WO2012040587A2 - Compositions and methods for altering matrix rigidity to regulate cancer cell growth and phenotype - Google Patents

Abstract

A method of determining tumor cell phenotype is carried out by culturing a tumor cell or progeny thereof on a first substrate and on a second substrate and measuring growth, morphology, migration, apoptosis, or protein expression, or gene expression. The substrates are characterized by different elastic moduli.

Description

COMPOSITIONS AND METHODS FOR ALTERING MATRIX

RIGIDITY TO REGULATE CANCER CELL GROWTH AND PHENOTYPE

RELATED APPLICATIONS

This application claims the benefit of priority under 35 U.S.C. § 119(e) to U.S.

Provisional Application No: 61/385,617, filed September 23, 2010, which is incorporated herein by reference in its entirety.

STATEMENT AS TO FEDERALLY SPONSORED RESEARCH

This invention was made with Government support under NIH CA40042 awarded by the National Institutes of Health. The Government has certain rights in the invention.

FIELD OF THE INVENTION

This invention relates to cell culture, as well as reagents, tools, compositions and methods to study, diagnose and/or treat diseases, such as cancer.

BACKGROUND OF THE INVENTION

Cell proliferation is regulated by complex interactions with the surrounding microenvironment, including exposure to growth factors, contact with adjacent cells, and adhesion to components of the extracellular matrix (ECM).

SUMMARY OF THE INVENTION

The invention provides methods for characterizing cells based on their biological function on soft versus stiff substrates. The methods are useful to predict behavior of cells in vivo, e.g. , sensitivity to therapeutic agents, by observing their behavior in vitro on substrate of defined rigidity. For example, cell responsiveness to therapeutic agents is tested on substrates that have the same or similar elastic modulus as those conditions encountered in the body. The ability to make such predictions is critical in drug screening assays, because certain cell types (e.g. , tumor cells) behave differently on soft substrates (pliant or spongy tissue such as lung) compared to stiff substrates (firm or inflexible tissue such as scars or hard tumor masses). Accordingly, a method of determining tumor cell phenotype is carried out by culturing a tumor cell or progeny thereof on a first substrate and on a second substrate and measuring growth, morphology, migration, apoptosis, protein expression or gene expression. The first and second substrates vary in stiffness (as measured by elastic modulus) and a change or difference in any one of above-listed characteristics or activities indicates that the tumor cell comprises a rigidity-dependent phenotype. To determine the phenotype or a rigidity profile of the tumor cell, a difference or change in any one of the above-listed characteristics or activities is detected and the difference between the level associated with culture on the first substrate compared to the level associated with culture on the second substrate (or a plurality of substrates, each of which has a different elastic modulus) is calculated, e.g., the gathered data is transformed using a machine or computer. For example, the method comprises measuring apoptosis or another attribute or activity after growth of the tumor cell in culture for five days on the first substrate and on the second substrate. In another example, protein expression or gene expression comprises protein or gene expression of E-cadherin or Slug. In this manner, a rigidity profile is established for the tumor cell. The methods are useful to determine a rigidity profile for any cell type, e.g. , cell lines, and primary cells obtained from normal or cancerous tissue.

The cell culture substrates vary in their stiffness, e.g. , as measured by elastic modulus (Pascals; Pa). In certain embodiments, the elastic moduli of the substrates are in the range of 100 Pa to 150,000 Pa, e.g. , 100 Pa to 19200 Pa or 600 Pa to 4800 Pa. For example, the elastic moduli of the substrates may be about 1200 Pa. As another example, the first and the second substrates may vary by at least 10 Pa, at least 100 Pa., at least 500 Pa, at least 1000 Pa, or at least 4000 Pa. In one embodiment, the first substrate comprises an elastic modulus of 100 Pa to 1200 Pa, and the second substrate comprises an elastic modulus of 4800 Pa to 19200 Pa. In another example, the first substrate is characterized by an elastic modulus of 100-300 Pa, and the second substrate is characterized by an elastic modulus of 4800 Pa or 9600 Pa.

Biological activity of the cells cultured on each of the different substrates is evaluated. For example, a change in cell number between the number of cells seeded and the number of cells present in a given well after culture for a defined period of time (e.g., 2, 3, 4, 5, 7, and up to 10 days) on the first substrate and second substrate (same temperature, C0₂%, duration) indicates a rigidity dependent phenotype. A change of at least 10%, 20%, 50%, 75% or more reflects the cell's sensitivity to mechanical characteristics of the microenvironment. At least a 10%, 20%, 50% or more, e.g. , a 2-fold change, in cell number after culture on the first substrate compared to the second substrate indicates a rigidity-dependent phenotype. For example, less than a 2-fold change in cell number after on the first substrate compared to the second substrate indicates a rigidity independent phenotype. At least 5, 10, 15, or 20-fold change in cell number after culture on the first substrate compared to the second substrate indicates a strongly rigidity-dependent phenotype.

In one example, at least a 2-fold difference between tumor cell number on the first substrate after culture for 5 days and tumor cell number on the second substrate after culture for 5 days indicates a rigidity dependent phenotype. In another embodiment, less than a 2- fold difference indicates a rigidity independent phenotype. In yet another example, at least a 10-fold difference between tumor cell number on the first substrate after culture for 5 days and tumor cell number on the second substrate after culture for 5 days indicates a rigidity dependent phenotype. The method optionally includes culturing the tumor cell on a plurality of different substrates, e.g., a third, fourth, or fifth substrate (or more), wherein each substrate has a different elastic modulus. For example, the method further comprises culturing the tumor cell on five substrates, wherein the five substrates have an elastic modulus of 150 Pa, 1200 Pa, 2400 Pa, 4800 Pa and 9600 Pa, respectively. In one embodiment, each of the substrates is present on a single multi-well culture plate. In addition to differences in elastic modulus, the substrates may differ in other way such as the composition of the substrate itself. For example, at least one substrate comprises collagen covalently coupled to a polyacrylamide gel. In some embodiments, at least one substrate comprises one or more extracellular matrix (ECM) proteins covalently coupled to a matrix, e.g., collagen, elastin, fibronectin and/or laminin.

The method optionally further comprises culturing the cell or progeny thereof on a third, fourth, or fifth substrate (or more), each of the substrates comprising a different elastic modulus. In one embodiment, the method further comprises culturing the tumor cell (or progeny) on two to five different substrates, wherein the different substrates range in elastic modulus from 100 Pa to 100000 Pa, wherein the difference in elastic modulus between the different substrates is at least 500 or 1000 or 5000 Pa. In certain embodiments, each of substrates is present on a single multi-well culture plate (e.g. , a 96- well or 384-well plate). To accurately predict behavior of a cell in vivo, the mechanical properties of the in vitro culture environment must accurately mimic the in vivo microenvironment. The culture conditions are particularly important in screening for anti-cancer drugs, because a drug may inhibit cell proliferation of cells cultured on a hard substrate but promote cell proliferation of the same cells cultured on a soft substrate (or vice versa). Reliance on screening assays performed on standard tissue culture plates constructed out of hard plastic limits and protracts the drug discovery process, and at worst, leads to identification of compounds that may be detrimental in clinical use.

The systems and methods described herein provide a solution to such shortcomings of conventional drug screening systems. Thus, a method of identifying anti-proliferative drugs or a method for screening candidate compounds to identify an anti-tumor agent is carried out by providing a rigidity dependent cell (i.e., a cell that has been previously profiled for rigidity) and contacting the cell with a candidate compound on a rigidity-matched substrate. A rigidity-matched substrate is characterized as having an elastic modulus similar to that in which the cell or cell type resides in vivo. To mimic early stages of tumor development, substrate stiffness is matched to that of the tissue site of origin of the tumor. Similarly, to mimic early metastatic colonization by tumor cells, stiffness is matched to the tissue site of metastasis. In contrast, primary tumor and metastatic growth are both associated with tissue remodeling and matrix deposition, potentially altering the stiffness of the microenvironment in which the tumor cells reside. To mimic these environments, the stiffness of the substrate can be matched to the tumor microenvironment extant at various stages of tumor

development.

Each cell or cell type is evaluated as described above to determine a rigidity profile prior to undertaking a screening program. A decrease in level of proliferation on a rigidity- matched substrate in the presence of the candidate compound compared to that in the absence of the candidate compound indicates that the candidate compound comprises anti-tumor activity. Accordingly, in one embodiment, a method of identifying an anti-tumor composition is carried out by contacting a rigidity dependent cell, e.g., identified using the methods described above, with a candidate compound on a rigidity-matched substrate. In each of the assays or methods described herein, an amount or level of proliferation (or another activity or characteristic) is detected and the difference between the level associated with culture on the first substrate compared to the level associated with culture on the second substrate (or a plurality of substrates, each of which has a different elastic modulus) is calculated, e.g., the gathered data is transformed using a machine or computer. A decrease in cell proliferation in the presence of the candidate compound as compared to cell proliferation in the absence of the candidate compound indicates that the candidate compound comprises anti-tumor activity. In another example, a method for identifying an anti-tumor composition includes the following steps: (a) preparing or obtaining a cell line derived from a cancer cell present in an in vivo environment in a tumor in a mammal; (b) identifying the elastic modulus (stiffness) of the in vivo environment of the cancer cell in the mammal; (c) identifying a rigidity-matched substrate for the cell line; (d) culturing the cell line on the rigidity-matched substrate, and measuring 5-day cell growth (or growth after about 2 or more days) of the cell line on the substrate in the presence of a candidate composition; (e) culturing the cell line on the rigidity-matched substrate, and measuring 5-day cell growth (or growth for the same amount of time as in (d)) of the cell line on the substrate in the absence of the candidate composition; (f) comparing the growth rates measured in (d) and (e), wherein a decrease by at least two-fold in 5 -day (or other time length of) growth on the substrate in the presence of the candidate composition, as compared to 5-day (or other time length of) growth on the substrate in the absence of the candidate composition, indicates that the candidate composition comprises an anti tumor composition; and (g) identifying an antitumor composition. In one embodiment, the cells of (d) and (e) above are cultured for at least about 2 days, such as about 2-10 days. In yet another example, the method for identifying an anti-tumor composition comprises the following steps: (a) preparing or obtaining a cell line derived from a cancer cell present in an in vivo environment in a tumor in a mammal; (b) measuring the elastic modulus of the in vivo environment of the cancer cell in the mammal; (c) identifying a composition that reduces 5-day growth of the cancer cell line by at least two fold when grown in culture on a substrate having the elastic modulus identified in (b) in the presence of the composition, as compared to 5 -day growth in culture on the substrate having the elastic modulus identified in (b) in the absence of the composition. In the latter example, the composition identified in step (c) does not reduce, or reduces by less than two fold, 5-day growth of the cancer cell line in culture on a substrate having an elastic modulus greater than 100 kPa in the presence of the composition, as compared to 5-day growth on a substrate having an elastic modulus greater than 100 kPa in the absence of the composition. Another variation of the method is carried out as follows: (a) preparing or obtaining a cell line derived from a cancer cell present in an in vivo environment in a tumor in a mammal; (b) identifying the elastic modulus of the in vivo environment of the cancer cell in the mammal; (c) identifying a rigidity-matched substrate for the cell line; (d) culturing the cell line on the rigidity-matched substrate, and measuring 5-day cell growth of the cell line on the substrate in the presence of a candidate composition; (e) culturing the cell line on the rigidity-matched substrate, and measuring 5-day cell growth of the cell line on the substrate in the absence of the candidate composition; (f) comparing the 5 day growth rates measured in (d) and (e), wherein a decrease by at least two-fold in 5 -day growth on the substrate in the presence of the candidate composition, as compared to 5-day growth on the substrate in the absence of the candidate composition, indicates that the candidate composition comprises an anti tumor composition; (g) repeating (d) - (f) with one or more additional different candidate compositions; and (h) identifying an anti-tumor composition.

Also within the invention is a method for determining whether a composition is a "false positive" anti-tumor composition comprising: (a) preparing or obtaining a cell line derived from a cancer cell present in an in vivo environment in a tumor in a mammal; (b) identifying a composition that reduces 5-day growth rate of the cell line by at least two fold when grown in culture on a substrate having an elastic modulus greater than 100 kPa in the presence of the composition, as compared to 5 -day growth rate of the cell line on the substrate having the elastic modulus greater than 100 kPa in the absence of the composition; (c) identifying the elastic modulus of the in vivo environment of the cancer cell in the mammal or in a tumor of the mammal; and (d) determining whether the composition identified in (b) is a "false positive" anti-tumor composition by measuring whether the composition identified in (b) fails to reduce, or reduces by less than two fold, 5-day growth rate of the cell line in culture on a substrate having the elastic modulus identified in (c) in the presence of the composition, as compared to 5 -day growth rate of the cell line on a substrate having the elastic modulus identified in (c) in the absence of the composition.

The methods are also applicable to customize treatment of an individual, i.e., personalized medicine. Thus, a method for treating a cancer in a mammal comprises the following steps: (a) preparing or obtaining a cell line derived from a cancer cell present in an in vivo environment in a tumor in a first mammal; (b) identifying the elastic modulus of the in vivo environment of the cancer cell in the mammal; (c) identifying a composition that reduces growth of the cancer cell line by at least two-fold when grown in culture over 5 days in the presence of the composition on a substrate having the elastic modulus identified in (b), as compared to growth over 5 days in the absence of the composition on a substrate having the elastic modulus identified in (b); (d) identifying a second mammal that comprises cancer cells having the same phenotype as the cancer cell present in the tumor of the first mammal; and (e) reducing or inhibiting growth of the cancer cells in the second mammal by administering to the second mammal the composition identified in (c).

