain tumors, the tumors were again excised, minced, and then implanted into the flank of VM mice. For flank implantation, recipient VM mice were anesthetized with isofluorane, and 0.2 ml of tumor chunks suspended in 0.3 ml PBS were injected into the right flank of the animals using a 1-ml syringe.

A tumor was dissected from the flank of a recipient mouse, and placed in ice-cold, sterile PBS. The tumor was rinsed with ice-cold, sterile PBS and transferred to a Petri dish containing 5 ml of Dulbecco's Modified Eagle Medium (DMEM) with 10% fetal bovine serum (FBS). The tumor was minced into small pieces in the medium using sterile scissors, and the medium containing the tumor pieces was transferred to a cell culture flask. The cells were grown at 37°C, and the growth medium was changed every other day. When the cells reached confiuency, they were trypsinized and frozen in DMEM with 10% dimethylsulfoxide (DMSO).

A similar procedure was used to derive the VM-M3 and VM-NMl cell lines from two independent spontaneous VM mouse brain tumors.

Example 2: Re-implantation of Cells into VM Mice

VM-M2 cells were grown in DMEM medium with 10% (v/v) FBS to near confiuency. The cells were trypsinized, resuspended in 8 ml of DMEM, and centrifuged for 5 minutes at 2000 rpm. The cells were resuspended and counted using a hemocytorneter. About 2^χ 10⁶ d^'ellk wife ciritrifuged¹ as above, and resuspended in 0.3 ml of sterile, ice cold PBS. The cells were then injected subcutaneously into the flank of a VM mouse.

For brain implantation, VM-M2 cells were trypsinized and counted, and about 2x10⁶ cells were centrifuged as above. The cell pellet was inserted into the trocar of a VM mouse through a hole drilled in the skull above the right cerebral hemisphere.

Tumors from these mice were dissected as in Example 1 for analysis, cell line derivation, and/or re-implantation.

Example 3: Development of Metastatic Tumors

The flank-derived tumors from Example 1 were used to study the metastatic potential of the VM-M2, VM-M3, and VM-NMl cells. Tumors were excised from the flank as described, minced, and then implanted into the flank of VM mice. The recipient VM mice were anesthetized with isofluorane, and 0.2 ml of tumor chunks suspended in 0.3 ml PBS were injected into the right flank of the animals using a 1-ml syringe. The developing tumors were observed over time. At times post-implantation, the mice were sacrificed and necropsies were performed to characterize the VM tumor and its metastatic involvement. The composite results of multiple experiments are depicted in Table 1.

Table 1. Metastatic sites for naturally occurring VM tumors

Lymph Heart Skeletal Spinal

Tumor Lung Spleen Liver Kidney Brain Node Muscle Cord

VM-M2 + + + + + + + + +

VM-M3 + + + + + + + + +

VM-NMl — — — - — — — — —

" + and — indicate the presence or absence of tumor cells, respectively.

Nearly 100% of animals implanted with VM-M2 or VM-M3 tumors exhibited some metastasis, with multicentric nodal presentation the most common. The metastases were typically characterized by systemic/disseminated spread. The observed latency for metastasis was 4-6 weeks for VM-M2 and 3-4 weeks for VM-M3. Metastases within the brain were frequently observed before primary tumor morbidity. No metastasis was observed in animals implanted with VM-NMl tumors. Ex^'am^'ple k: lipid Analysis of Cultured Cells

To analyze the ganglioside content of the VM-M2, VM-M3, and VM-NMl cells, the cells were grown for 48 hours in DMEM containing ¹⁴C-galactose. Astrocyte and macrophage cells were grown concurrently in DMEM with ¹⁴C-galactose as controls. After isolation and purification of labeled cell gangliosides (Seyfried et al. (1978) J. Neurochem. 31:21-27; Seyfried et al. (1987) Cancer Res. 47:3538-42), an aliquot containing approximately 3000 disintegrations per minute was spotted on an HPTLC plate along with unlabelled ganglioside standards. The plate was developed with chloroform:niethanol:water (55:45:10) (v/v/v) containing 0.02% CaCl₂-2H₂O ,exposed to medical x-ray film to visualize ¹⁴C-labeled gangliosides, and then sprayed with resorcinol reagent to identify lanes with unlabelled ganglioside standards. The results are shown in FIG. 1. Metastatic VM-M2 and VM-M3 cells showed ganglioside patterns similar to that of macrophages, with low levels of GM2 and almost no detectable GM3. The non-metastatic VM-NMl cells expressed high levels of gangliosides GM3 and GDIa.

Example 5: Gene Expression Analysis of Cultured Cells

The metastatic cell lines VM-M2 and VM-M3 expressed markers characteristic of macrophages. For the expression analysis, total RNA was isolated using TRIZOL3 Reagent (hivitrogen, Carlsbad, CA) from cell pellets of cell lines VM-M2, VM-M3, and VM-NMl₃ as well as isolated astrocytes and macrophages. The concentration and purity of the total RNA was assessed by absorbance at 260 nm and 280 nm. Single-stranded cDNA was made by reverse transcription of 3 μg of total RNA using oligo (dT) primers (Promega, Madison, WI) in a 20 Tl reaction with Moloney murine leukemia virus reverse transcriptase (Promega) according to the manufacturer's protocol. The reverse transcribed cDNA was then amplified by polymerase chain reaction (PCR) with Tag polymerase (Promega) and 2 mM MgCl₂ using the conditions described below for each primer set. Gradient PCR was performed initially to determine the optimal annealing temperature for each primer set. Amplification of /5-actin cDNA was performed with primers Beta Actin F207 (5'-tgtgatggtgggaatgggtcag-3'; SEQ ID NO:1) and Beta Actin R719 (5'-tttgatgtcacgcacgatttcc-3⁵; SEQ ID NO:2). The PCR protocol was an initial denaturation at 94 °C for 2 minutes, followed by 22 cycles of denaturation at 94 ⁰C for 1 minute; annealing at 58 °C for 35 seconds; and extension at 72°C for 1 minute. A final extension at 72 °C for 6 minutes followed the last cycle.

Amplification of CD68 cDNA was performed with primers CD68 F95 (5'-catccttcacgatgacacctacag-3'; SEQ ID NO:3) and CD68 R638 (5^cfctgatgtaggtcctgllfgaatc-3'; SEQ ID NO:4). The PCR protocol was an initial denaturation at 94 °C for 2 minutes, followed by 27 cycles of denaturation at 94 °C for 1 minute; annealing at 65 ⁰C for 35 seconds; and extension at 72 °C for 1 minute. A final extension at 72 °C for 6 minutes followed the last cycle.

Amplification of Ibal cDNA was performed with primers IbalF50

(5'-aggcccagcaggaagagagg-3'; SEQ DD NO:5) and IbalR443 (5'-cagggcagctcggagatagc-3'; SEQ ID NO:6). The PCR protocol was an initial denaturation at 94 °C for 2 minutes, followed by 34 cycles of denaturation at 94 °C for 1 minute; annealing at 66 °C for 35 seconds; and extension at 72 °C for 1 minute. A final extension at 72 °C for 6 minutes followed the last cycle.