The invention also encompasses multiwall plates. For example, a 96-well plate composition comprises five sections of plates, each section of which comprises two columns of plates comprising collagen covalently coupled to polyacrylamide gels, and wherein the plates in each section have an elastic modulus of, for example, 150 Pa, 1200 Pa, 2400 Pa, 4800 Pa and 9600 Pa, respectively. In one embodiment, the invention further comprises plate sections having two to five different elastic modulus, wherein the different elastic modulus range from 100 Pa to 100000 Pa, and for example, wherein the difference in elastic modulus on the different sections of plates is at least 500 or 1000 or 5000 Pa. Optionally, the 96-well plate composition further comprises at least one cancer cell line having a rigidity dependent phenotype.

In one embodiment, the methods utilize a mulitwell plate that includes well with a plurality of different substrates, at least 2 of which are characterized by a different elastic modulus as described above. For example, the plate used is a 96 well plate (e.g., SoftPlate96 assay). Certain cancer cell types do not grow well on soft substrates, although they grow well on stiff or rigid substrates. Such "rigidity-dependent" cancer cells include rigidity-dependent cancer cell lines. The present invention also discloses that rigidly-dependent cancer cells exhibit a phenotype similar to dormant cancer cells. Soft substrates in the SoftPlate96 assay mimic soft cell tissue in vivo located outside of a "rigid" tumor. When certain cancer cells (such as cancer stem cells) metastasize from a tumor to a different "soft" location in the body, they become dormant. Dormant cancer cells are particularly difficult to kill using

chemotherapy in vivo.

A method for identifying whether a cancer population or sample comprises mammalian dormant cancer cells is carried out as follows: (a) preparing or obtaining a cancer cell line derived from a cancer cell present in a mammal; (b) culturing cancer cells from the cell line on a first soft substrate; (c) culturing cancer cells from the cell line on a second more rigid substrate; and (d) measuring in the cancer cells of (b) and (c) cultured for the same amount of time at least one selected from the group consisting of (i) cell cycle length, length of at least one cell cycle phase, or both; (ii) cellular ATP levels; and (iii) protein synthesis; and (e) identifying whether the cancer cell line comprises mammalian dormant cancer cells based on results obtained upon measuring in (d), wherein at least one of an increase in (i), a decrease in (ii) and a decrease in (iii) in the cancer cells of (b), as compared to in the cancer cells of (c), indicates that the cell line comprises mammalian dormant cancer cells. In certain embodiments, the cancer cell population is obtained from an individual, e.g., a mammalian subject such as a human patient, or is a cell line, e.g., a a cloned, immortalized cell line. Changes in cell behavior are generally detectable after at least 24 hours of culture. For example, differences in (i) and (iii) are detectable after ~2 days or later. For (ii), differences are detected after as short as 24 hours. At least two days of culturing permits detection of differences in (i) and (iii). For example, differences in (iii) are seen after culturing for about four days or longer.

In certain embodiments, the first soft substrate has an elastic modulus in the range of about 100 Pa to about 1500 Pa, and the second more rigid substrate has an elastic modulus of about 4500 Pa to about 150000 Pa. For example, the first soft substrate has an elastic modulus in the range of about 100 Pa to about 300 Pa, and the second more rigid substrate has an elastic modulus of about 4500 Pa to about 76800 Pa.

In the dormant cancer cell assay described above, the method optionally includes the following variations: step (d) further comprises measuring in the cancer cells of (b) and (c) cellular protein levels of at least one protein involved in regulating cytoskeleton structure, wherein a decrease in the protein levels in the cancer cells of (b), as compared to in the cancer cells of (c), indicates that the cell line comprises mammalian dormant cancer cells or step (d) further comprises measuring in the cancer cells of (b) and (c) cellular protein levels of at least one of tubulin, actin, phosphofructokinase-1 (PFK) and Slug, wherein a decrease in the protein levels in the cancer cells of (b), as compared to in the cancer cells of (c), indicates that the cell line comprises mammalian dormant cancer cells. Exemplary proteins whose synthesis is most affected by rigidity (levels are significantly decreased on soft substrates) and exemplary proteins whose synthesis is least affected by rigidity (levels are relatively sustained on soft substrates) are provided in the Examples below.

With regard to assays to determine differential protein levels, the method optionally includes the following step (d), which further comprises measuring in the cancer cells of (b) and (c) the rate of protein synthesis of at least one of E-cadherin, epoxide hydrolase, nicotinamide phosphoribosyltransferase, aldose reducase, kynureninase and aldo-keto reductase family 1 , wherein a lack of difference or increase in the rates of protein synthesis in the cancer cells of (b), as compared to in the cancer cells of (c), indicates that the cell line comprises mammalian dormant cancer cells.

Dormant cancer cell assays optionally comprise step (d) which comprises measuring cellular ATP levels in the cancer cells of (b) and (c), wherein a decrease in cellular ATP levels by 50% or more in the cancer cells of (b), as compared to in the cancer cells of (c), when the cancer cells of (b) and (c) are cultured for about 24 hours or longer, indicates that the cell line comprises mammalian dormant cancer cells. The mammalian dormant cancer cells comprise, have properties of, or have the phenotype of, mammalian stem cancer cells.

Also within the invention is a method of using a cancer cell line to obtain information regarding a mammalian dormant cancer cell. This method comprises culturing, and obtaining information regarding, a cancer cell line that comprises mammalian dormant cancer cells, wherein the cancer cell line is identified by the method described above. The information obtained comprises information relating to cellular metabolic state, signal transduction pathways, and gene expression profiles.

A method for identifying a cancer cell line that exhibits a morphology or phenotype similar to that of a mammalian dormant cancer cell, comprises: (a) preparing or obtaining a cancer cell line derived from a cancer cell present in a mammal; (b) culturing cancer cells from the cell line on a first soft substrate; (c) culturing cancer cells from the cell line on a second more rigid substrate; and (d) measuring in the cancer cells of (b) and (c) cultured for the same amount of time at least one selected from the group consisting of (i) cell cycle length, length of at least one cell cycle phase, or both; (ii) cellular ATP levels; and (iii) protein synthesis; and (e) identifying whether the cancer cell line exhibits a morphology or phenotype similar to that of a mammalian dormant cancer cell based on results obtained upon measuring in (d), wherein at least one of an increase in (i), a decrease in (ii) and a decrease in (iii) in the cancer cells of (b), as compared to in the cancer cells of (c), indicates that the cell line exhibits a morphology or phenotype similar to that of a mammalian dormant cancer cell.

A method for identifying whether a candidate composition comprises a composition that can kill dormant cancer cells comprises: (a) identifying a cancer cell line that comprises, or exhibits a morphology or phenotype similar to, a mammalian dormant cancer cell, comprising identifying a cancer cell line that exhibits at least one of (i) an increase in cell cycle length, (ii) a decrease in cellular ATP levels, and (iii) a decrease in protein synthesis when the cancer cell line is grown on a soft substrate, as compared when grown for the same amount of time on a more rigid substrate; (b) culturing the cancer cell line identified in (a) on the soft substrate, and measuring cell death when growing the cell line in the presence of a candidate composition; (c) culturing the cancer cell line identified in (a) on the soft substrate, and measuring cell death when growing the cell line in the absence of the candidate composition; (d) comparing the cell deaths as measured in (b) and (c), wherein an increase by at least two-fold in cell death when growing the cell line for 2-5 days in the presence of the candidate composition, as compared to when growing the cell line for 2-5 days in the absence of the candidate composition, indicates that the candidate composition comprises a composition that can kill dormant cancer cells; and (e) identifying whether the candidate composition comprises a composition that can kill dormant cancer cells. The method optionally further comprises identifying whether the candidate composition comprises a composition that can kill stem cancer cells.

Advantages of the methods described herein include reduced time and cost for screening. In addition to increased efficiency of the screening process, testing in this manner ultimately results in increased safety. Screening with this method increases the biofidelity of the culture environment, and permits identifying candidate compounds whose function is specific to certain substrate stiffness conditions that are relevant to in vivo environments. Hence, the method detects functional compounds that exert negligible effects in traditional rigid tissue culture environments but are functional in certain in vivo environments.

Screening using the methods described herein reduces false positives, in which compounds are highly effective in traditional rigid tissue culture environments, but are ineffective in soft environments more analogous to in vivo environments.

"Tumor cell phenotype" refers to observable characteristics or traits in a cancer cell, such as morphology, phenotype, migration, growth (e.g., rate of growth, cell cycle length), cell signal transduction, gene expression, protein expression, cell metabolism (e.g., cellular ATP levels), cell death/apoptosis, and other biochemical or physiological properties.

"Culture" refers to a process by which cells are grown under controlled conditions, such as in a tissue culture dish or plate in an incubator. A "candidate compound" is a compound that is being tested for, or has or potentially has, certain effects on one or more cells, such as tumor/cancer cells in vivo or in vitro, when contacted with the cell, cells or tissue comprising cells, or when administered to a subject.

A "rigidity dependent cell" is one that grows differently depending on the elasticity of the substrate on which it is grown or cultured. One example is a cancer cell that grows well on a rigid substrate (e.g., having an elastic modulus of about 4500 Pa to about 150000 Pa or greater), but less well on a soft substrate (e.g., having an elastic modulus in the range of about 100 Pa to about 1500 Pa.

A "rigidity-matched substrate" refers to a substrate having an elastic modulus (e.g., a hard or soft substrate) that allows for greater growth or metabolism of a cell of interest, as compared to a reduced growth or metabolism of the cell on a substrate with a different elastic modulus (e.g., a hard or soft substrate). In one embodiment, a "rigidity-matched substrate" is a substrate having an elastic modulus in a range that allows for the highest growth or metabolism of a cell of interest, as compared to substrates having different elastic modulus.

An "elastic modulus" is the measure of stiffness of a substance. It is the degree of strain that the object undergoes in response to a defined stress applied to it. It is measured in Pascal's (Pa), i.e., force per unit of area.

An "anti-tumor composition" is a composition that kills or inhibits growth of a tumor cell when contacted with the cell, cells or tissue comprising cells, or when administered to a subject.

A "false positive anti-tumor composition" is a composition that kills or inhibits growth of a tumor cell when contacted with the cell when grown in culture on a substrate having an elastic modulus greater than 100 kPa (100000 Pa), but does not kill or inhibit growth of the tumor cell at all or as well when contacted with the cell when grown in culture on a substrate having the same elastic modulus as that in the in vivo environment of the cancer cell (e.g., about 100 Pa to about 9600 Pa).

Other features and advantages of the invention will be apparent from the following description of embodiments thereof, and from the claims. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention, suitable methods and materials are described below. All published foreign patents and patent applications cited herein are incorporated herein by reference. Genbank and NCBI submissions indicated by accession number cited herein are incorporated herein by reference. All other published references, documents, manuscripts and scientific literature cited herein are incorporated herein by reference. In the case of conflict, the present specification, including definitions, will control. In addition, the materials, methods, and examples are illustrative only and not intended to be limiting.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1A is a diagram showing design of a typical 5-day growth assay using the SoftPlate96 yields a "growth profile," which reflects the effect of rigidity on the proliferation of the cell line. Collagen- coated polyacrylamide gels of varying rigidity are attached to the bottom of the wells of a 96-well plate. Cells are cultured on the gels for 5 days, and growth is measured by quantitation of cellular DNA. The resulting "growth profile" is generated for each cell line and summarizes the response of that cell line to extracellular rigidity. Figure IB is a bar graph showing a 5-day growth assay of four cancer cell lines on plastic. Figure 1C is a bar graph showing 5-day growth assays of the four cancer cell lines on the SoftPlate96. Data are expressed as fold change over the number of cells initially plated. Results are representative of at least three experiments. These figures show growth of cancer cell lines on flexible substrates.

Figure 2A is a bar graph showing relative cell adhesion. A549 and MDA-MB-231 cells were plated on the SoftPlate96 and total cell numbers per well were counted after 6 hours of attachment. Data are expressed as percent of adhesion to the 150 Pa gels. Figure 2B is a bar graph showing % cell population and cell cyde stage. A549 and PC-3 cells were cultured on 150 Pa or 4800 Pa gel substrates for 5 days followed by cell cycle analysis. Results show average of at least three experiments + SEM. * p <0.05. Figure 2C is a series of representative micrographs of A549 and MDA-MB-231 cells plated for 5 days on 150 Pa or 4800 Pa gel substrates. All cell nuclei are stained with DAPI (blue) and TUNEL-positive cells are labeled with fluorescein (green). Bar = 100 μιη. Figure 2D is a bar graph showing quantitation of TUNEL staining. Average of 2 experiments + SEM with a total at least 400 cells counted for each condition. These figures represent an analysis of adhesion, cell cycle, and apoptosis of cancer cell lines on soft gels. Figure 3A is a dot plot and a line graph showing the growth of cancer cell lines in mouse lung tissue. GFP-labeled MDA-MB-231, A549, PC-3, or mPanc96 cells were seeded into the lungs of nude mice. (Left) The number of GFP-positive cells in the lung was determined 4 hours and 14 days after injection, and the change in the number of GFP-positive cells in the lung over the 14 days was scored. * p <0.05 vs. MDA-MB-231 cells. (Right) GFP-labeled A549 and mPanc96 cells were seeded into the lungs and the percentage of GFP- positive cells were scored at intervals over 24 hours. Figure 3B is a series of

photomicrographs showing the histology of the mouse lung at 14 days following injection of A549 cells (left panel) and mPanc96 cells (right panel). Arrows indicate micrometastases. Figure 3C is a bar graph showing a comparison of the growth of cell lines on plastic (taken from Figure IB) and on 1200 Pa substrates (taken from Figure 1C).

Figure 4A is a series of micrographs of A549, MDA-MB-231, PC-3, and mPanc96 cells that were plated on 150 or 4800 Pa gel substrates for 20 hours. Figure 4B is a bar graph showing areas of cells that were plated for 20 hours on 150 or 4800 Pa gel substrates. Results show mean fold increase over an unspread cell + SEM of at least 20 cells counted for each condition. Figure 4C is a bar graph showing cell velocity. A549, MDA-MB-231, PC-3, and mPanc96 cells were plated for 2 hours, then filmed for an additional 18 hours. Mean cell velocity + SEM in jim/hr was determined by tracing and measuring the paths of 15 cells per rigidity per cell line. * p <0.05. These figures demonstrate rigidity-dependent changes in morphology and migration correlate with rigidity-dependent cell proliferation.

Figure 5A is a series of photomicrographs showing cells strained for actin, paxillin, or E-cadherin. A549 cells were cultured for 3 days on gels with rigidities of 150, 4800, or 19200 Pa. Cells were fixed and stained for actin (green) and paxillin (red). Arrows indicate focal adhesions. Figure 5B are photomicrographs showing stained cells. A549 cells were cultured on gels with rigidities of 150 or 19200 Pa. Cells were fixed and stained for actin (green) and E-cadherin (red). Figure 5C is a photograph of the results of a western blot assay. A549 cells cultured on PA gels for 3 days were lysed and blotted for expression of E-cadherin, vimentin, and actin. Figure 5D is a bar graph showing the relative levels of Slug and E-cadherin mRNA in A549 cells cultured on PA gels for 3 days as measured by real-time RT-PCR. Results show mean + SEM of three independent experiments. These figures demonstrate that substrate rigidity regulates E-cadherin expression in A549 cells. Figs. 6A-D are bar graphs showing that culturing rigidity-dependent cells on soft substrates does not select for a subpopulation of rigidity-independent cells. A549 cells (A, B) or MDA-MB-231 cells (C, D) were cultured on plastic (A, C) or a 150 Pa substrate (B, D) for 15 days. The cells were then subjected to a 5-day growth assay on a Soft-Plate 96. Each cell line exhibited its typical Soft-Plate profile.

Figure 7A is a photograph of the results of a western blot assay. Cells were cultured on 150 Pa or 4800 Pa gels for 5 days and FAK autophosphorylation levels were detected by immunoblotting for phospho-Y397 (top) and total FAK (bottom). Numbers refer to fold increase in FAK autophosphorylation over the 150 Pa control. A representative blot is shown. Figure 7B is a bar graph showing a quantitation of blots as shown in Figure 7A. Results are a mean + SEM of at least 4 independent experiments. These figures show FAK phosphorylation in cancer cell lines cultured on PA gels.

Figure 8 is a series of photographs of a western blots and a series of bar charts showing cyclin Dl expression and cell cycle analysis in rigidity-dependent cancer cells growing on soft and stiff gels. Figure 8A is a photograph of a western blot, wherein A549 cells and MDA-MB-231 cells were cultured on 150 Pa, 4800 Pa, or 19200 Pa polyacrylamide gels for 2 or 5 days. Cells were lysed and analyzed by western blot for the expression of cyclin Dl (top panel). The expression of GAPDH was analyzed as a loading control (bottom panel). Despite a lack of FAK activity and a decrease in growth on soft gells, A549 and MDA-MB- 231 cells continue to express cyclin D. Figure 8B is a photograph of a western blot showing cyclin Dl expression as compared to actin expression. Figure 8C is a series of bar charts showing cell cycle profiles of A549 and MDA-MB-231 cells cultured on soft (150 Pa) or still (19200 Pa) gels for 2 or 5 days as measured by DNA staining and flow cytometry. Both cell lines show a slight accumulation of cells in the Gl stage of the cell cycle after 5 days. The A549 cells do not show an accumulation of cells in any one stage of the cell cycle after 2 days of culture on soft gels.

Figure 9 is a series of scatter plots and bar charts demonstrating BrdU pulse-chase of cell cycle progression. Figure 9A shows that the cells are "pulsed" for 30 minutes with the nucleotide analog BrdU, resulting in the majority of cells in S phase incorporating the BrdU label (first row). The BrdU+ population is then tracked over time to calculate the rate at which the cells are progressing through the cell cycle (second row). A549 cells were pulsed with BrdU for 30 minutes following growth on soft or stiff gels for 2 days. Figure 9A shows scatter plot histograms of BrdU-labeled cells on soft (third row) or stiff (fourth row) gels, stained for DNA content (X-axis) and BrdU (Y-axis). The times indicated are the times, in hours, after the BrdU pulse. Figure 9B shows cell cycle progression analysis was performed on the scatter plot histograms from the cells grown on gels for 2 days (left) or 5 days (right).

Figure 10 is a series of bar charts showing ATP levels in rigidity-dependent cells cultured on soft or stiff gels. ATP levels were measured in A549 cells (left) or MDA-MB- 231 cells (right) following culture on a Softplate96 for 2 days. Data represents the average of two experiments performed in triplicate + S.E., of cells on soft (300 Pa) or stiff (19200 Pa) gels. Total ATP levels were normalized to cell numbers.

Figure 11 is a schematic and a scatter plot showing differential analysis of protein expression by "stable isotope labeling of amino acids in cell culture" (SILAC). Protein synthesis is decreased in rigidity-dependent cells cultured on soft gels. A549 cells were subjected to SILAC analysis to determine rates of protein synthesis on soft or stiff gels. Figure 11 A shows an overview of the SILAC procedure. A549 cells were cultured for 3-4 passages in media containing stable heavy isotopes of lysine and arginine, resulting in -100% of cellular proteins labeled with the heavy amino acids. These cells were then cultured on soft or stiff gels for 4 days in the presence of the heavy isotopes, then the isotopes were removed and the cells were cultured in regular "light" growth media for 24 hours. Cells were then lysed, and cellular proteins were digested by trypsin, and the resulting peptides were analyzed by mass spectrometry. The ratio of "heavy" to "light" peptides for each peptide sequence corresponds to the rate of synthesis of that protein over the 24-hour period that the cells were in the light media. Figure 11B shows a scatter plot of heavy to light (H/L) ratios of proteins identified by SILAC/mass spectroscopy from A549 cells (left) or mPanc96 cells (right). Each dot represents an individual protein, and the dotted lines represent the means of the H/L ratios of proteins from cells on soft (blue) or stiff (orange) gels.

Figure 12 is a dot plot showing the identification of proteins that are differentially regulated by rigidity. The graph shows the proteins with the greatest (green) or least (red) changes in H/L ratios between cells on soft and stiff gels. Proteins with the greatest changes in H/L ratios (most sensitive to changes in rigidity) include structural proteins and proteins involved in glucose metabolism. Proteins with the least changes in H/L ratios (levels are relatively maintained on soft versus stiff gels) include proteins involved in the metabolism of cellular macromolecules and in NAD-producing salvage pathways. Figure 13 shows a schematic illustrating potential signaling processes regulated by extracellular rigidity.

Figure 14 shows the cell cycle analysis of A549 cells cultured on soft (150 Pa) or stiff (19200 Pa) gels. A549 cells were cultured on gels for 2 days and cell cycle analysis was performed by BrdU incorporation. The cells on the soft gels exhibited longer Gl and S phases than the cells on the stiff gels. The 1:4 ratio of growth of A549 cells on soft versus stiff gels is similar to what is seen experimentally.

Figure 15 shows the validation of the SILAC results presented above. Cells were grown for 5 days on gels, and plated on plastic for the indicated time.

DETAILED DESCRIPTION OF THE INVENTION

The mechanical properties of the extracellular matrix have an important role in cell differentiation. However prior to the invention, cancer cells responsiveness to changes in microenvironmental rigidity was unclear. A 96-well assay system that arrays extracellular matrix-conjugated polyacrylamide gels that vary in stiffness by at least 2 fold across the plate was established. This assay was used to determine how changes in the rigidity of the ECM modulate the biological properties of tumor cells. The cell lines tested fell into one of two categories based on their proliferation on substrates of differing stiffness: "rigidity dependent" (those which show an increase in cell growth as extracellular rigidity is increased), and "rigidity independent" (those which grow equally on both soft and stiff substrates). Cells which grew poorly on soft gels also showed decreased spreading and migration under these conditions. More importantly, the ability of cells to grow on soft substrates in vitro correlated with their ability to grow in a soft tissue environment in vivo. The lung carcinoma line A549 responded to culture on soft gels by expressing the differentiated epithelial marker E-cadherin and decreasing the expression of the mesenchymal transcription factor Slug. These observations indicated that the mechanical properties of the matrix environment play a significant role in regulating the proliferation and the morphological properties of cancer cells. Further, the multiwell format of the Soft-plate 96 assay is a useful and effective tool for growing cancer cells in pre-determined, defined stiffness cultures.

Mechanical properties of extra cellular matrix (ECM).

The control of epithelial cell (EC) differentiation and proliferation is critical for tissue homeostasis. EC proliferation is regulated by complex interactions with the surrounding microenvironment, including exposure to growth factors, contact with adjacent cells, and adhesion to components of the extracellular matrix (ECM). Alteration of the signaling pathways that regulate the response to these microenvironmental cues is a critical event in tumor initiation, progression and metastasis.

The mechanical properties of the ECM have been identified as an important factor regulating the differentiation and proliferation of a multitude of cell types both in vitro and in vivo. Specifically, the rigidity ("stiffness") of the ECM, defined by its elastic modulus (E) in units of force per area (Pa), affects the growth, differentiation, and functionality of many cell types, including stem cells, fibroblasts, glial cells, and cardiomyocytes. In addition, disease states are often accompanied by a local increase in ECM rigidity. Cancer progression in soft tissues is typically associated with an increase in rigidity due to local accumulation of a dense, crosslinked collagen matrix allowing detection of the tumor by physical palpation.

Accordingly, nontumorigenic mammary epithelial cells, which normally reside in the soft (E =150 pascals [Pa] or N/m2) microenvironment of the breast, show increased proliferation when cultured on stiffer matrices (E=4500 Pa), along with increased migration, augmented ERK signaling, and loss of cellular polarity. These attributes are considered hallmarks of tumor cells and are characterized as being an integral component of a transition from a relatively quiescent to a "malignant" phenotype, driven by a local increase in ECM rigidity.

Cancer cells are responsive to variations in microenvironmental rigidity. Fibroblasts transformed with oncogenic H-Ras no longer show inhibition of growth on soft substrates. In addition, the growth properties of clonal populations of the breast cancer cell line MDAMB- 231 differ in response to rigidity, and they correlate with the ability to grow in the soft lung or stiff bone in vivo. This indicates that the growth properties of a particular cancer cell line in response to substrate rigidity may be determined by its genetic or epigenetic composition.

Analysis of human cancer cell lines is generally performed using cells cultured on rigid plastic, or in matrigel or soft agar, the mechanical properties of which are poorly defined and difficult to modulate. A method for culturing cells on biologically relevant "soft" substrates using ECM-conjugated polyacrylamide (PA) gels that can span the stiffness range of 100 Pa - 150,000 Pa was developed. A 96-well assay system arrays PA gels of varying in stiffness in user-defined increments across the plate.

This system was used to determine how changes in the rigidity of the ECM modulate the biological properties of tumor cells, including growth, morphology, and migratory properties. The cell lines tested diverged into two categories based on their proliferation profiles: "rigidity dependent" lines generally exhibited increasing cell growth as extracellular rigidity increased, while "rigidity independent" lines grew equally well across the entire tested spectrum of matrix stiffness. Cells which grew poorly on soft gels also showed decreased spreading and migration under these conditions. The growth of four representative cell lines selected from these two categories was assessed in vivo by introducing the cells into the soft tissue environment of the lung. The two rigidity-independent cell lines (PC-3 and mPanc96) grew well in soft (lung) tissue, while the rigidity dependent cell lines (A549 and MDA-MB- 231) did not grow well in the lung. The lung carcinoma line A549 responded to culture on soft gels by expressing the differentiated epithelial marker E-cadherin and decreasing the expression of the mesenchymal transcription factor Slug. These observations indicated that the mechanical properties of the matrix environment play a significant role in regulating the proliferation and the morphological properties of cancer cells, and that the "rigidity profile" is an intrinsic property of each cancer cell line.

The following materials and methods were used to generate the data described herein.

Example 1 : Matrix Rigidity Regulates Cancer Cell Growth and Cellular Phenotype

Cell culture and antibodies

Cancer cell lines were obtained from the ATCC except: MDA-MB-231(SA) cells were a gift from Amy Bouton and Theresa Guise (UVa), VMM 18 and VMM39 were a gift from Victor Engelhard (UVa), and HPSC cells were a gift from Rosa Hwang (M.D. Anderson). Cells were routinely cultured in RPMI supplemented with 10% fetal bovine serum (FBS), except for the BT549, 22Rvl, and mPanc96 cell lines, which were maintained in DMEM with 10% FBS. The MCF-IOA human mammary cells were maintained (Debnath et al. , Methods, 30: 256-268, 2003). Monoclonal antibodies to E-cadherin and FAK were purchased from Cell Signaling Technologies. Monoclonal anti-actin and vimentin, and polyclonal anti-FAK pY397 were purchased from Sigma.

Polyacrylamide substrates and Soft-plate96 fabrication

Flexible polyacrylamide substrates were generated on glass coverslips or in 96- well arrays and adapted for cell culture using the method of Pelham and Wang (Wang et al, Methods Enzymol., 298: 489- 496, 1998). Analysis of cell growth, migration, cell cycle, and apoptosis

Cell growth and adhesion assays were performed using the Soft-plate96 (USSN: 12/675,882 and 12/675,839). Analysis of cell spreading, migration, cell cycle, and apoptosis were performed on gels on glass coverslips.

In vivo lung colonization assays

Cancer cell lines were fluorescently labeled by infecting with a lentivirus encoding GFP. 1 x 106 cells in 200μ1 PBS were injected into the tail vein of 6-8 week-old nude mice (Taconic). Lungs were removed at 2-24 hours or 14 days following tumor cell injection and digested in collagenase (0.5 mg/ml in growth media) overnight at 37° C. Lung homogenates were fixed for 20 minutes at room temperature with 2% paraformldehyde. Samples were analyzed by flow cytometry on a FACSCalibur Benchtop Analyzer and data acquired with Cell Quest software (Beckton Dickinson). 5x105 events were collected, and GFP positive cancer cells counted with FlowJo v.8.8.6.