Amplification of Macl cDNA was performed with primers MAC1F1700 (5'-acagccaccggatcataggc-3'; SEQ ID NO:7) and MAC1R2256 (5'-cagaactggtcggaggttcc-3'; SEQ ID NO:8). The PCR protocol was an initial denaturation at 94 °C for 2 minutes, followed by 28 cycles of denaturation at 94 ⁰C for 1 minute; annealing at 65 °C for 35 seconds; and extension at 72 °C for 1 minute. A final extension at 72 °C for 6 minutes followed the last cycle.

Amplification of F4/80 cDNA was performed with primers F4_80F1777 (5'-cctatctgtgtctcctggaac-3'; SEQ ID NO:9) and F4_80R2271 (5'-gtgcagcatcttgatgttgcg-3'; SEQ ID NO: 10). The PCR protocol was an initial denaturation at 94 ⁰C for 2 minutes, followed by 27 cycles of denaturation at 94 ⁰C for 1 minute; annealing at 62 °C for 35 seconds; and extension at 72 °C for 1 minute. A final extension at 72 °C for 6 minutes followed the last cycle.

The PCR products were visualized by separation on a 1% agarose gel containing ethidium bromide. FIG. 2 shows the results for Ibal and CD68, with β-actin as a control. FIG. 3 shows the results for F4/80 and Macl, again with /5-actin as a control. Both the macrophages and the metastatic cell lines (VM-M2 and VM-M3) expressed all four macrophage markers tested (Ibal, CD68, F4/80, and Macl). By contrast, none of these genes were expressed detectably in astrocytes or the non-metastatic cell line VM-NMl. These results demonstrate that the metastatic cell lines described herein share at least some gene expression features with macrophages.

Example 6: Novel Metastatic Mouse Tumor Cells Express Multiple Properties of Macrophages. Metastasis involves the detachment of tumor cells from the primary neoplasm and their invasion of surrounding tissues and distant organs. Conventional therapies are largely ineffective against tumor metastasis, the primary cause of morbidity and mortality for cancer patients. This has been due in large part to the absence of in vivo metastatic models that represent the full spectrum of metastatic disease. Here it is described two new autochthonously arising tumors in the inbred VM mouse strain, which reliably express all of the major biological processes of metastasis to include local invasion, intravasation, immune system survival, extravasation, and secondary tumor formation involving liver, kidney, spleen, lung, and brain. The metastatic VM tumor cells also expressed multiple properties of macrophages including morphological appearance, surface adhesion, phagocytosis, total lipid composition (glycosphingolipids and phospholipids), and gene expression (CDlIb, Ibal, F4/80, CD68, CD45, and CXCR4). These findings support previous observations with human metastatic tumors and present the evidence that metastatic cancer cells can arise from macrophages or macrophage-like cells.

Methods Mice

Mice of the VM/Dk (VM) strain were obtained as gifts from G. Carlson (McLaughlin Research Institute, Great Falls, Montana) and H. Fraser (University of Edinburgh, Scotland). All VM mice used in this study were housed and bred in the Boston College Animal Care Facility using husbandry conditions as described previously (Ranes, M.K. et al. (2001) Br J Cancer 84: 1107-1114). All animal procedures were in strict accordance with the NIH Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care Committee.

Intracerebral and subcutaneous tumor implantation

The VM tumors were implanted either i.e. or s.c. as previously described Ranes, M.K. et al. (2001) Br J Cancer 84: 1107-1114.). Briefly, VM mice were anaesthetized with Avertin (O.lmL/ 1Og body weight) and small tumor pieces (1 mm³, estimated using a 1 mm x 1 mm grid) from a donor VM tumor were implanted into the right cerebral hemisphere of anesthetized recipient VM mice using a trocar. For s.c. implantation, VM mice were anaesthetized with Isoflurane (Halocarbon, River Edge, NJ), and small tumor pieces were ' ^ILl"

mice recovered from their surgical procedure and were returned to their cages when they became fully active.

Histology

Brain tumor samples were fixed in 10% neutral buffered formalin (Sigma) and embedded in paraffin. The brain-tumor samples were sectioned at 5 μm, were stained with haematoxylin and eosin (H & E) at the Harvard University Rodent Histopathology Core Facility (Boston, MA), and were examined by light microscopy as previously described (Mukherjee, P. et al. (2002) Br J Cancer 86: 1615-1621).

Metastasis analysis

Metastatic spread of VM tumors was analyzed in host VM mice following s.c implantation of VM tumor pieces as described above. Metastasis to liver, spleen, and kidney was confirmed by visual inspection of these organs for the presence of gross tumor nodules. In tissues where gross nodules were not visually apparent (e.g. brain) metastasis to that tissue was confirmed by the formation of a tumor following s.c. implantation of that tissue in additional VM mice. To exclude the possibility that tumors formed from circulating tumor cells entrapped within the tissues, animals were perfused with phosphate buffered saline (PBS) prior to tissue collection.

Cell lines and culture conditions

Cell lines were established from each flank-grown VM tumor as previously described (Seyfried, T.N. et al. (1992) MoI Chem Neuropathol 17: 147-167). The macrophage RAW 264.7 and the astrocyte C8-D30 (Astrocyte type III clone) cell lines were purchased from American Type Culture Collection (Manassas, VA). The cell lines were grown in Dulbecco's Modified Eagle medium (DMEM, Sigma, St. Louis, MO) with high glucose (25 mM) supplemented with 10% fetal bovine serum (FBS, Sigma) and 50 μg/ml penicillin- streptomycin (Sigma). The cells were cultured in a CO₂ incubator with a humidified atmosphere containing 95% air and 5% CO₂ at 37°C.

Morphology

All cell lines were grown under identical culture conditions as described above. Pictures of the cells were taken at approximately 50% confluency with a Cyber-shot digital still camera (Sony) attached to a phase-contrast microscope (Nikon). C T./ U S «3 B S NHS :L JL «3

Phagocytosis assay

Cell phagocytic capacity was determined by using a modification of a standard fluorometric assay (Oda, T. & Maeda, H. (1986) J Immunol Methods 88: 175-183). Fluoresbrite® YG carboxylate microspheres (1 μm diameter, Polysciences, Warrington, PA) were opsinized with FBS and were suspended in PBS. For the assay, cells were seeded in 8- well Lab-Tek Chambered Coverglass Systems (Nunc, Rochester, NY) and allowed to adhere in DMEM. The cells were then incubated with the beads for 1 hr. The cells were rinsed with DMEM to remove excess beads. Phagocytosis was detected by confocal microscopy (Leica TCS SP2, Wetzlar, Germany) and the cells were photographed using Lecia software.

Lipid Isolation, Purification, and Quantification

Total lipids were isolated and purified from lyophilized cell pellets by modified procedures previously described (Macala, LJ. et al. (1983) J Lipid Res 24: 1243-1250; Seyfried, T.N. et al (1984) Exp Neurol 84: 590-595; Kasperzyk, J.L. et al. (2005) J Lipid Res 46: 744-751; Hauser, E.C. et al. (2004) Biochem Genet 42: 241-257). Briefly, neutral lipids and acidic lipids were separated using DEAE-Sephadex (A-25; Pharmacia Biotech, Upsala, Sweden) column chromatography. The neutral lipid fraction contained cholesterol esters, cholesterol, ceramide, cerebrosides, phosphatidylethanolamine, phosphatidylcholine, sphingomyelin, and neutral glycosphingolipids. The acidic lipid fraction contained the gangliosides, free fatty acids, cardiolipin, phosphatidylserine, phosphatidylinositol, phosphatide acid, and sulfatides. The gangliosides were separated from the acidic lipid fraction by Folch partitioning (Folch, J. et al. (1957) J Biol Chem 226: 497-509) and were treated with mild base (0.15 M NaOH) and desalted. Ganglioside sialic acid was quantified by the gas-liquid chromatographic method of Yu and Leeden (Yu, R.K. & Ledeen, R. W. (1970) J Lipid Res 11: 506-516).

Ganglioside biosynthesis in cultured tumor cells

Synthesized gangliosides were isolated and purified from control and tumor cell lines as previously described (El-Abbadi, M. et al. (2001) Br J Cancer 85: 285-292). Briefly, the cells were grown for 48 h in medium containing ¹⁴C-galactose (Sigma, St. Louis, MO). The cells were removed from the flask with a cell scraper in PBS and pelleted. Total lipids were extracted from the radiolabeled cells in chloroform: methanol (1: 1 by vol). The gangliosides were separated from the total lipids by Folch partitioning (Folch, J. et al. (1957) J Biol Chem

Aimer purified using DEAE-Sephadex column chromatography, followed by mild base treatment and desalting as described above. Prior to ganglioside isolation, unlabelled mouse brain ganglioside sialic acid was as a carrier (El-Abbadi, M. et al. (2001) Br J Cancer 85: 285-292). The concentration of labeled sialic acid was determined by scintillation counting.

High performance thin-layer chromatography

All lipids were analyzed by high-performance thin-layer chromatograph (HPTLC) as previously described (Macala, LJ. et al. (1983) J Lipid Res 24: 1243-1250; Kasperzyk, JX. et al. (2005) J Lipid Res 46: 744-751; Hauser, E.C. et al. (2004) Biochem Genet 42: 241- 257). Lipids were spotted on 10 x 20 cm Silica gel 60 HPTLC plates (E. Merck, Darmstadi, Germany) using a Camag Linomat V auto-TLC spotter (Camag Scientific Inc., Wilmington, NC). An internal standard (oleoyl alcohol) was added to the neutral lipid and acidic lipid samples to control for the amount of lipids spotted (Macala, LJ. et al. (1983) J Lipid Res 24: 1243-1250; Kasperzyk, J.L. et al. (2005) J Lipid Res 46: 744-751). Purified lipid standards were purchased from Matreya Inc. (Pleasant Gap, PA). The amount of neutral lipids and acidic lipids spotted on the HPTLC for each cell line was equivalent to approximately 70 μg and 400 μg of cell dry weight, respectively (Fig. 4A and Fig. 4B). The individual neutral and acidic lipid bands were visualized by charring with the cupric acetate, phosphoric acid solution. Approximately 3000 dpm of radiolabeled gangliosides were spotted on the HPTLC for each cell line and the individual gangliosides were visualized as an autoradiogram (Fig. 4C). The ganglioside standards on this plate were from VM mouse brain, B, and purified GM3 and were visualized with the resorcinol spray. The individual gangliosides were labeled according to the nomenclature system of Svennerholm (Svennerholm, L. (1963) J Neurochem 10: 613-623). For GAl analysis, approximately 2 mg of cell dry weight were spotted for each cell line and visualized following immunostaining with anti-asialoGMl (GAl) antibody (Fig. 4D). Neutral lipid, acidic lipid, and ganglioside plates were developed and visualized as previously described (Macala, LJ. et al. (1983) J Lipid Res 24: 1243-1250; Seyfried, T.N. et al (1984) Exp Neurol 84: 590-595; Denny, CA. et al. (2006) J Neurosci Res 83: 1028-1038; Seyfried, T.N. et al. (1994) J Lipid Res 35: 993-1001).

Neutral glycosphingolipid purification

Neutral glycosphingolipids were purified from the neutral lipid fraction following DEAE-Sephadex column chromatography as mentioned above. Briefly, a 4 ml aliquot of the neutral lipid fraction was evaporated under a stream of nitrogen and treated with mild base as described above. The solution was then converted to chloroform: methanol: water (8: 4: 3) by the addition of 4 ml of chloroform: methanol (2: 1, by vol). Folch partitioning was then used to separate the neutral glycosphingolipids from the saponified phospholipids. The upper aqueous phase was discarded and the lower organic phase was washed once with 1.72 ml of chloroform: methanol: water (3: 48: 47, by vol). The combined lower phases were evaporated under a stream of nitrogen and resuspended in 4 ml chloroform: methanol (2: 1, by vol) prior to immunostaining.

Asialo-GMl immunostaining mimunostaining for asialo-GMl (GAl) was performed as previously described (Ecsedy, J.A. et al. (1998) J Lipid Res 39: 2218-2227; Saito, M. et al. (1985) Anal Biochem 148: 54-58). An aliquot of neutral glycosphingolipids representing approximately 2 mg of the original sample dry weight was used for immunostaining. GAl antiserum was obtained from Dr. Robert Yu, Medical College of Virginia, Richmond, VA and diluted 1:40. Perioxidase-conjugated anti-rabbit IgG secondary antibody (1 :150, Sigma) was used to visualize GAl.

Semiquantitative RT-PCR

AU cell lines were grown under identical culture conditions as described above. Total RNA was isolated from homogem^'zed cell pellets of the control and tumor cell lines using the Trizol Reagent (Invitrogen, La Jolla, CA), according to the manufacture's protocol. Single- strand cDNA was synthesized from total RNA and used for PCR amplification as was previously described (Abate, L.E. et al. (2006) J Neurochem 98: 1973-1984). Primers were optimized for annealing temperatures and cycle number as previously described (Abate, L.E. et al. (2006) J Neurochem 98: 1973-1984). Primer sequences and amplicon information for CD68 and F4/80 can be viewed at NCBI (National Center for Biotechnology Information, Pubmed) using accession numbers DQ167574 and DQ167573 and CDl Ib, Ibal, CD45, CXCR4, nestin, SATII, GFAP and NF200 can be viewed using accession numbers EF101553-EF101560. PCR products were separated on 1-1.5% agarose gels containing ethidium bromide and visualized by UV light. RT-PCR was performed on the total RNA of each sample in the absence of reverse transcriptase to control for possible DNA contamination. Metastasis is the process by which cancer cells disseminate from the primary neoplasm and invade surrounding tissue and distant organs. This process involves cancer cell detachment from the primary tumor, intravasation into the circulation, evasion of immune attack, extravasation at distant capillary beds, and invasion and proliferation in distant organs (Welch, D.R. (2006) Defining a Cancer Metastasis. AACR Education Book: 11-17; Chambers, A.F. et al. (2002) Nat Rev Cancer 2: 563-572; Fidler, IJ. (2003) Nat Rev Cancer 3: 453-458). In addition, the metastatic cells establish a microenviroήment facilitating colonization (angiogenesis and further proliferation), resulting in macroscopic malignant secondary tumors (Welch, D.R. (2006) Defining a Cancer Metastasis. AACR Education Book: 11-17; Chambers, A.F. et al. (2002) Nat Rev Cancer 2: 563-572; Fidler, LJ. (2003) Nat Rev Cancer 3: 453-458; Steeg, P.S. (2006) Nat Med 12: 895-904). Metastatic cells preferentially invade those organs (lymph nodes, lung, liver, brain, bone, pleura and peritoneum) that promote tumor cell growth and survival consistent with the 'seed and soil' hypothesis (Fidler, IJ. (2003) Nat Rev Cancer 3: 453-458; Paget, S. (1889) Lancet 1: 571- 573; Munzarova, M. & Kovarik, J. (1987) Lancet 1: 952-954). The metastatic and invasive potential of cancer cells is often correlated with abnormalities in phospholipids and in cell- surface glycolipids, which contribute to tumor progression (Iorio, E. et al. (2005) Cancer Res 65: 9369-9376; Yates, AJ. et al. (1979) J Lipid Res 20: 428-436; Ito, A. et al. (2001) FEBS Lett 498: 116-120). Hence, aberrant cellular migration and proliferation characterize the metastatic phenomenon.