Cell staining, immunoblotting, and quantitative PCR

Immunofluorescence staining and immunoblotting were performed using standard methods. Quantitative PCR was performed on cDNA generated from cells cultured for three days on PA gel-coated coverslips.

Rigidity-dependent growth of cancer cell lines

To measure the growth of cancer cell lines as a function of matrix rigidity, 96- well assay system ("Soft-plate96") that uses collagen covalently coupled to polyacrylamide gels as substrates in place of ECM-coated rigid plastic was used. The Soft-plates were comprised of five sections, each containing two columns of collagen-coated PA gels of a specific elastic modulus (Figure 1A), 150 Pa and 1200 Pa (comparable to lung and breast), 2400 Pa and 4800 Pa (comparable to a mammary tumor), and 9600 Pa (approximating striated muscle). These elastic moduli were chosen based on known measurements of the rigidity of soft tissues and tumors (Engler et al, Cell, 126: 677-689, 2006; Flanagan et al, Neuroreport, 13: 2411-2415, 2002; Paszek et al, Cancer Cell, 8: 241-254, 2005; Engler et al, J.Cell Biol., 166: 877-887, 2004), and on data showing that the greatest changes in rigidity-dependent cell occurred between 150 Pa and 4800 Pa.

The growth profile of fourteen cancer cell lines was determined by plating the cells on the Soft-plate96 and measuring the fold change in cell number after five days using a fluorescent DNA-binding dye (Table 1). In addition, the growth profiles of nontumorigenic mammary epithelial cells (MCF-IOA) and two fibroblast lines were determined. Cell growth on defined matrices generated a qualitative "growth profile" for each cell line (Figure 1 A). The growth profiles of the cell lines fell into one of two categories: "rigidity-dependent" cells, at least a 2-fold change in cell number across the range of extracellular rigidity tested (e.g., MDA-MB-231 breast cancer cells and A549 lung cancer cells), and "rigidity-independent" cells which grew equally well across the range of tested matrix stiffness (e.g., PC-3 prostate cancer cells and mPanc96 pancreatic cancer cells) (Table 1). There was no correlation between the shape of the stiffness-dependent growth profile and the tissue of origin, or whether the cells were originally cultured from the primary tumor or from a metastatic lesion.

Table 1

SoftPlate96 growth profiles of cancer cell lines

Table 1 is a compilation of 5-day growth assays for 17 cell lines. Included in the table are original source of the cells (indicated by literature citations), the ability to grow on 150 Pa and 9600 Pa substrates (from SoftPlate96 assays), and the Soft-plate96 growth profile for each cell line. Grey profiles indicate rigidity-dependent lines and black profiles indicate rigidity- independent lines. Growth is measured as follows: - < 1 fold; + 1-5 fold; ++ 5-10 fold; +++ 10-15 fold; ++++ 15-20 fold; +++++ > 20 fold increase in cell number over 5 days.

Further characterization was carried out on two cell lines which showed rigidity- dependent growth (MDA-MB-231 and A549) and two cell lines which showed rigidity- independent growth (mPanc96 and PC-3). Each cell line demonstrated robust cell growth on rigid collagen-coated plastic (Figure IB). The rigidity-dependent cells demonstrated a 4-5 fold increase in number on the more rigid gels (4800-9600 Pa) relative to the soft (100-1200 Pa) gels (Figure 1C, top panels). In contrast, the two rigidity-independent cell lines demonstrated nearly equivalent numbers on the soft and rigid gels (Figure 1C, bottom panels). To determine if the differential growth on soft or rigid substrates represented the selection of a population of cells exhibiting preferential growth on the different substrates, A549 or MDA-MB-231 cells were cultured on plastic or 150 Pa substrates for 15 days. These cells were then harvested and subjected to a 5-day growth assay on a Soft-Plate 96. No change in the Soft-Plate growth profile was observed after prolonged culturing on soft substrates (Figure 6). These data establish cell line specific differences in the ability to grow on soft versus rigid substrates and indicate that the "rigidity profile" is an intrinsic property of each cell line.

Properties of rigidity-dependent and -independent cell lines on different substrates

Studies were carried out to determine whether the decreased growth of rigidity- dependent cells on soft gels was due to defects in adhesion to the substrate, a block in cell cycle, or induction of apoptosis. A549 and MDA-MB-231 cells were plated on the Soft- plate96, allowed to adhere for six hours, and the number of attached cells measured. Both cell lines attached efficiently to the collagen-gels irrespective of elastic moduli (Figure 2A), indicating that lower cell numbers on the gels after five days was not due to a lack of cell attachment.

A lack of adhesion signaling in anchorage-dependent cells results in a block at the Gl/S checkpoint of the cell cycle. A549 cells cultured on a 150 Pa gel for five days showed a modest but significant accumulation in the Gl phase of the cell cycle with a corresponding decrease in the percentage of cells in the S phase (Figure 2B), consistent with a block at the Gl/S checkpoint. In contrast, MDA-MB-231 cells exhibited no significant change in their cell cycle profile, similar to the rigidity-independent cell lines PC-3 and mPanc96 (Figure 2B). However, both of the rigidity-dependent cell lines (A549 and MDAMB-231) exhibited significant apoptosis when cultured on soft gels for five days, while the rigidity-independent PC-3 and mPanc96 cell lines did not (Figure 2C-D). None of the four cell lines exhibited significant apoptosis when cultured on the more rigid (4800 Pa) gels. These data indicate that the "rigidity profile" of cells does not reflect differences in adhesion to the matrix, but more likely reflects rigidity -dependent changes in cell cycle progression and cell apoptosis.

The ability of cancer cell lines to form colonies in soft tissue correlates with their Softplate96 profiles

Studies were undertaken to determine whether the differential ability to grow on soft substrates exhibited by the rigidity-dependent and -independent cell lines was predictive of the ability of these cells to grow in a soft tissue environment in vivo. Two rigidity-dependent lines (MDA-MB-231 and A549) and two rigidity-independent lines (PC-3 and mPanc96) were stably transduced with a GFP-encoding lentivirus, and injected into the tail vein of nude mice. Either 2-24 hours or 14 days post-injection the GFP-positive cell population in the lung homogenates was determined by flow cytometry and histochemistry. Each of the cell lines exhibited significant number of cells in the lungs post injection (Fig 3A). However, after two weeks the lungs of mice injected with the two rigidity-dependent cell lines (MDA-MB-231 and A549) contained fewer GFP-positive cells, compared to the lungs of mice injected with rigidity-independent cell lines (PC-3 and mPanc96) (Figure 3A). Hematoxylin and eosin (H&E) staining of paraffin-embedded sections of the lungs two weeks post-injection of mPanc96 cells showed microcolonies within the alveoli, while the lungs of the mice that were injected with the A549 cells and MDA-MB-231 containied no detectable microcolonies (Figure 3B). Thus the growth of the rigidity-independent lines in the lung correlated with their efficiency of growth on the 1200 Pa gels of the SoftPlate96 assay (Figure 3C), a rigidity similar to that of lung tissue. The correlation between relative cell growth rates on soft substrates, but not rigid dishes, with the growth of the same cell lines in the lung indicates that the cells' ability to grow on soft gels in vitro predicts their ability to grow in soft tissue in vivo. Increased proliferation and cell migration of rigidity-dependent cells correlates with cell spreading

Experiments were carried out to evaluate whether proliferation of rigidity-dependent cell lines correlated with the ability of cells to spread on different gel substrates. Both MDA- MB-231 and A549 cells exhibited significant increases in cell spreading on 4800 Pa gels compared to 150 Pa gels (Figure 4A-B). Similar results were obtained when BxPC-3 cells, a pancreatic line that exhibits a comparable growth profile to MDA-MB-231 and A549 cells (Table 1), were cultured on 150 Pa and 4800 Pa gels. PC-3 cells (rigidity-independent) were able to spread on 150 Pa gels to a similar extent as on the 4800 Pa gels, whereas mPanc96 cells (rigidity-independent) did not spread appreciably on either soft or rigid substrates (Figure 4A- B). The ability to spread on more rigid substrates also correlated with the ability of A549 and MDA-MB-231 cells to migrate. In contrast, both PC-3 and mPanc96 cells failed to show significant differences in migration when plated on soft versus more rigid substrates (Figure 4C). These results demonstrate that for rigidity-dependent cell lines, the ability of cells to spread correlates with increased proliferation and migration. For rigidity-independent cells these behaviors appear to be uncoupled.

Focal adhesion kinase (FAK) is a critical signaling component of integrin signaling and has been implicated in sensing the rigidity of the ECM (Provenzano et at, Oncogene, 28: 4326-4343, 2009; Tilghman et al. Semin.Cancer Biol., 18: 45-52, 2008). FAK activity, as measured by its autophosphorylation on tyrosine397, was only modestly activated as a function of matrix stiffness in A549 cells, and was not significantly altered in the other cell lines tested (Figure 7). These data emphasize that the behaviors of the different cancer cell lines on soft or rigid substrates cannot be simply attributed to alterations in general adhesion signaling through FAK activation. The mechanical properties of the microenvironment regulate the epithelial and mesenchymal properties of A549 cells

The conversion of normal epithelial cells to malignant, metastatic counterparts often involves the loss of expression E-cadherin and the acquisition of a more migratory phenotype - a process termed epithelial-to-mesenchymal transition (EMT). A549 cells grown on soft (150 Pa) gels for 5 days formed clusters with no visible focal adhesions or stress fibers (Figure 5A). In contrast, cells on more rigid (4800 Pa and 19200 Pa) gels were spread and more disperse exhibiting prominent stress fibers and focal adhesions (Figure 5A), all hallmarks of the mesenchymal phenotype. Immunofluorescence staining or western blot analysis of cells cultured on the 150 Pa gels or 4800 or 19200 Pa gels demonstrated significant upregulation of E-cadherin expression on soft substrates (Figure 5B-C). However, no significant change in the expression of the mesenchymal marker vimentin was observed in cells growing on the different substrates (Figure 5C).

The transcription repressor Slug is a member of the Snail family of DNA-binding elements that regulates E-cadherin expression (Hayashida et at, J.Biol. Chem., 281: 32469- 32484, 2006) and has been shown to be critical for conferring a metastatic phenotype in an experimental model of melanoma ( Shibue et at, Proc.Natl.Acad.Sci.U.S.A, 106: 10290- 10295, 2009). Quantitative RT-PCR analysis of A549 cells cultured on substrates of different rigidities revealed an upregulation of Slug mRNA when cells were grown on more rigid gels (4800 and 19200 Pa) compared to the soft gel (150 Pa) (Figure 5D). Slug mRNA levels inversely correlated with E-cadherin mRNA levels, which were lower in A549 cells cultured on the more rigid gels compared to cells cultured on the soft gel, consistent with changes observed in E-cadherin protein expression (Figure 5B-C). These data indicate that matrix rigidity can modulate E-cadherin and Slug expression in A549 cells. MDA-MB-231 cells, while exhibiting rigidity-dependent proliferation, did not express detectable levels of E- cadherin at either 150 Pa or 19200 Pa, suggesting that these cells, while morphologically similar to A549 cells on soft and rigid substrates, do not alter the expression of this epithelial marker when exposed to a soft microenvironment.

Mechanical properties of the ECM regulate cancer cell proliferation and survival

An efficient and flexible assay system was used to determine how changes in matrix rigidity influence cell properties. Analysis of 14 cancer cell lines revealed that altering the rigidity of the collagen-coated matrix prominently alters the growth of certain cancer cell lines ("rigidity-dependent" growth) while having little effect on other cancer cell lines ("rigidity- independent" growth) which grew robustly even on substrates of very low stiffness. The lower growth rates on soft gels in rigidity-dependent cell lines were caused at least in part by the selective alteration in cell cycle progression and the induction of apoptosis when cell lines were plated on soft matrices. Additionally, the rigidity-dependent lines showed a marked decrease in cell spreading and migration when plated on soft versus rigid substrates, and at least one of the cell lines (A549) exhibited a rigidity-dependent regulation of E-cadherin expression and reversible modulation of epithelial and mesenchymal phenotype. When seeded into mouse lungs, rigidity-dependent cell lines did not grow as well as rigidity-independent lines, indicating a correlation between the ability to grow on soft matrices in vitro and proliferative capacity in vivo in the lung.

The Soft-plate96 multiwell assay represents a relatively high-throughput approach to assess the role of substrate rigidity on the properties of cancer cells in culture. In this system, the method of Pelham and Wang (Wang et al, Methods EnzymoL, 298: 489- 496, 1998) has been adapted to generate a multiwell plate in which the substrate is comprised of

polyacrylamide gels of varying stiffness that have been functionalized to provide a binding surface for extracellular matrix proteins, e.g., collagen. In the studies described, above the plates were designed to encompass five levels of elastic moduli, ranging from 150 to 9600 Pa, however other formats are easily created. The endpoint of the assay was cell proliferation, but other endpoints, e.g., cell survival are readily measured. As illustrated, the assay system provides a rapid and reproducible method to assess the role of matrix rigidity on cell growth and survival.