Tumor cell metastasis is the primary cause of morbidity and mortality for cancer patients (Chambers, A.F. et al. (2002) Nat Rev Cancer 2: 563-572; Steeg, P.S. (2006) Nat Med 12: 895-904). While many primary tumors can be treated with conventional therapies, few treatments are effective against metastatic disease (Chambers, A.F. et al. (2002) Nat Rev Cancer 2: 563-572; Steeg, P.S. (2006) Nat Med 12: 895-904; Fidler, IJ. & Hart, LR. (1982) Science 217: 998-1003). The lack of effective therapies for metastasis has been due in large part to the absence of in vivo metastatic models that represent the pathophysiology and biological progression of metastatic disease (Steeg, P.S. (2006) Nat Med 12: 895-904; Paris, S. & Sesboue, R. (2004) Carcinogenesis 25: 2285-2292). For example, experimental metastasis models require injection of tumor cells directly into the host's circulation, thus bypassing the early stages of metastasis prior to local invasion and intravasation (Steeg, P.S. (2006) Nat Med 12: 895-904; Paris, S. & Sesboue, R. (2004) Carcinogenesis 25: 2285-2292; Khanna, C. & Hunter, K. (2005) Carcinogenesis 26: 513-523). While spontaneous models """ express'iήbfe ^f steps δ'jf 'metastasis than experimental models, many of these models require implantation of tumor cells as xenography into immune compromised animals. The use of immune compromised animals is a drawback because tumor cell evasion of the host's immune system is a key rate-limiting process in metastatic spread (Chambers, A.F. et al. (2002) Nat Rev Cancer 2: 563-572; Steeg, P.S. (2006) Nat Med 12: 895-904; Khanna, C. & Hunter, K. (2005) Carcinogenesis 26: 513-523). Spontaneous metastasis models also do not reliably produce secondary lesions, while genetically engineered models are expensive and their metastatic spread is often sporadic and of long latency (several months) (Chambers, A.F. et al. (2002) Nat Rev Cancer 2: 563-572; Steeg, P.S. (2006) Nat Med 12: 895-904; Khanna, C. & Hunter, K. (2005) Carcinogenesis 26: 513-523; Hoffman, R.M. (1999) Invest New Drugs 17: 343-359). Finally, current metastatic mouse models do not naturally produce brain metastasis, a frequent occurrence in many human metastatic cancers (Khanna, C. & Hunter, K. (2005) Carcinogenesis 26: 513-523; Yang, M. et al. (1999) Clin Cancer Res 5: 3549-3559; Schackert, G. & Fidler, IJ. (1988) Lit J Cancer 41: 589-594). These shortcomings and limitations of current in vivo models hinder progress in developing new therapies for metastatic cancer.

This study included analysis of the morphological, behavioral, biochemical, and genetic properties of three new spontaneous malignant brain tumors in the inbred VM mouse strain. Spontaneous tumors of the central nervous system (CNS) are rare in mice and occur with an incidence of approximately 0.01% (Swenberg, J.A. (1982) Neoplasms of the nervous system. The Mouse in Biomedical Research. New York: Academic Press, pp. 529-537). The VM mouse strain is unique in this regard as spontaneous CNS tumors arise in this strain at a relatively high incidence (about 1.5%) (Fraser, H. (1986) Food Chem Toxicol 24: 105-111). Although many spontaneous tumors arising in the VM mouse brain were described previously as astrocytomas (Fraser, H. (1986) Food Chem Toxicol 24: 105-111; Fraser, H. (1971) J Pathol 103: 266-270), one described VM brain tumor was not of neural origin and had metastatic properties when grown outside the brain (El-Abbadi, M. et al. (2001) Br J Cancer 85: 285-292). We prepared cell lines from the three spontaneous VM tumors that were identified as VM-M2, VM-M3, and VM-NMl.

The VM-M2 and the VM-M3 tumors were highly invasive when grown in the brain and metastasized to multiple organ systems when grown subcutaneously in the flank. The VM-NMl tumor was also malignant, but was not invasive in brain and did not metastasize when grown subcutaneously. We show that the VM-M2 and the VM-M3 tumors have multiple properties of macrophages and that the VM-NMl tumor has properties of neural έώpfδgenitόr cBlliVindrcating that spontaneous brain tumors in VM mice can arise from different cell types. Our study describes a novel metastatic tumor model system in mice that expresses all of the major biological processes found in human metastatic tumors. Moreover, this work presents the strongest evidence to date indicating that metastatic cancer cells can arise from macrophages or macrophage-like cells.

Results

The VM-M2, the VM-M3, and the VM-NMl tumors arose spontaneously in the cerebrum of three adult VM mice, two males and a female. The tumors were detected during routine examination of the VM mouse colony over a period of several years (1993-2000). Each tumor-bearing mouse expressed cranial swelling and appeared lethargic with the males also expressing priapism. The tumors were grossly identified in the cerebrum as poorly defined masses (about 3 x 1 x 1 mm) similar to those described previously for other spontaneous tumors in the VM mouse brain (Eraser, H. (1986) Food Chem Toxicol 24: 105- 111; El-Abbadi, M. et al. (2001) Br J Cancer 85: 285-292). In order to preserve in vivo viability, each tumor was immediately resected and implanted intracerebrally (i.e.) into host VM mice as described in Materials and Methods. As soon as cranial domes appeared, the tumors were passaged again into several host VM mice. After a total of three i.e. passages, the tumors were grown subcutaneously (s.c.) and cell lines were prepared from each tumor as described in Materials and Methods. All cell lines were grown under identical conditions to reduce environmental variability.

In vivo growth and invasive behavior

The VM-M2 and the VM-M3 tumors were diffusely invasive into the neural parenchyma with the VM-M3 also displaying streams of invading tumor cells. The VM-M2 and the VM-M3 tumors achieved comparable size after approximately 4 weeks and 3 weeks of i.e. growth, respectively. Both the VM-M2 and the VM-M3 tumors showed multiple tumor foci in the hippocampal region well in advance of the main tumor mass and formed perivascular pseudorossettes. At higher magnification, the cells of both the VM-M2 and the VM-M3 tumors appeared disorganized and pleomorphic.

In contrast to the VM-M2 and the VM-M3 tumors, the VM-NMl tumor was noninvasive and formed a sharp boundary between the main tumor mass and the neural parenchyma. No tumor foci in advance of the main tumor mass or perivascular pseudorossettes were detected in brains containing the VM-NMl tumor. At higher

of 'Me VM-NMl tumor appeared homogenous in shape with poorly defined cytoplasm. The VM-NMl tumor was highly malignant and caused morbidity after approximately 1 week of i.e. growth. These findings indicate that the VM-NMl tumor differed markedly from the VM-M2 and the VM-M3 tumors with respect to invasive behavior, morphology, and growth rate.

When grown s.c, the VM-M2 and the VM-M3 tumors metastasized to multiple organ systems (lung, liver, spleen, kidney, and brain) with 100% fidelity (Table 2). The tumors were implanted subcutaneously in the flank as described in the methods section. Cells from the VM-M2 and VM-M3 tumors metastasized to form numerous secondary tumor nodules in the liver and the spleen. Spleen size was noticeably larger in mice bearing the VM-M2 and VM-M3 tumors than in mice bearing the VM-NMl tumor. No metastatic cells or secondary tumor nodules were found in the liver or the spleen of mice bearing the VM-NMl tumor. The morphology and size of the liver and the spleen from VM-NMl -tumor bearing mice appeared the same as that of normal (non-tumor bearing) mice.

Table 2: Metastatic Sites for VM Tumors ' ^'

Sites Examined

Tumor Lung Liver Spleen Kidney Brain

VM-M2 + + + + + VM-M3 + + + + +

VM-NMl - - - - - ^a Latency for metastasis was 3-4 weeks for VM-M2 and was 2-3 weeks for VM-M3. ^bThe + and - indicate the presence or absence of metastatic spread detectable by either visualization of nodules or by the development of tumors as described in the methods, respectively.

The latency for metastatic tumors in various organs was 3-4 weeks for the VM-M2 tumor and was 2-3 weeks for the VM-M3 tumor, consistent with the differences in i.e. growth rate

contrast to the VM-M2 and the VM-M3 tumors, the VM-NMl tumor did not metastasize following s.c. implantation. However, the growth rate of the VM- NMl tumor was significantly faster than that of the VM-M2 and the VM-M3 tumors and caused morbidity within 14 days of s.c. implantation (Table 2). These findings indicate that the VM-M2 and the VM-M3 tumor cells differ from the VM-NMl cells with respect to growth rate and metastatic potential.

In vitro morphology, adhesion, and phagocytic behavior

The cultured VM tumor cells were compared to the astrocyte C8-D30 cell line (AC) and the macrophage RAW 264.7 cell line. The morphological appearance of the VM-M2 and the VM-M3 cells were similar to each other and to that of the RAW 264.7 cells. Each of these cell lines expressed mixed morphology consisting of large flat cells with protoplasmic extensions and small round haloed cells. The mixed morphology of these cells was not due to the presence of multiple cell types in these cultures, but rather to single cells that changed their morphology during the cell cycle. For example, prior to dividing, a large flat cell would transform gradually into a small round haloed cell, divide, and then transform again into a large flat cell (not shown). In contrast, the VM-NMl and the AC cells expressed a spindle shaped morphology that remained relatively constant throughout the cell cycle. The VM-M2, the VM-M3 and the RAW 264.7 cells were strongly adhesive to the tissue culture flask and were resistant to trypsin treatment. Scraping was necessary to remove these cells from the flask. In contrast, the VM-NMl cells and the AC cells were susceptible to trypsin treatment and were easily removed from the culture flask without scraping.

Phagocytosis of fluorescent beads was noticeably greater in the VM-M2, the VM-M3 and the RAW 264.7 cells than in the VM-NMl and AC cells. In contrast to the VM-M2, the VM-M3 and the RAW 264.7 cells, in which most fluorescent beads were internalized, most of the fluorescent beads associated with the VM-NMl and the AC cells remained on the cell surface. These findings indicate that the VM-M2 and the VM-M3 cells were more similar to the RAW 264.7 cells than to the AC cells with respect to morphology, adhesion, and phagocytic behavior. On the other hand, these properties in the VM-NMl cells were more similar to those of the AC cells.

Lipid Distribution

Lipids were evaluated because these molecules can provide specific information on cell origin, function, and behavior (Seyfried, T.N. et al. (1983) J Neurochem 41: 491-505; ScJe(Iy! fϊ ?et S^'tlAsJ'jtipid Res 39: 2218-2227; Bai, H. & Seyfried, T.N. (1997) J Lipid Res 38: 160-172; Iwabuchi, K. et al. (2000) J Biol Chem 275: 15174-15181; Kotani, M. et al. (1993) Glycobiology 3: 137-146). In general, the distribution of neutral and acidic lipids was more similar in the VM-M2, the VM-M3, and the RAW 264.7 cells than in the VM-NMl and AC cells (Figure 4A, 4B and Table 3). This was especially noteworthy for sphingomyelin and phosphatidic acid levels, which were markedly higher in the VM-M2, the VM-M3, and the RAW 264.7 cells than in the VM-NMl and the AC cells. Indeed, the levels of phosphatidic acid were about 6 to 7 fold higher in the VM-M2 and the VM-M3 cells than in the VM-NMl cells. The elevated phosphatidic acid levels in the VM-M2 and the VM-M3 cells were also associated with reduced levels of phosphatidylcholine relative to the VM- NMl cells, the lipid precursor of phosphatidic acid. Phosphatidylinositol levels were highest in the VM-NMl and the AC cells, whereas levels of cholesterol, cardiolipin, phosphatidylethanolamine and phosphatidylserine were generally similar among all the cell lines. Several lipids were either undetectable or found in only trace amounts and included, cholesterol esters, triglycerides, ceramide, lysophosphatidylcholine, cerebrosides, and sulfatides.

Table 3: Lipid distributions in metastatic and non-metastatic cell lines ^a

Cell Line

a Values represent percentages of individual lipids and are expressed as means of 2-3 independent samples from each cell line. b The percentage distributions for neutral lipids, acidic lipids, and gangliosides were generated from the densitometric scanning of HPTLC plates similar to those shown in Figure 4A, 4B, and 4C, respectively. c ND= not detectable.