A panel of cancer cell lines was surveyed to determine how changes in the mechanical properties of the matrix influence cell proliferation. Nine of the cancer cell lines exhibited a dependence on matrix rigidity for growth, growing significantly better on stiff/rigid matrices than on the less rigid/soft matrices. Rigidity-independent cell lines exhibited virtually no changes in growth rate over the range of matrix stiffness used on the plates. Remarkably, all of the cancer cell lines examined in this study were capable of proliferating on soft substrates, whereas normal fibroblasts, smooth muscle and epithelial cells exhibit a strict dependence on matrix rigidity for growth. This observation reflects the "oncogenic" transformation of the cancer cell lines relative to normal cells, events that reflect the multiple mutations that characterize cancer cells. The results parallel the analysis of a series of ghoma cell lines propagated on fibronectin-coated polymeric substrates of defined mechanical rigidity. On highly rigid substrates (> 100 kPa), the ghoma cells spread extensively, formed prominent stress fibers and mature focal adhesions, and migrated rapidly. However, when cultured on less rigid matrices (values comparable with normal brain tissue), the ghoma cells appeared rounded and failed to productively migrate, similar to the rigidity-dependent cell hnes described above. Ghoma cell motility on highly compliant substrates was rescued by pharmacologic inhibition of actinomyosin-based contractility, indicating that actinomyosin contractility is a critical component of the mechanosensory apparatus. FAK, ERK, and the small GTPase Rho may also play a role in the regulation of growth in response to rigidity, or an increase in cyclin D levels downstream of Rac activation.

Extracellular rigidity affects the growth of certain cancer cell lines, and the ability of a cell line to grow on a soft substrate in vitro predicts its ability to grow in a soft environment in vivo. In addition, a cell line's response to extracellular rigidity in vitro predicts its reaction to the desmoplastic response in vivo, i.e., whether an increase in rigidity of the microenvironment in vivo will favor growth of the tumor cells. A cell line's Soft-plate growth profile also predicts its sensitivity to therapeutic drugs in soft tissue. For example, the DNA-crosslinker mitomycin C has been shown to inhibit proliferation of mesenchymal stem cells more efficiently on rigid versus soft substrates. In addition, pancreatic cancer cell hnes that express epithelial markers such as E-cadherin and lower levels of the mesenchymal marker vimentin are more responsive to erlotinib treatment. Therefore, if a cell line (such as A549) were to become more epithelial-like when cultured in a soft environment, it would be predicted to be more sensitive to erlotinib. These data demonstrate the predictive capacity of the Soft-plate assay in determining cancer cell responses to in vivo soft tissue environments and therapeutic potency within such environments.

Cellular plasticity, or the ability to transition back and forth between a sessile epithelial cell and a migratory mesenchymal cell, is a phenomenon that is critical to several physiological processes, including embryonic development, wound healing, and cancer progression. A common feature to EMT is a downregulation of cell-cell adhesion, primarily through inhibition of E-cadherin expression, and the acquirement of a motile phenotype along with increased expression of certain infrastructural components such as vimentin. However, this transition may not always be complete, as there are many examples of cells which undergo a "partial" EMT (p-EMT) in which cells become motile by transiently acquiring some but not all of the mesenchymal cell characteristics. Cells undergo EMT in a complex and stepwise manner, and not all EMT events are necessary to achieve a migratory phenotype. E-cadherin expression alone has been linked to inhibition of migration and Gl/S cell cycle arrest. The observations described herein indicate that certain cancer cell lines, while they may have mesenchymal characteristics when they are cultured on rigid substrates, when placed in a soft microenvironment they may respond accordingly by activating an epithelial-type program.

The data described herein demonstrate the response of cancer cell lines to changes in the rigidity of their microenvironment. The substrates and methods are useful to identify for tumors based on the mechanical milieu in which they thrive.

Example 2: Matrix Rigidity Regulates Cancer Cell Growth by Modulating Cellular

Metabolism and Protein Synthesis

Morphology of dormancy is seen in rigidity-dependent cancer cells identified by the methods described herein by virtue of the fact that when grown on soft substrates (e.g., 100- 300 Pa), rigidity-dependent cancer cells exhibit longer cell cycles, less active metabolism, and reduced protein synthesis, as compared to the same cells when grown on a rigid/stiff substrate for the same amount of time (e.g., 1-5 days). When grown on soft substrates, rigidity-dependent cancer cells continue to synthesize, at normal rates, proteins that are necessary to sustain cell growth— i.e., protein machinery needed to keep dormant cells alive even if growth and metabolism is slowed. See Table 5. Thus, rigidity-dependent cancer cells grown on soft substrates, such those identified using the SoftPlate96 assay, provide an excellent tool for the study of dormant cancer cells, such as cancer stem cells.

Described herein is the utilization of "softplate96" technology to identify the properties of rigidity-dependent cancer cell lines that regulate their differential growth on soft and rigid substrates. Compared to cells growing on more rigid/stiff substrates, cells on soft substrates (100-300 Pa) exhibited a longer cell cycle, due predominantly to an extension of the Gl phase of the cell cycle, and were metabolically less active, showing decreased levels of intracellular ATP and a marked reduction in protein synthesis. The rates of protein synthesis of over 1200 cellular proteins under conditions of growth on soft and rigid/stiff substrates were measured using stable isotope labeling of amino acids in culture (SILAC) and mass spectrometry. Cellular proteins whose synthesis was preferentially inhibited by exposure to soft substrates and proteins whose synthesis was relatively resistant to plating on soft matrices were identified. The former category included proteins that regulate cytoskeletal structures (e.g. , tubulins) whereas the latter category included proteins that regulate key metabolic pathways required to sustain cell growth, e.g.. nicotinamide phosphoribosyltransferase, a regulator of the NAD salvage pathway. The cellular properties of rigidity-dependent cancer cells growing on soft matrices are reminiscent of the properties of cancer stem cells or dormant cancer cells, e.g. , slow growth rate and reduced metabolism. The soft plate technology provides a unique platform for the study of stem cells and dormant cancer cells, allowing the growth and molecular analysis of cells in different mechanical environments that reflect the changing microenvironment of cancers in experimental animal models and in patients.

Sensing the mechanical properties of the extracellular matrix (ECM) is a central mechanism for regulating the differentiation and proliferation of a multitude of cell types both in vitro and in vivo. Evidence implicates alterations in the signaling pathways that regulate the response of cells to microenvironmental cues as critical events in tumor initiation, progression, metastasis and perhaps tumor dormancy.

Tumor dormancy is defined as a stage in cancer progression in which residual disease is present but is asymptomatic (Aguirre-Ghiso, J.A. 2007. Models, mechanisms and clinical evidence for cancer dormancy. Nat Rev Cancer. 7:834-46). Twenty to forty-five percent of breast and prostate cancers will relapse years or decades later. Most cancer types are associated with disseminated disease that, after treatment, might persist as minimal residual disease. For example, many dormant cells survive chemotherapy, and disseminated breast cancer cells are growth- arrested and resistant to doxorubicin. It is hypothesized that dormant cells survive because they are not dividing; however, prior to the invention described herein, it was unknown whether the cells are not dividing because they have

upregulated/downregulated expression of certain regulatory proteins, which can be targeted. Prior to the invention described herein, the lack of mechanistic insight into this stage has been a major shortcoming in our understanding of the full complexities of metastatic growth. To fully understand dormancy, cells must be characterized in their dormant state.

The increase in tissue rigidity due to local accumulation of a dense, crosslinked collagen matrix is a hallmark of cancer progression in soft tissues and is the basis for detection of many types of tumors by physical palpation (Huang, S., and D.E. Ingber. 2005. Cell tension, matrix mechanics, and cancer development. Cancer Cell. 8: 175-6; Levental, K.R., H. Yu, L. Kass, J.N. Lakins, M. Egeblad, J.T. Erler, S.F. Fong, K. Csiszar, A. Giaccia, W. Weninger, M. Yamauchi, D.L. Gasser, and V.M. Weaver. 2009. Matrix crosslinking forces tumor progression by enhancing integrin signaling. Cell. 139:891-906).

Analysis of human cancer cell lines in cell culture is almost always performed using cells cultured on rigid plastic, or less often in Matrigel or soft agar, the mechanical properties of the latter being poorly defined and/or difficult to modulate. A method for culturing cells on biologically relevant "soft" substrates using ECM conjugated polyacrylamide (PA) gels that can span the stiffness range of 100 pascals ([Pa] or N/m²) - 150,000 Pa has been described (Tilghman, R.W., C.R. Cowan, J.D. Mih, Y. Koryakina, D. Gioeli, J.K. Slack- Davis, B.R. Blackman, D.J. Tschumperlin, and J.T. Parsons. Matrix rigidity regulates cancer cell growth and cellular phenotype. PLoS One. 5:el2905). In the assay described herein, a 96-well assay system that arrays PA gels of varying stiffness in user defined increments across the plate was utilized (Mih, J.D., A.S. Sharif, F. Liu, A. Marinkovic, M.M. Symer, and D.J. Tschumperlin. A multiwell platform for studying stiffness-dependent cell biology. PLoS One. 6:el9929). This experiment has been used to assess how changes in the rigidity of the ECM modulate the biological properties of tumor cells, including growth, morphology, and migratory properties. The cancer cell lines tested were grouped into two categories based on their proliferation profiles: "rigidity dependent" lines exhibited increasing cell growth as extracellular rigidity increased, while ' 'rigidity independent' ' lines grew equally well across the entire tested spectrum of matrix stiffness. Cells which grew poorly on soft gels exhibited decreased spreading and migration under these conditions and grew poorly when introduced into the soft tissue environment of the lung. The rigidity-dependent lung carcinoma line A549 responded to culture on soft gels by expressing the differentiated epithelial marker E- cadherin, and decreasing the expression of the mesenchymal transcription factor Slug. These observations demonstrate that the mechanical properties of the matrix environment play a significant role in regulating the proliferation and the morphological properties of cancer cells, and that the "rigidity profile" is an intrinsic property of each cancer cell line.

As described in detail below, the properties of rigidity-dependent cancer cell lines that contribute to their differential growth on soft and rigid substrates were investigated using the soft plate technology. Compared to cells growing on more rigid/stiff substrates, cells on soft substrates (100-300 Pa) exhibited a longer cell cycle, due predominantly to an extension of the Gl phase of the cell cycle, and were metabolically less active, showing decreased levels of intracellular ATP and a marked reduction in protein synthesis. The rates of protein synthesis of over 1200 cellular proteins under conditions of growth on soft and rigid/stiff substrates were measured using stable isotope labeling of amino acids in culture (SILAC) and mass spectrometry. Whereas overall rates of protein synthesis decrease markedly in rigidity- dependent cells grown on soft substrates, cellular proteins whose synthesis was preferentially inhibited by exposure to soft substrates and proteins whose synthesis was relatively resistant to plating on soft matrices were identified. The former category included proteins that regulate cytoskeletal structures (e.g. , tubulins) whereas the latter category included proteins that regulate key metabolic pathways required to sustain cell growth, e.g. , nicotinamide phosphoribosyltransferase, a regulator of the NAD salvage pathway. The cellular properties of rigidity-dependent cancer cells growing on soft matrices are reminiscent of the properties of cancer stem cells or dormant cancer cells, e.g. , slow growth rate and reduced metabolism. The soft plate technology provides a unique platform for the study of stem cells and dormant cancer cells, allowing the growth and molecular analysis of cells in different mechanical environments that reflect the changing microenvironment of cancers in experimental animal models and in patients.

Cell lines and antibodies

Cancer cell lines were obtained from the ATCC. All cells were routinely cultured in RPMI supplemented with 10% fetal bovine serum (FBS). Antibodies to cyclin Dl were purchased from Abeam, and the actin antibody was purchased from Sigma.

Polyacrylamide substrates

Flexible polyacrylamide substrates were generated on glass coverslips or in 96- well arrays and adapted for cell culture using the method of Pelham and Wang (Wang et at, Methods Enzymol., 298: 489- 496, 1998). Polyacrylamide gels contained 3% (150 Pa) or 7% acrylamide (4800 and 19200 Pa), and 0.04% (150 Pa), 0.05% (4800 Pa), or 0.24% (19200 Pa) bisacrylamide. The gels were polymerized on acid- washed, silanated, and glutaraldehyde- treated 22 mm glass coverslips. Each gel was placed in a well of a 6-well dish and activated using the heterobifunctional crosslinker Sulfo-SANPAH followed by coating with collagen I (10 μg/ml) for four hours at room temperature or overnight at 4°C. The gels were soaked in the appropriate growth media at 37°C for at least 20 minutes prior to the addition of cells. Fabrication of Soft-plate96 plates

Glass-bottom 96-well plates (Matrical) were treated with a 0.4% aqueous solution of c- methacryloxypropyltrimethoxysilane (Acros Organics) to enable covalent attachment of acrylamide to the glass during gel polymerization. Solutions containing 0.075% ammonium persulfate, 0.15% tetramethylethylenediamine, and variable ratios of acrylamide:

bisacrylamide (all from Bio-rad) were delivered into the well plate with a multichannel pipettor. A 96-pin block with affixed, hydrophobic glass squares corresponding to the diameter of the wells was inserted, sandwiching the polymerization solutions between two glass surfaces. Gel thickness was controlled by placing 100 mm-thick spacers in the corner wells. Following polymerization, the block was removed and the gels were immersed in 0.5 mg/ml of the heterobifunctionalcrosslinker sulfosuccinimidyl-6(49-azido-29- nitrophenylamino) hexanoate diluted in 50 mM HEPES buffer, pH 8.5. After a 5 minute UV exposure, the crosslinker solution was removed and the gels were rinsed once with HEPES buffer. Monomelic collagen (PureCol) diluted in PBS at 100 mg/ml was delivered to each well and incubated for 4 hours at room temperature. The well plate was rinsed in PBS and UV-sterilized prior to cell seeding. The soft-plate 96 assay system was also described by Mih et al. (Mih, J.D., A.S. Sharif, F. Liu, A. Marinkovic, M.M. Symer, and D.J. Tschumperlin. A multiwell platform for studying stiffness-dependent cell biology. PLoS One. 6:el9929). Cell cycle analysis

A549 cells were cultured on 150 Pa or 19200 Pa gels for 3 or 5 days. BrdU was added to the cells for 30 minutes; cells were washed twice with PBS and incubated for 2-10 hours in fresh growth media. Labeled cells were collected by trypsinization, washed twice with PBS, and resuspended in 0.5 ml PBS. The cells were fixed overnight after the addition of 4 ml 70% ethanol. The fixed cells were collected by centrifugation, and DNA was denatured for 30 minutes with 0.3 ml of a solution of 0.2 mg/ml pepsin in 2M hydrochloric acid. Following denaturation, the acid was neutralized with 1 ml of a solution of 3.8% sodium tetraborate. The cells were then washed once with 1 ml of FACS buffer (2% BSA, 0.1% Triton-X 100 in PBS) and incubated with anti-BrdU antibody (diluted 1 :8 in FACS buffer) for 1 hour at room temperature. The cells were washed once with FACS buffer and resuspended in PI staining solution (0.1 % Triton X 100, 0.1% RNAase cocktail [Ambion], lmg/ml PI in PBS). The stained cells were analyzed by flow cytometry on a BD FACSCalibur Benchtop Analyzer. Scatter charts were analyzed with FlowJo v8.8.6. The length of each cell cycle phase was calculated as previously described (Terry, N.H., and R.A. White. 2006. Flow cytometry after bromodeoxyuridine labeling to measure S and G2+M phase durations plus doubling times in vitro and in vivo. Nat Protoc. 1 :859-69).