The distribution of gangliosides was markedly similar in the VM-M2, the VM-M3, and the RAW 264.7 cells, in which GMl and GDIa were the major species (Figure 4C and Table 3). These gangliosides migrated as double bands due to structural heterogeneity of molecular species (El-Abbadi, M. et al. (2001) Br J Cancer 85: 285-292). Ganglioside GM3 was undetectable in these cells. In contrast, GM3 was the major ganglioside expressed in the VM-NMl cells, and was the only ganglioside detectable in AC cells. The VM-NMl cells also expressed GDIa, GM2, GMl, and GD3. GD3 expression is interesting because this ganglioside is a marker for stem cells or cells of neural progenitor origin (Seyfried, T.N. & Yu, R.K. (1985) MoI Cell Biochem 68: 3-10; Yanagisawa, M. et al. (2005) J Neurochem 95: 1311-1320). GD3 was not expressed in the VM-M2, the VM-M3, or the RAW 264.7 cells. In addition to the markedly similar distribution of gangliosides in the VM-M2, the VM-M3 and the RAW 264.7 cells, these three lines also expressed significant amounts of the neutral glycosphingolipid GAl (asialo-GMl) (Figure 4D). GAl is enriched in macrophages and is a reliable marker for tumor-associated macrophages (Ecsedy, J.A. et al. (1998) J Lipid Res 39: 2218-2227; Taki, T. et al. (1981) J Biochem (Tokyo) 90: 1653-1660). GAl was undetectable in the VM-NMl and AC cells. None of the cell lines expressed gangliosides (GTIb and GQIb) characteristic of neurons. Viewed together, these findings indicate that the distribution of total lipids in the two metastatic VM tumor cell lines was similar to that of the RAW 264.7 cells and was different from that of the non-metastatic VM-NMl cells and from the AC cells. !3/^'4»5:it;:tJ9

Gene expression profiles

Based on the differences in behavior and lipid composition experiments were performed next to examine the expression of genes characteristic of macrophages, glia, neurons, and neural stem/progenitor cells in the various cell lines. The genes characteristic of macrophages included CDlIb , Ibal, F4/80, CD68, CD45 and CXCR4 (Guillemin, GJ. & Brew, BJ. (2004) J Leukoc Biol 75: 388-397; Springer, T. et al. (1979) Eur J Immunol 9: 301-306; Kanazawa, H. et al. (2002) J Biol Chem 277: 20026-20032; Austyn, J.M. & Gordon, S. (1981) Eur J Immunol 11 : 805-815; Micklem, K. et al. (1989) Br J Haematol 73: 6-11; Rossi, D. & Zlotnik, A. (2000) Annu Rev Immunol 18: 217-242). These macrophage genes are also expressed in microglia, the resident macrophages of brain (Guillemin, GJ. & Brew, BJ. (2004) J Leukoc Biol 75: 388-397; Albright, A. V. et al. (1999) J Virol 73: 205-, 213). The genes characteristic of neurons and glia included glial fibrillary acidic protein- (GFAP) andNF200, respectively (Trojanowski, J.Q. et al. (1986) J Neurosci 6: 650-660; Eng, L.F.et al. (1971) Brain Res 28: 351-354). The genes characteristic of neural stem/progenitor cells included nestin and sialyltransferase II (SAT II) (Seyfried, T.N. & Yu, R.K. (1985) MoI Cell Biochem 68: 3-10; Lendahl, U. et al. (1990) Cell 60: 585-595). In general, the expression of macrophage-type genes was restricted to the VM-M2, the VM-M3 and the RAW 264.7 cells (Figure 5A), whereas expression of the neural stem/progenitor-type genes was restricted to the VM-NMl cells (Figure 5B). The AC cells also expressed the nestin gene, but not the SATII gene. None of the VM tumor cell lines expressed genes for mature astrocytes (GFAP) or neurons (NF200) (Figure 5B). Also, none of the VM tumor cell lines expressed CD 19, a gene characteristic of B lymphocytes (not shown) (Engel, P. et al. (1995) Immunity 3: 39-50). Viewed together, these findings indicate that gene expression in the two metastatic VM tumor cell lines was similar to that in the RAW 264.7 cells and differed markedly from that in the non-metastatic VM-NMl and the AC cells.

Discussion

The findings show that the autochthonously arising VM-M2 and VM-M3 tumors expressed all the major biological processes of metastasis to include local invasion, intravasation, immune system survival, extravasation, and secondary tumor formation. The selection of these tumors in their orthotopic site in vivo, rather than selection as cultured cells in vitro, may have contributed to the development and discovery of their highly metastatic phenotype (Khanna, C. & Hunter, K. (2005) Carcinogenesis 26: 513-523). Following s.c.

the VM-M3 tumors metastasized to all major organ systems (liver, kidney, spleen, and lung) and are the only known experimental mouse tumors that reliably metastasize to brain (Khanna, C. & Hunter, K. (2005) Carcinogenesis 26: 513-523; Weil, RJ. et al. (2005) Am J Pathol 167: 913-920). An advantage of the metastatic VM tumors over other metastatic mouse models is that every VM mouse bearing a subcutaneous VM-M2 or VM-M3 tumor developed widespread systemic metastasis within a relatively short period of time (~ 4 weeks). Another advantage is that the syngeneic host strain used for metastatic tumor growth (inbred VM mice) has a functional immune system. Hence, the metastatic VM tumors provide a quick, reliable, and cost effective system for modeling all the major biological steps of cancer metastasis in a natural host environment.

Although the metastatic VM-M2 and VM-M3 tumors arose in the brains of VM mice, they did not express genes or lipid markers characteristic of mature neurons (NF200, and complex gangliosides GTIb and GQIb), astrocytes (GFAP and ganglioside GM3) or oligodendrocytes (cerebrosides and sulfatides) (Seyfried, T.N. et al. (1983) J Neurochem 41: 491-505; Trojanowski, J.Q. et al. (1986) JNeurosci 6: 650-660; Eng, L.F.et al. (1971) Brain Res 28: 351-354; Asou, H et al. (1989) Cell Struct Funct 14: 561-568; Raff, M.C. et al. (1978) Nature 274: 813-816). The absence of CD19 gene expression argues against an origin from lymphocytes. The VM-M2 and VM-M3 tumors also did not express the neural stem/progenitor cell markers for nestin, ganglioside GD3, or SAT II, making it unlikely that the VM-M2 and VM-M3 tumors arose from neural stem cells or neural progenitor cells (Seyfried, T.N. & Yu, R.K. (1985) MoI Cell Biochem 68: 3-10; Lendahl, U. et al. (1990) Cell 60: 585-595). However, nestin, ganglioside GD3 and SATII were expressed in the VM-NMl tumor suggesting that this non-metastatic tumor arose from a neural stem/progenitor cell. These findings indicate that spontaneous brain tumors in VM mice can arise from different cell types and that the highly invasive and metastatic VM-M2 and VM-M3 tumors are not of neural origin.

On the other hand, the VM-M2 and the VM-M3 tumor cells expressed several morphological, behavioral, and genetic characteristics of macrophages or macrophage-like cells. For example, the metastatic VM tumor cells were strongly adhesive to the culture dish and were highly phagocytic. These are classical properties of macrophages (Burke, B. & Lewis, CE. (2002) The Macrophage. 2 ed. New York: Oxford University Press Inc.). Additionally, the in vitro morphology and growth characteristics of VM-M2 and VM-M3 cell lines were remarkably similar to those of the RAW 264.7 macrophage cell line. The VM- M2, VM-M3 and RAW 264.7 cells also expressed multiple genes (CDl Ib, Ibal, F4/80, M' "are expressed in macrophages and microglia, the resident brain macrophage (Guillemin, GJ. & Brew, BJ. (2004) J Leukoc Biol 75: 388-397; Springer, T. et al. (1979) Eur J Immunol 9: 301-306; Kanazawa, H. et al. (2002) J Biol Chem 277: 20026- 20032; Austyn, J.M. & Gordon, S. (1981) Eur J Immunol 11: 805-815; Micklem, K. et al. (1989) Br J Haematol 73: 6-11; Rossi, D. & Zlotnik, A. (2000) Annu Rev Immunol 18: 217- 242). It is unlikely that the expression of these macrophage characteristics in the VM-M2 and VM-M3 cell lines resulted from the in vitro culture conditions since none of these characteristics were expressed in the VM-NMl or AC cells that were cultured under identical conditions. Rather, the expression of these macrophage characteristics in the VM-M2 and VM-M3 cell lines suggests that these cells arose from macrophages or macrophage-like cells.

Further evidence for the macrophage origin of the VM-M2 and VM-M3 tumors comes from the analysis of lipid composition. The VM-M2, VM-M3, and RAW 264.7 cells had remarkably similar distributions of phospholipids and glycosphingolipids suggesting that these cells share a common origin. Indeed, the ganglioside pattern of the VM-M2 and VM- M3 cells was nearly identical to that of the RAW 264.7 cells. The elevated levels of sphingomyelin and gangliosides GMl and GDIa together with undetectable GM3 expression in the VM-M2 and VM-M3 cells is consistent with pro-angiogenic activities and phagocytosis involving lipid rafts and caveolin-1 (Iwabuchi, K. et al. (2000) J Biol Chem 275: 15174-15181; Abate, L.E. et al. (2006) J Neurochem 98: 1973-1984; Lugini, L. et al. (2006) Cancer Res 66: 3629-3638; Lugini, L. et al. (2003) Lab Invest 83: 1555-1567; Nichols, B. (2003) J Cell Sci 116: 4707-4714; Williams, T.M. & Lisanti, M.P. (2005) Am J Physiol Cell Physiol 288: C494-506). High levels of GM3, as seen in the VM-NMl and AC cells, are associated with cell-cell adhesion and reduced angiogenesis (B ai, H. & Seyfried, T.N. (1997) J Lipid Res 38: 160-172; Abate, L.E. et al. (2006) J Neurochem 98: 1973-1984). The high phosphatidic acid levels in the VM-M2 and VM-M3 cells are intriguing since phosphatidic acid is known to participate in key macrophage functions to include invasion, phagocytosis, inflammation, and the respiratory burst (Lugini, L. et al. (2006) Cancer Res 66: 3629-3638; Lugini, L. et al. (2003) Lab Invest 83: 1555-1567; Corrotte, M. et al. (2006) Traffic 7: 365-377; Lee, H.S. et al. (2004) J Cell Biochem 92: 481-490; Sliva, D. et al. (2000) Biochem Biophys Res Commun 268: 471-479; Lim, H.K. et al. (2003) J Biol Chem 278: 45117-45127; McPhail, L.C. et al. (1995) Proc Natl Acad Sci U S A 92: 7931-7935). These lipid biochemical data add further to the data on behavior, morphology, and gene expression indicating that the VM-M2 and VM-M3 cells express characteristics of macrophages or

VM-M3 cells will serve as an important new model cell system for evaluating the role of lipids in metastatic disease.

Macrophages are among the most versatile cells of the body in terms of their ability to move, to change shape, and to secrete growth factors and cytokines (Burke, B. & Lewis, CE. (2002) The Macrophage. 2 ed. New York: Oxford University Press Inc.). In their response to tissue injury or disease, normal macrophages express several hallmarks of metastatic tumor cells. For example, activated macrophages extravasate from the circulation into inflamed or diseased tissue where they proliferate and establish a microenvironment to facilitate angiogenesis and wound healing. This process involves release of inflammatory cytokines and phagocytosis of debris. Following these activities, macrophages intravasate back into the circulation where they take up residence in the lymph nodes and participate in the immune response (Sunderkotter, C. et al. (1994) J Leukoc Biol 55: 410-422; Seyfried, T.N. (2001) Perspect Biol Med 44: 263-282; Bellingan, GJ. et al. (1996) J Immunol 157: 2577-2585). Previous studies in several human cancers showed that metastatic tumor cells express molecular and behavioral characteristics of macrophages (Chakraborty, A.K. et al. (2001) Gene 275: 103-106; Rachkovsky, M. & Pawelek, J. (1999) Cell Growth Differ 10: 517-524; Munzarova, M. et al. (1992) Melanoma Res 2: 127-129; Kerbel, R.S. et al. (1983) MoI Cell Biol 3: 523-538; De Baetselier, P. et al. (1984) Cancer Metastasis Review 3: 5-24; Ruff, M.R. & Pert, CB. (1984) Science 225: 1034-1036; Calvo, F. et al. (1987) Br J Cancer 56: 15-19). Ruff and Pert found that antigens recognized only on macrophages (CD36, C3bi, CDl Ib) were also expressed on tumor cells from human small cell lung carcinoma (Ruff, M.R. & Pert, CB. (1984) Science 225: 1034-1036). Other investigators reported that macrophage specific antigens (CDl Ib, CD14, CD15 and CD45) were found on metastatic melanoma and breast carcinoma (Munzarova, M. et al. (1992) Melanoma Res 2: 127-129; Calvo, F. et al. (1987) Br J Cancer 56: 15-19). Tumor cells from metastatic melanoma, Meth A sarcoma, breast carcinoma, and medulloblastoma also display macrophage-like activity with respect to invasive behavior, chemotactic mobility, and phagocytosis (Lugini, L. et al. (2006) Cancer Res 66: 3629-3638; Lugini, L. et al. (2003) Lab Invest 83: 1555-1567; Montcourrier, P. et al. (1994) J Cell Sci 107 ( Pt 9): 2381-2391; Busund, L.T. et al. (2003) Int J Cancer 106: 153- 159; Ghoneum, M. & Gollapudi, S. (2004) Cancer Detect Prev 28: 17-26; Youness, E. et al. (1980) Arch Pathol Lab Med 104: 651-653). Considered together, these findings suggest that some metastatic human cancers might arise from macrophages or macrophage-like cells.

Although macrophages or macrophage-like cells have long been proposed as the origin of several human metastatic cancers, this hypothesis has not been widely accepted or

Nat Rev Cancer 3: 453-458; Fidler, IJ. & Hart, LR. (1982) Science 217: 998-1003; Pawelek, J.M. (2000) Melanoma Res 10: 507-514; Pawelek, J.M. (2005) Lancet Oncol 6: 988-993; Hanahan, D. & Weinberg, R.A. (2000) Cell 100: 57-70; Fidler, IJ. & Kripke, MX. (1977) Science 197: 893-895; Nowell, P.C. (1976) Science 194: 23-28). Rather, macrophages are generally considered part of the tumor stroma that either inhibit or facilitate tumor growth (Seyfried, T.N. (2001) Perspect Biol Med 44: 263-282; Lewis, C. & Murdoch, C. (2005) Am J Pathol 167: 627-635; Luo, Y. et al. (2006) J Clin Invest 116: 2132-2141). Difficulty in accepting the macrophage hypothesis of metastatic cancer has been due in part to its reliance on a fusion hybrid mechanism whereby metastatic cancer cells arise following hybridization between macrophages or between macrophages and other cells (Munzarova, M. & Kovarik, J. (1987) Lancet 1: 952-954; Munzarova, M. et al. (1992) Melanoma Res 2: 127-129; Pawelek, J.M. (2000) Melanoma Res 10: 507-514; Pawelek, J.M. (2005) Lancet Oncol 6: 988-993; Vignery, A. (2005) Trends Cell Biol 15: 188- 193; Duelli, D. & Lazebnik, Y. (2003) Cancer Cell 3: 445-448). This proposed mechanism of cancer metastasis stands apart from more widely recognized mechanisms involving somatic mutations and epithelial-mesenchymal transitions (Fidler, IJ. (2003) Nat Rev Cancer 3: 453-458; Fidler, IJ. & Hart, LR. (1982) Science 217: 998-1003; Hanahan, D. & Weinberg, R.A. (2000) Cell 100: 57-70; Fidler, IJ. & Kripke, MX. (1977) Science 197: 893-895; Nowell, P.C. (1976) Science 194: 23-28; Thiery, LP. (2002) Nat Rev Cancer 2: 442-454; Yang, J. et al. (2004) Cell 117: 927-939). Although potential mechanism by which the different spontaneous VM tumors arose in the VM mice has not been addressed, these results with the VM metastatic tumors provide the strongest and most comprehensive evidence to date supporting the hypothesis that metastatic cancer cells can arise from macrophages or macrophage-like cells.