ATP assay

Cells were cultured on a SoftPlate96 with an elastic modulus range of 300 Pa ("soft") to 76800 Pa ("stiff) for two days. The cells were lysed, and cellular ATP levels were measured using the ATPlite assay (Perkin Elmer), as per the manufacturer's instructions.

SILAC of A549 and MPanc96 cells

A549 cells were grown for two passages in SILAC media (Lysine and Arginine replaced with Lys 13C6 and Arg 13C6, respectively) to incorporate the heavy isotopes into the cellular proteins. The labeled cells were then cultured on soft (150 Pa) or stiff (19200 Pa gels for 4 days In the presence of heavy amino acids. On the fourth day, the cells were washed twice in PBS, and incubated for 24 hours in unlabeled ("light") media. The cells were then collected by trypsinization, counted, and lysed in sample buffer. Lysates were separated by SDS-PAGE. Protein bands were cut from the gel and digested with trypsin. The resulting peptides were analyzed and identified by mass spectrometry and for each individual protein the number of heavy (H) or light (L) peptides was determined. The mean H/L ratios of peptides derived from the A549 cell lysates were normalized by dividing each mean by itself, and proteins were identified as being at least one standard deviation away from the normalized mean.

Cell cycle progression of rigidity-dependent cells on soft gels

To understand the molecular basis for the slow growth of rigidity-dependent cell lines on soft matrices, key regulators of cell cycle progression as well as the time required for cells to transverse the cell cycle on soft or stiff substrates were measured. Cyclin Dl is critical for cells to enter the cell cycle, and the loss of cyclin Dl expression is a marker for cells that have exited the cell cycle and are quiescent. The expression of cyclin Dl in rigidity- dependent cells was measured to determine whether the cells exit the cell cycle when cultured on soft gels. Rigidity-dependent A549 (lung carcinoma)or MDA-MB-231 (breast caracinoma) cells were cultured on 150 Pa, 4800 Pa, or 19200 Pa gels for 2 or 5 days, and cyclin Dl expression was measured by western blot. As shown in Figure 8, both cell lines still expressed cyclin Dl even when cultured on the soft (150 Pa) gels. These data indicate that rigidity-dependent cells do not exit the cell cycle, even on soft gels where the cells exhibit slower growth.

Since rigidity-dependent A549 cells do not exit the cell cycle when plated on soft gels, and only ~5% of the cells undergo apoptosis under these conditions (Tilghman, R.W., and J.T. Parsons. 2008. Focal adhesion kinase as a regulator of cell tension in the progression of cancer. Semin Cancer Biol. 18:45-52), it was determined if these cells are progressing through the cell cycle more slowly than when they are growing on stiff substrates. To examine the rate of progression through specific phases of the cell cycle, the rigidity- dependent lung carcinoma line A549 cells were cultured on soft (150Pa) or stiff (19200 PA) gels for two or five days and then pulsed for 30 minutes with the nucleotide analog bromodeoxyuridine (BrdU) to label the population of cells undergoing DNA synthesis. Time spent in each phase of the cell cycle was determined by tracking the BrdU-positive population as the cells progressed through the S and G2 phases and accumulated in the Gl phase (Tilghman, R.W., and J.T. Parsons. 2008. Focal adhesion kinase as a regulator of cell tension in the progression of cancer. Semin Cancer Biol. 18:45-52). Cells on soft gels progressed more slowly through the Gl phase of the cell cycle compared to the same cells growing on stiff matrices (Figure 9), consistent with a global decrease in cellular metabolism or anabolic processes that affects the cellular "growth" stage of the cell cycle. Because the cell cycle profiles are similar after 2 and 5 days on soft gels, it is likely that this represents a steady-state measurement, as opposed to the possibility that cell growth is gradually slowing down over time. Based on the calculated length of the cell cycle for cells growing on soft versus stiff gels after 5 days the ratio of cells on soft versus stiff matrices would be 1 :4.3. A similar ratio was observed by manually counting the number of cells present at five days following growth on soft versus stiff matrices, thereby validating the BrdU pulse-chase calculations.

The fold increase in cell number over five days is detailed below.

Calculated fold increase

Figure 14 shows the cell cycle analysis of A549 cells cultured on soft (150 Pa) or stiff (19200 Pa) gels. A549 cells were cultured on gels for 2 days and cell cycle analysis was performed by BrdU incorporation as described above. The cells on the soft gels exhibited longer Gl and S phases than the cells on the stiff gels. The 1 :4 ratio of growth of A549 cells on soft versus stiff gels is similar to what is seen experimentally.

Cellular metabolism in rigidity-dependent cells cultured on soft gels

The Gl phase of the cell cycle relies heavily upon cellular metabolic events and is critical for the synthesis of structural proteins and enzymes that contribute to the generation of new organelles and the overall growth of the cell. Because the Gl phase of the cell cycle was prolonged in rigidity-dependent cells grown on soft gels, it was determined if there are metabolic changes under conditions of growth of soft substrates. Culturing A549 or MDA- MB-231 cells on soft gels for 2 days resulted in an approximately 50% decrease in ATP levels compared with cells growing on stiff gels (Figure 10). The lower levels of cellular ATP are consistent with a decrease in cellular metabolism when rigidity-dependent cells are cultured on soft gels.

Effect of rigidity on protein synthesis in cells cultured on soft gels

Because of the decrease in ATP levels in cells cultured on soft gels, it was determined whether matrix rigidity may regulate net protein synthesis in rigidity-dependent cells. To assess the changes in global protein synthesis and to identify proteins that might be differentially synthesized under soft versus stiff conditions, "stable isotope labeling of amino acids in cell culture" (SILAC) was performed. Heavy-to-light (H/L) ratios for individual proteins from both stiff and soft conditions sorted by IPI Protein index number are shown in Figure 11. Rigidity-dependent A549 cells and rigidity- independent mPanc96 cells were grown in SILAC media for two generations on plastic tissue culture dishes to fully label the proteome with "heavy" amino acids, e.g. , lysine and arginine (Figure 11 A). Cells were then plated on either soft or stiff gels and cultured in "heavy" amino acids for an additional 4 days. Cells were then incubated for 24 hours ("pulse") in "light" media (normal tissue culture media), and cellular proteins were isolated from each of the samples and resolved on SDS- PAGE. Proteins present in soft and stiff samples were analyzed by taking ten gel slices from each protein track and subjecting each to digestion with trypsin. Net protein synthesis during the 24 hour "pulse" was measured by determining the ratio of heavy to light peptides by mass spectrometry for proteins in the soft and stiff samples. Mean H/L ratios between stiff and soft were significantly different (p<0.0001). Both sets of data follow a normal distribution with unequal means and variances. Figure 1 IB plots the heavy to light ratio for peptides derived from individual proteins (e.g. , each dot is an identified protein with 2-20 labeled peptides) for A549 and mPanc96 cells. The average heavy/light ratio of peptides in the A549 cells on soft gels was significantly higher (p<0.05) than in the cells on stiff gels, indicating that net protein synthesis was decreased in cells on soft gels, (e.g. , fewer light amino acids were incorporated during the pulse in cell growing on soft substrates compared to stiff substrates). By contrast, the heavy/light ratios of peptides in mPanc96 cells were not significantly different between cells growing on soft or stiff gels (Figure 11B).

To assess the relative rates of synthesis of individual cellular proteins in A549 cells under conditions of growth on soft or stiff gels the ratio of heavy to light peptides was compared for individual proteins and identified those proteins whose H/L ratios were greater than one standard deviation from the mean (Figure 12). Proteins where Soft Ratio »Stiff Ratio or

Soft Ratio « Stiff Ratio have the largest decreases (green) or the least changes (red) in their rates of synthesis when the cells are cultured on soft compared to stiff substrates, respectively. Specifically, proteins whose synthesis was most sensitive to the shift from stiff to soft matrix (largest change in heavy/light ratio (Figure 12, green circles) are associated with cytoskeletal structures or glucose/sugar metabolism, whereas the proteins that exhibit the least change in rate of protein synthesis (smallest change in heavy to light ratio, red circles, Figure 12) are proteins involved in the metabolism of cellular macromolecules required to sustain growth on soft matrices. These results are consistent with the hypothesis that matrix rigidity modulates cellular metabolic processes, including energy production and protein synthesis.

The proteins with the largest decrease in rate of synthesis upon shift from stiff to soft are shown below. Proteins GO terms

Aspartate aminotransferase, mitochondrial Cellular macromolecular complex assembly Glutamate dehydrogenase 1 , mitochondrial Cellular protein complex assembly

Tubulin beta-3 chain Translational elongation

Leucine-rich PPR motif-containing protein, mitochondrial Macromolecular complex assembly

Tubulin alpha-1 C chain Microtubule-based movement 60 kDa heat shock protein, mitochondria! Protein polymerization

Elongation factor 1 -gamma Translation

Tubulin beta-2A chain Protein complex biogenesis 40S ribosomal protein S3a Protein complex assembly

Tubulin, beta Microtubule-based process Elongation factor 2 Mitotic cell cycle

Tubulin beta-4 chain Regulation of apoptosis

40S ribosomal protein S6 Regulation of programmed cell death 40S ribosomal protein S3 Regulation of cell death

40S ribosomal protein S8 Cell cycle process

The proteins with the least change in rate of synthesis upon shift from stiff to soft are shown below.

Proteins GO terms

Epoxide hydrolase 1 Oxidation reduction

Aldo-keto reductase family 1 member C2 Nicotinamide metabolic process Aldo-keto reductase family 1 member B10 Alkaloid metabolic process Aldo-keto reductase family 1 member C1 Pyridine nucleotide metabolic process Nicotinamide phophoribosyltransferase Oxidoreduction coenzyme metabolic process

Aldose reductase Cellular amide metabolic process Isocitrate dehydrogenase [NADP] cytoplasmic Cellular aldehyde metabolic process Aldo-keto reductase family 1 , member B1 variant NADP metabolic process

Isoform 1 of Fatty aldehyde dehydrogenase Xenobiotic metabolic process

Kynuren Aromatic compound catabolic process Cathepsin D Steroid metabolic process

Gene Ontology (GO) terms are words or a string of words describing the properties of a gene product. See, The Gene Ontology Consortium (Jan 2008). "The Gene Ontology project in 2008." Nucleic acids research 36 (Database issue): D440-4. Listed above are GO Terms of the Biological Process domain associated with proteins that have significantly different synthesis rates in soft and stiff.

Proteins with the greatest H/L Ratio differences and with significant difference

(p<0.05) were used to find the top related GO Terms. Note the decreased synthesis of mitochondrial proteins, tubulins, and proteins related to translation, and maintained synthesis of proteins related to oxidation/reduction reactions and the breakdown of aromatic

compounds when transitioning from stiff to soft culture conditions.

Figure 15 shows the validation of the SILAC results presented above. Cells were grown for 5 days on gels, and plated on plastic for the indicated time. These results indicate that suitable targets for dormant cancer cells include tubulin, nicotinamide

phosphoribosyltransferase, phosphofructokinase, and epoxide hydrolase. Taken together, rigidity-dependent cells (A549 and MDA321) cultured on soft matrix have less Y397-phosphorylated FAK, have unaltered levels of cyclin D, exhibit an extended Gl phase, and have a decreased overall rate of protein synthesis. Moreover, proteomic analysis (SILAC) revealed that proteins whose synthesis exhibits the largest decrease in rate upon shift from stiff to soft culture conditions; these include tubulin subunits, translational components, and mitochondrial proteins. Proteomic analysis (SILAC) also revealed that proteins whose synthesis exhibits the least change in rate upon shift from soft culture conditions; these include proteins involved in the regulation of oxidation/reduction reactions and in the breakdown of aromatic compounds. The results presented herein also indicate that mechanical properties of the matrix can selectively regulate the translation of cellular proteins, contributing to the changes in cell growth and metabolism - properties that are relevant to cancer cell dormancy. Overall, these data indicate that cancer cell proliferation and protein synthesis are regulated by the mechanical properties of the microenvironment, and this is an important factor in determining the "dormancy" of tumor cells at metastatic sites. Exemplary proteins whose synthesis is most affected by rigidity (levels are significantly decreased on soft substrates) and exemplary proteins whose synthesis is least affected by rigidity (levels are relatively sustained on soft substrates) are listed below.

Proteins whose synthesis is most affected by rigidity (levels are significantly decreased on soft substrates):

Protein ID (IPI #) Name

IPI00009790 6-phosphofructokinase type C

IPI00645452 Tubulin, beta

IPI00217223 Multifunctional protein ADE2

IPI00299573 60S ribosomal protein L7a

IPI00171199 Isoform 2 of Proteasome subunit alpha type- 3

IPI00186290 Elongation factor 2

IPI00937615 Elongation factor 1 -gamma

IPI00644079 HNRPU protein

IPI00455383 Isoform 2 of Clathrin heavy chain 1

IPI00220740 Isoform 2 of Nucleophosmin

IPI00021439 Actin, cytoplasmic 1

IPI00024911 Endoplasmic reticulum resident protein 29

IPI00216308 Voltage-dependent anion-selective channel protein 1

IPI00759832 Isoform Short of 14-3-3 protein beta/alpha

IPI00794663 cDNA FLJ58687, highly similar to Tubulin alpha-4 chain

IPI00216587 40S ribosomal protein S8

IPI00792677 cDNA FLJ60097, highly similar to Tubulin alpha-ubiquitous chain

IPI00003918 60S ribosomal protein L4 IPI00550069 Ribonuclease inhibitor

IPI00294578 Isoform 1 of Protein-glutamine gamma- glutamy transferase 2

IPI00908881 Glucose-6-phosphate isomerase

IPI00744692 Transaldolase

IPI00029012 Eukaryotic translation initiation factor 3 subunit A

IPI00939174 Isoform 1 of Ubiquitin thioesterase OTUB 1

IPI00916517 Uncharacterized protein

IPI00019502 Isoform 1 of Myosin-9

IPI00793930 TUBA1B protein (tubulin, alpha)

IPI00011253 40S ribosomal protein S3

IPI00021840 40S ribosomal protein S6

IPI00453476 29 kDa protein

IPI00016801 Glutamate dehydrogenase 1, mitochondrial

IPI00216319 14-3-3 protein eta

IPI00917420 30 kDa protein

IPI00376215 Isoform 2 of DNA-dependent protein kinase catalytic subunit

IPI00018206 Aspartate aminotransferase, mitochondrial

Proteins whose synthesis is least affected by rigidity (levels are relatively sustained on soft substrates):

Protein ID (IPI #) Name

IPI00009896 Epoxide hydrolase 1

IPI00291483 Aldo-keto reductase family 1 member C3

IPI00967892 cDNA, FLJ92557

IPI00257508 Dihydropyrimidinase-related protein 2

IPI00479722 Proteasome activator complex subunit 1

IPI00003818 Kynureninase

IPI00027223 Isocitrate dehydrogenase [NADP] cytoplasmic

IPI00916111 Malate dehydrogenase

IPI00037448 Glyoxylate reductase/hydroxypyruvate reductase

IPI00007755 Ras-related protein Rab-21

IPI00013894 Stress-induced-phosphoprotein 1

IPI00784154 60 kDa heat shock protein, mitochondrial

IPI00184330 DNA replication licensing factor MCM2

IPI00947368 Protein

IPI00619966 NAD(P)H dehydrogenase [quinone] 1 isoform b

IPI00018873 Nicotinamide phosphoribosyltransferase

IPI00009342 Ras GTPase-activating-like protein IQGAP1

IPI00796116 Uncharacterized protein

IPI00009896 Epoxide hydrolase 2

IPI00291483 Aldo-keto reductase family 1 member C4

IPI00257508 Dihydropyrimidinase-related protein 3

IPI00479722 Proteasome activator complex subunit 2

IPI00178440 Elongation factor 1-beta

IPI00843975 Ezrin IPI00942979 Transketolase

IPI00018352 Ubiquitin carboxyl-terminal hydrolase isozyme LI

IPI00011229 Cathepsin D

IPI00413641 Aldose reductase

IPI00025512 Heat shock protein beta- 1

IPI00915808 Putative uncharacterized protein GARS

IPI00909570 cDNA FLJ56548, highly similar to Elongation factor 2

IPI00012837 Kinesin-1 heavy chain

IPI00027107 Tu translation elongation factor, mitochondrial precursor

IPI00783271 Leucine-rich PPR motif-containing protein, mitochondrial

IPI00555744 Ribosomal protein L14 variant

Figure 13 shows a schematic illustrating potential signaling processes regulated by extracellular rigidity. In nontransformed cells, rigidity-dependent cell cycle progression is regulated by the mechanosensor FAK through Rac activity and cyclin D expression. Also in nontransformed cells, adhesion-dependent regulation of gene expression occurs through the MAPK pathway, and protein synthesis is regulated by the AKT/mTOR pathway. In cancer cells, mutations that constituitvely activate these signaling pathways could lead to uncoupling of the FAK/Src mechanosensory complex with its downstream effectors. Protein synthesis, and perhaps other aspects of cellular metabolism, are regulated by the rigidity of the microenvironment, and this contributes to slower growth in soft tissue and tumor dormancy at distant metastatic sites.

As described herein, the rigidity of the extracellular matrix regulates cellular metabolism and protein synthesis in cancer cells. The rigidity-dependent cancer cell lines A549 and MDA-MB-231 sustain the expression of cyclin Dl when cultured on

polyacrylamide gels that have mechanical properties similar to that of soft tissue such as lung or breast. In spite of cyclin Dl expression, A549 cells, when cultured on soft gels, show a lengthening of the Gl phase of the cell cycle, suggesting that there may be defects in the synthesis of the structural and enzymatic components necessary for the transition into S phase. Accordingly, the rigidity-dependent cell lines show lower levels of cellular ATP levels when cultured on soft gels. Additionally, protein synthesis is slower under this condition, with the synthesis of specific structural proteins and glycolytic enzymes such as tubulin, actin, and phosphofructokinase especially sensitive to the decrease in extracellular rigidity. Proteins that were less sensitive to the change in rigidity included enzymes such as epoxide hydrolase and nicotinamide phosphoribosyltransferase that are involved in the metabolism of cellular macromolecules. Tumor dormancy is defined as a stage in cancer progression in which residual disease is present but is asymptomatic (Aguirre-Ghiso, J.A. 2007. Models, mechanisms and clinical evidence for cancer dormancy. Nat Rev Cancer. 7:834-46). Dormant cancer cells can reside undetected at sites distant from the primary tumor, pending subsequent growth and clinical recurrence. Studies support a model whereby cancer cells from the primary tumor will disseminate early during disease progression, so that by the time the primary tumor is detected, there may be dormant cancer cells already residing at distant sites. These cells are either temporarily nonproliferative, or there is a balance between proliferation and cell death, or the immune system is keeping the cell numbers in check. The latter possibility has come under scrutiny because of studies showing a lack of tumor recurrence in cancer patients that have undergone immuosuppressive therapy (Uhr, J.W., and K. Pantel. Controversies in clinical cancer dormancy. Proc Natl Acad Sci U S A. 108: 12396-400).

Studies have implicated the microenvironment as a key regulator of tumor cell dormancy and recurrence. Specifically, the properties of the extracellular matrix at sites of metastasis may play important roles in determining the state of disseminated tumor cells (Barkan, D., H. Kleinman, J.L. Simmons, H. Asmussen, A.K. Kamaraju, M.J. Hoenorhoff, Z.Y. Liu, S.V. Costes, E.H. Cho, S. Lockett, C. Khanna, A.F. Chambers, and J.E. Green. 2008. Inhibition of metastatic outgrowth from single dormant tumor cells by targeting the cytoskeleton. Cancer Res. 68:6241-50). However, prior to the invention described herein, mechanistic insight as to how dormancy occurs was scarce, since dormant cancer cells were difficult to detect and study in vivo, and there was a paucity of in vitro models for tumor cell dormancy.

Prior to the invention described herein, in vitro models for tumor dormancy have been limited to cells grown on chorioallantoic membranes (CAMs) or on 3D matrix derived from the Engelbreth-Holm-Swarm tumor (EHS matrix, or Matrigel) (Barkan, D., L.H. El Touny, A.M. Michalowski, J.A. Smith, I. Chu, A.S. Davis, J.D. Webster, S. Hoover, R.M. Simpson, J. Gauldie, and J.E. Green. Metastatic growth from dormant cells induced by a col-I-enriched fibrotic environment. Cancer Res. 70:5706-16; Barkan, D., H. Kleinman, J.L. Simmons, H. Asmussen, A.K. Kamaraju, M.J. Hoenorhoff, Z.Y. Liu, S.V. Costes, E.H. Cho, S. Lockett, C. Khanna, A.F. Chambers, and J.E. Green. 2008. Inhibition of metastatic outgrowth from single dormant tumor cells by targeting the cytoskeleton. Cancer Res. 68:6241-50). Matrigel is a mixture of laminin, collagen IV, and entactin, in addition to a variety of proteases and growth factors such as TGFb, FGF, EGF, PDGF, and IGF (Kleinman, H.K., and G.R. Martin. 2005. Matrigel: basement membrane matrix with biological activity. Semin Cancer Biol. 15:378- 86). Because it is generated from a tumor, the specific composition of Matrigel is not well defined, and it may vary from batch to batch, producing variability in experimental results. While the mechanical properties of Matrigel have been analyzed, they also are subject to variability because its polymerization is affected by its composition and other experimental factors such as temperature (Soofi, S.S., J.A. Last, S.J. Liliensiek, P.F. Nealey, and C.J. Murphy. 2009. The elastic modulus of Matrigel as determined by atomic force microscopy. J Struct Biol. 167:216-9). In addition, prior to the invention described herein, the only way to alter the mechanical properties of Matrigel was to add additional matrix components such as collagen I (Paszek, M.J., N. Zahir, K.R. Johnson, J.N. Lakins, G.I. Rozenberg, A. Gefen, C.A. Reinhart-King, S.S. Margulies, M. Dembo, D. Boettiger, D.A. Hammer, and V.M. Weaver. 2005. Tensional homeostasis and the malignant phenotype. Cancer Cell. 8:241-54), which may also complicate the interpretation of experimental results.

Described herein are polyacrylamide gels that mimic the mechanical properties of the soft tissue where metastases commonly occur, such as the lung, liver, and bone marrow. Unlike Matrigel, polyacrylamide is a stable, homogeneous polymer, and its mechanical properties are not sensitive to changes in temperature (Sunyer, R., X. Trepat, J.J. Fredberg, R. Farre, and D. Navajas. 2009. The temperature dependence of cell mechanics measured by atomic force microscopy. Phys Biol. 6:025009). Its rigidity is easily tunable by modulating the amount of bis crosslinker without affecting its ability to crosslink ECM molecules to its surface (Mih, J.D., A.S. Sharif, F. Liu, A. Marinkovic, M.M. Symer, and D.J. Tschumperlin. A multiwell platform for studying stiffness-dependent cell biology. PLoS One. 6:el9929). Acrylamide encompasses a spectrum of elastic moduli that includes a range of human tissues from fat to skeletal muscle, and a variety of ECM molecules can be conjugated to its surface, including collagen, fibronectin, and unpolymerized Matrigel. In addition, a high-throughput system was developed to enable the screening of small molecule inhibitors with cells cultured on polyacrylamide gels spanning a range of stiffnesses (Mih, J.D., A.S. Sharif, F. Liu, A. Marinkovic, M.M. Symer, and D.J. Tschumperlin. A multiwell platform for studying stiffness-dependent cell biology. PLoS One. 6:el9929; Tilghman, R.W., C.R. Cowan, J.D. Mih, Y. Koryakina, D. Gioeli, J.K. Slack-Davis, B.R. Blackman, D.J. Tschumperlin, and J.T. Parsons. Matrix rigidity regulates cancer cell growth and cellular phenotype. PLoS One. 5:el2905).

The growth of certain cancer cell lines slows when cultured on soft gels (Tilghman, R.W., C.R. Cowan, J.D. Mih, Y. Koryakina, D. Gioeli, J.K. Slack-Davis, B.R. Blackman, D.J. Tschumperlin, and J.T. Parsons. Matrix rigidity regulates cancer cell growth and cellular phenotype. PLoS One. 5 :e 12905). As described herein, cellular metabolic events are key regulators in this process. Prior to the invention described herein, little was known about the regulation of cellular metabolism by cell adhesion to the extracellular matrix. Prior to the invention described herein, Penman and colleagues showed that fibroblasts in suspension exhibited a decrease in protein synthesis that was rapidly recovered when the cells were replated onto plastic (Ben-Ze'ev, A., S.R. Farmer, and S. Penman. 1980. Protein synthesis requires cell-surface contact while nuclear events respond to cell shape in anchorage- dependent fibroblasts. Cell. 21 :365-72; Benecke, B.J., A. Ben-Ze'ev, and S. Penman. 1978. The control of mRNA production, translation and turnover in suspended and reattached anchorage-dependent fibroblasts. Cell. 14:931-9; Farmer, S.R., A. Ben-Ze'av, B.J. Benecke, and S. Penman. 1978. Altered translatability of messenger RNA from suspended anchorage- dependent fibroblasts: reversal upon cell attachment to a surface. Cell. 15:627-37), although the mechanism(s) that leads to the halt in protein synthesis are unclear. It is interesting that one of the proteins that is especially sensitive to changes in rigidity is phosphofructokinase-1 (PFK). PFK is a key regulator of glucose metabolism by catalyzing the conversion of fructose 6-phosphate to fructose 1,6-biphosphate, which is the first "committed" step in glycolysis. Levels of PFK are regulated by insulin and nutrients (Gehnrich, S.C., N. Gekakis, and H.S. Sul. 1988. Liver (B-type) phosphofructokinase mRNA. Cloning, structure, and expression. J Biol Chem. 263: 11755-9), and its levels are decreased in a rat model of diabetes (Dunaway, G.A., G.L. Leung, J.R. Thrasher, and M.D. Cooper. 1978. Turnover of hepatic phosphofructokinase in normal and diabetic rats. Role of insulin and peptide stabilizing factor. J Biol Chem. 253:7460-3), suggesting that PFK is an important regulatory point in glucose metabolism. This is in agreement with the observation that ATP levels are lower in cells cultured on soft gels.