A macrophage origin of invasive and metastatic tumors could explain numerous previous findings of macrophage properties expressed by tumor cells. The possibility that some metastatic cancers represent a macrophage disease would have important implications for cancer diagnosis and therapy. The VM tumor model system will have considerable utility for better defining the major biological processes of cancer metastasis and for evaluating potential therapies that can target these processes for tumor management.

Example 7 - Labeling of cell lines (expression of a detectable label) •

Stable expression of a detectable protein (e.g., a bioluminescent or fluorescent protein) was induced in a cell line of the invention using standard procedures. Expression of the ptόlemWas ϊEήieWbf transfecting a cell (e.g. a VM-M2, VM-M3) cell with a vector (plasmid or viral) containing nucleic acid encoding the labeling protein of interest. Many suitable vectors are commercially available and may be used for labeling cell lines in accordance with the manufacturer's instructions. After a clonal line of transgenic cells that expressed the detectable label was produced, transgenic cells from the line were administered into a non-human animal (e.g. a VM mouse) and cells that were administered and their progeny were then visualized in vivo using a microscope with a florescent filter. An in vivo imaging system is also used for real-time visualization of the administered cells and/or their progeny. Use of transgenic cells that express a detectable label allowed tracking and quantitation of tumor growth and metastatic spread.

Other Embodiments

A number of embodiments of the invention have been described. Nevertheless, it will be understood that various modifications may be made without departing from the spirit and scope of the invention. Accordingly, other embodiments are within the scope of the following claims.

All references, including patent documents, disclosed herein are incorporated by reference in their entirety.

Claims

C T./ 1.1 S O B / NHS .11 '3 What is claimed is:

1. An isolated cell of a mouse brain tumor that comprises less than about 1 % GM3 as a percentage of total ganglioside content, wherein following subcutaneous administration of a plurality of said cells to a mouse, the cells spontaneously metastasize to one or more distant organ sites including one or more of the lymph node, skeletal muscle, heart, spinal cord, brain, liver, spleen, kidney, or lung.

2. The cell of claim 1, wherein the cell further comprises less than 5% GM2 as a proportion of total ganglioside content of the cell.

3. The cell of claim 1 or 2, wherein the cell expresses one or more of CD68, Ibal (allograft inflammatory factor 1 (AIFl)), macrophage receptor 1 (Macl), F4/80, CDlIb, CD45, or CXCR4.

4. The cell of any of claims 1-3, wherein the ganglioside content is analyzed by high performance thin-layer chromatography.

5. The cell of any of claims 1-4, wherein the cell is of the cell line VM-M2 or VM-M3.

6. A cell of the cell line VM-M2, VM-M3, or VM-NMl .

7. A cell of a cell line that originated with VM-M2, VM-M3, or VM-NMl .

8. A composition comprising the cell of any of claims 1-7.

9. The composition of claim 8, further comprising a tissue culture medium.

10. The composition of claim 8 or 9, further comprising a cryoprotective agent.

11. A nonhuman subject comprising a cell of any of claims 1 -7.

12. The nonhuman subject of claim 11, wherein the subject is a rodent.

13. The nonhuman subject of claim 12, wherein the rodent is a mouse.

14. The nonhuman subject of claim 13, wherein the mouse is a VM mouse.

15. The nonhuman subj ect of any of claims 11-14, wherein the subj ect is immunocompromised.

16. The nonhuman subject of any of claims 11-15, wherein the subject comprises a cell of the cell line VM-M2 or VM-M3 and a cell of the cell line VM-NMl .

17. A kit comprising a first container comprising a cell of the VM-M2 cell line or the VM-M3 cell line, and instructions for use.

18. The kit of claim 17, further comprising a second container comprising a cell of the VM-NMl cell line, and instructions for use.

19. The kit of claim 17, further comprising a second container comprising a cell of the VM-M2 cell line or the MV-M3 cell line that is not the cell line contained in the first container of the kit.

20. A method of identifying a candidate chemotherapeutic agent, the method comprising:

(a) providing a nonhuman subject;

(b) administering to the subject a plurality of cells of any of claims 1-7 and a test compound; and

(c) measuring the size and/or number of tumors in the subject, wherein a test compound that reduces the size and/or number of tumors in the subject compared to a reference standard is a candidate chemotherapeutic agent.

21. The method of claim 20, wherein the reference standard is:

(i) a second nonhuman subject that has been subjected to the same method as the nonhuman subject of claim 20 with the exception of the administration of the test compound, which is omitted or administered in an inactive form; or

one would expect in an untreated subject.

22. The method of claim 20 or 21 , wherein the nonhuman subj ect is a rodent.

23. The method of claim 22, wherein the rodent is a mouse.

24. The method of claim 23, wherein the mouse is a VM mouse.

25. The method of any of claims 20-24, wherein administering the plurality of cells comprises administering the cells subcutaneously or intravenously.

26. A method of identifying an antimetastatic agent, the method comprising: providing a first nonhuman subject; administering to the first subject a plurality of cells of a VM-M2 or VM-M3 cell line and a test compound; providing a second nonhuman subject; administering to the second subject a plurality of cells of a VM-NMl cell line and the test compound; and measuring the size and/or number of tumors in the first subject and the second subject, wherein a test compound that reduces the size and/or number of tumors in the first subject relative to the second subject is a candidate antimetastatic agent.

27. A method of inducing a tumor in a mouse, the method comprising: providing a mouse subject; and administering to the mouse subject one or more of the cells of any of claims 1-7.

28. An isolated cell of a mouse brain tumor that expresses one or more of CD68, Ibal (allograft inflammatory factor 1 (AIFl)), macrophage receptor 1 (Macl), F4/80, CDlIb, CD45, or CXCR4, wherein following subcutaneous administration of a plurality of said cells to a mouse, the cells spontaneously metastasize to one or more distant organ sites including one or more of the lymph node, skeletal muscle, heart, spinal cord, brain, liver, spleen, kidney, or lung.

cell derived from a cell of any of claims 1-7.

30. The genetically modified cell of claim 29, wherein the cell is genetically modified to express a detectable label.

31. A nonhuman subject comprising a cell of claim 29 or 30.