As described herein, the synthesis of other metabolic enzymes such as epoxide hydrolase (EH) and nicotinamide phosphoribosyltransferase (Nampt) are maintained when A549 cells are moved from stiff to soft environments. EH is involved in the detoxification of byproducts resulting from the breakdown of cellular components, and is also known to be important for the detoxification steps during the metabolism of small molecule inhibitors (Fretland, A.J., and C.J. Omiecinski. 2000. Epoxide hydrolases: biochemistry and molecular biology. Chem Biol Interact. 129:41-59). Nampt is the rate-limiting step in the salvage pathway to generate NAD from nicotinamide (Rongvaux, A., F. Andris, F. Van Gool, and O. Leo. 2003. Reconstructing eukaryotic NAD metabolism. Bioessays. 25:683-90). The conservation of EH and Nampt in cells cultured on soft gels suggests that these cells are undergoing a form of autophagy to promote survival by metabolizing cellular components.

In summary, the observations described herein indicate a "slowing down" in the growth and metabolism of cancer cells as they move from a stiff to a soft environment. Future studies will be necessary to determine if these metabolic changes occur in vivo when cancer cells are placed in soft tissues to mimic the dissemination of the disease.

The patent and scientific literature referred to herein establishes the knowledge that is available to those with skill in the art. All United States patents and published or unpublished United States patent applications cited herein are incorporated by reference. All published foreign patents and patent applications cited herein are hereby incorporated by reference. All other published references, documents, manuscripts and scientific literature cited herein are hereby incorporated by reference.

OTHER EMBODIMENTS

While the invention has been described in conjunction with the detailed description thereof, the foregoing description is intended to illustrate and not limit the scope of the invention, which is defined by the scope of the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims.

The patent and scientific literature referred to herein establishes the knowledge that is available to those with skill in the art. All United States patents and published or unpublished United States patent applications cited herein are incorporated by reference. All published foreign patents and patent applications cited herein are hereby incorporated by reference. Genbank and NCBI submissions indicated by accession number cited herein are hereby incorporated by reference. All other published references, documents, manuscripts and scientific literature cited herein are hereby incorporated by reference. While this invention has been particularly shown and described with references to embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention encompassed by the appended claims.

Claims

CLAIMS What is claimed:

1. A method of determining tumor cell phenotype, comprising culturing a tumor cell on a first substrate and on a second substrate and measuring growth, morphology, migration, apoptosis, protein expression or gene expression in the tumor cell and calculating the difference in said growth, morphology, migration, apoptosis, protein expression or gene expression compared to said second substrate, wherein the first and the second substrates vary in stiffness, and wherein a change in growth, morphology, migration, apoptosis, protein expression or gene expression in the tumor cell present on the first substrate compared to on the second substrate indicates that the tumor cell has a rigidity dependent phenotype.

2. The method of claim 1, wherein the first and the second substrates have an elastic modulus in the range of 100 Pa to 150,000 Pa.

3. The method of claim 2, wherein the elastic modulus of the first substrate differs from the elastic modulus of the second substrate by at least 500 Pa.

4. The method of claim 1, wherein the first substrate has an elastic modulus in the range of 150 Pa to 1200 Pa and the second substrate has an elastic modulus in the range of 4800 Pa to 9600 Pa.

5. The method of claim 1, wherein the first substrate has an elastic modulus of 150 Pa and the second substrate has an elastic modulus of 4800 Pa or 9600 Pa.

6. The method of claim 1 , wherein at least a 2-fold difference between tumor cell number on the first substrate after culture for 5 days and tumor cell number on the second substrate after culture for 5 days indicates a rigidity dependent phenotype, and wherein less than a 2-fold difference indicates a rigidity independent phenotype.

7. The method of claim 1, wherein each of the substrates is present on a single multi-well culture plate.

8. A method of identifying an anti-tumor composition, comprising contacting a rigidity dependent cell with a candidate compound on a rigidity-matched substrate, detecting and calculating a difference in a level of proliferation in the presence of said candidate compound compared to that in its absence, wherein a decrease in cell proliferation in the presence of the candidate compound as compared to cell proliferation in the absence of the candidate compound indicates that the candidate compound comprises anti-tumor activity.

9. A method for identifying an anti-tumor composition, wherein the method comprises:

(a) preparing or obtaining a cell line derived from a cancer cell present in an in vivo environment in a tumor in a mammal;

(b) identifying the elastic modulus (stiffness) of the in vivo environment of the cancer cell in the mammal;

(c) identifying a rigidity-matched substrate for the cell line;

(d) culturing the cell line on the rigidity-matched substrate, and measuring 5- day cell growth of the cell line on the substrate in the presence of a candidate composition;

(e) culturing the cell line on the rigidity-matched substrate, and measuring 5- day cell growth of the cell line on the substrate in the absence of the candidate composition;

(f) comparing the 5 day growth rates measured in (d) and (e), wherein a decrease by at least two-fold in 5-day growth on the substrate in the presence of the candidate composition, as compared to 5-day growth on the substrate in the absence of the candidate composition, indicates that the candidate composition comprises an anti tumor composition; and

(g) identifying an anti-tumor composition.

10. A method for identifying an anti-tumor composition comprising:

(a) preparing or obtaining a cell line derived from a cancer cell present in an in vivo environment in a tumor in a mammal;

(b) measuring the elastic modulus of the in vivo environment of the cancer cell in the mammal;

(c) identifying a composition that reduces 5-day growth of the cancer cell line by at least two fold when grown in culture on a substrate having the elastic modulus identified in (b) in the presence of the composition, as compared to 5-day growth in culture on the substrate having the elastic modulus identified in (b) in the absence of the composition.

11. The method of claim 10, wherein the composition identified in step (c) does not reduce, or reduces by less than two fold, 5-day growth of the cancer cell line in culture on a substrate having an elastic modulus greater than 100 kPa in the presence of the composition, as compared to 5-day growth on a substrate having an elastic modulus greater than 100 kPa in the absence of the composition.

12. A method for determining whether a composition is a "false positive" antitumor composition comprising:

(a) preparing or obtaining a cell line derived from a cancer cell present in an in vivo environment in a tumor in a mammal;

(b) identifying a composition that reduces 5-day growth rate of the cell line by at least two fold when grown in culture on a substrate having an elastic modulus greater than 100 kPa in the presence of the composition, as compared to 5-day growth rate of the cell line on the substrate having the elastic modulus greater than 100 kPa in the absence of the composition;

(c) identifying the elastic modulus of the in vivo environment of the cancer cell in the mammal;

(d) determining whether the composition identified in (b) is a "false positive" anti-tumor composition by measuring whether the composition identified in (b) fails to reduce, or reduces by less than two fold, 5-day growth rate of the cell line in culture on a substrate having the elastic modulus identified in (c) in the presence of the composition, as compared to 5-day growth rate of the cell line on a substrate having the elastic modulus identified in (c) in the absence of the composition.

13. A method for treating a cancer in a mammal comprising:

(a) preparing or obtaining a cell line derived from a cancer cell present in an in vivo environment in a tumor in a first mammal;

(b) identifying the elastic modulus of the in vivo environment of the cancer cell in the mammal; (c) identifying a composition that reduces growth of the cancer cell line by at least two-fold when grown in culture over 5 days in the presence of the composition on a substrate having the elastic modulus identified in (b), as compared to growth over 5 days in the absence of the composition on a substrate having the elastic modulus identified in (b);

(d) identifying a second mammal that comprises cancer cells having the same phenotype as the cancer cell present in the tumor of the first mammal; and

(e) reducing or inhibiting growth of the cancer cells in the second mammal by administering to the second mammal the composition identified in (c).

14. A 96-well plate composition comprising five sections of plates, wherein each section comprises two columns of plates comprising collagen covalently coupled to polyacrylamide gels, and wherein the plates in each section have an elastic modulus of 150 Pa, 1200 Pa, 2400 Pa, 4800 Pa and 9600 Pa, respectively, wherein the 96-well plate composition further comprises at least one cancer cell line having a rigidity dependent phenotype.

15. A method for screening for an anti-tumor composition, wherein the method comprises:

(a) preparing or obtaining a cell line derived from a cancer cell present in an in vivo environment in a tumor in a mammal;

(b) identifying the elastic modulus of the in vivo environment of the cancer cell in the mammal;

(c) identifying a rigidity-matched substrate for the cell line;

(d) culturing the cell line on the rigidity-matched substrate, and measuring 5 -day cell growth of the cell line on the substrate in the presence of a candidate composition;

(e) culturing the cell line on the rigidity-matched substrate, and measuring 5 -day cell growth of the cell line on the substrate in the absence of the candidate composition;

(f) comparing the 5 day growth rates measured in (d) and (e), wherein a decrease by at least two-fold in 5-day growth on the substrate in the presence of the candidate composition, as compared to 5-day growth on the substrate in the absence of the candidate composition, indicates that the candidate composition comprises an anti tumor composition;

(g) repeating (d) - (f) with one or more additional different candidate compositions; and (h) identifying an anti-tumor composition.

16. A method for identifying whether a cancer cell population comprises mammalian dormant cancer cells, comprising:

(a) preparing or obtaining a cancer cell line derived from a cancer cell present in a mammal;

(b) culturing cancer cells from the cell line on a first soft substrate;

(c) culturing cancer cells from the cell line on a second more rigid substrate; and

(d) measuring in the cancer cells of (b) and (c) cultured for the same amount of time at least one selected from the group consisting of (i) cell cycle length, length of at least one cell cycle phase, or both; (ii) cellular ATP levels; and (iii) protein synthesis; and

(e) identifying whether the cancer cell line comprises mammalian dormant cancer cells based on results obtained upon measuring in (d), wherein at least one of an increase in (i), a decrease in (ii) and a decrease in (iii) in the cancer cells of (b), as compared to in the cancer cells of (c), indicates that the cell line comprises mammalian dormant cancer cells.

17. The method of claim 16, wherein said cancer cell population comprises cells obtained from a mammalian subject or comprises cells from an immortalized cell line.

18. The method of claim 16, wherein (d) further comprises measuring in the cancer cells of (b) and (c) cellular protein levels of at least one protein involved in regulating cytoskeleton structure, wherein a decrease in the protein levels in the cancer cells of (b), as compared to in the cancer cells of (c), indicates that the cell line comprises mammalian dormant cancer cells.

19. The method of claim 16, wherein (d) further comprises measuring in the cancer cells of (b) and (c) cellular protein levels of at least one of tubulin, actin, phosphofructokinase- 1 (PFK) and Slug, wherein a decrease in the protein levels in the cancer cells of (b), as compared to in the cancer cells of (c), indicates that the cell line comprises mammalian dormant cancer cells.

20. The method of claim 19, wherein (d) further comprises measuring in the cancer cells of (b) and (c) the rate of protein synthesis of at least one of E-cadherin, epoxide hydrolase, nicotinamide phosphoribosyltransferase, aldose reducase, kynureninase and aldo-keto reductase family 1, wherein a lack of difference or increase in the rates of protein synthesis in the cancer cells of (b), as compared to in the cancer cells of (c), indicates that the cell line comprises mammalian dormant cancer cells.

21. The method of claim 16, wherein (d) comprises measuring cellular ATP levels in the cancer cells of (b) and (c), wherein a decrease in cellular ATP levels by 50% or more in the cancer cells of (b), as compared to in the cancer cells of (c), when the cancer cells of (b) and (c) are cultured for about 24 hours or longer, indicates that the cell line comprises mammalian dormant cancer cells.

22. The method of claim 16, wherein the mammalian dormant cancer cells comprise, have properties of, or have the phenotype of, mammalian stem cancer cells.

23. A method of using a cancer cell line to obtain information regarding a mammalian dormant cancer cell, wherein the method comprises culturing, and obtaining information regarding, a cancer cell line that comprises mammalian dormant cancer cells, wherein the cancer cell line is identified by the method of claim 16.

24. The method of claim 23, wherein the information obtained comprises information relating to cellular metabolic state, signal transduction pathways, and gene expression profiles.

25. A method for identifying a cancer cell line that exhibits a morphology or phenotype similar to that of a mammalian dormant cancer cell, comprising:

(a) preparing or obtaining a cancer cell line derived from a cancer cell present in a mammal;

(b) culturing cancer cells from the cell line on a first soft substrate

(c) culturing cancer cells from the cell line on a second more rigid substrate; and (d) measuring in the cancer cells of (b) and (c) cultured for the same amount of time at least one selected from the group consisting of (i) cell cycle length, length of at least one cell cycle phase, or both; (ii) cellular ATP levels; and (iii) protein synthesis; and

(e) identifying whether the cancer cell line exhibits a morphology or phenotype similar to that of a mammalian dormant cancer cell based on results obtained upon measuring in (d), wherein at least one of an increase in (i), a decrease in (ii) and a decrease in (iii) in the cancer cells of (b), as compared to in the cancer cells of (c), indicates that the cell line exhibits a morphology or phenotype similar to that of a mammalian dormant cancer cell.

26. A method for identifying whether a candidate composition comprises a composition that can kill dormant cancer cells, wherein the method comprises:

(a) identifying a cancer cell line that comprises, or exhibits a morphology or phenotype similar to, a mammalian dormant cancer cell, comprising identifying a cancer cell line that exhibits at least one of (i) an increase in cell cycle length, (ii) a decrease in cellular ATP levels, and (iii) a decrease in protein synthesis when the cancer cell line is grown on a soft substrate, as compared when grown for the same amount of time on a more rigid substrate;

(b) culturing the cancer cell line identified in (a) on the soft substrate, and measuring cell death when growing the cell line in the presence of a candidate composition;

(c) culturing the cancer cell line identified in (a) on the soft substrate, and measuring cell death when growing the cell line in the absence of the candidate composition;

(d) comparing the cell deaths as measured in (b) and (c), wherein an increase by at least two-fold in cell death when growing the cell line for 2-5 days in the presence of the candidate composition, as compared to when growing the cell line for 2-5 days in the absence of the candidate composition, indicates that the candidate composition comprises a composition that can kill dormant cancer cells; and

(e) identifying whether the candidate composition comprises a composition that can kill dormant cancer cells.

27. The method of claim 12, wherein (e) further comprises identifying whether the candidate composition comprises a composition that can kill stem cancer cells.