ATP-based cell sorting and hyperproliferative cancer stem cells

Abstract

High mitochondrial ATP is a metabolic trait that confers hyperproliferation, stemness, anchorage-independence, antioxidant capacity, and multidrug resistance to cancer cells. In this approach, intracellular ATP levels can be used as a metabolic biomarker to identify, isolate, and purify aggressive, hyperproliferative cancer stem cell ("CSC") phenotypes. Furthermore, ATP can be combined with other CSC markers, such as CD44 or ALDH activity, to usefully fractionate CSC populations into subpopulations. For example, the ATP-high / CD44-high CSC subpopulation exhibited two-fold greater anchorage-independent growth than the ATP-low / CD44-high CSC subpopulation. Complementary bioinformatic data implicating mitochondrial ATP synthesis in stemness, metastasis, and the detection of circulating tumor cells ("CTCs"), as well as five members of the ATP-related metastasis gene signature (ABCA2, ATP5F1C, COX20, NDUFA2, and UQCRB), are also disclosed. The gene signature of this approach can be used to identify CSCs with dramatically increased cell migration and invasive capacity in vitro and spontaneous metastasis in vivo. This approach also provides a cellular platform for systematically targeting cancer cell stemness, multidrug resistance, and metastasis.

Description

Technical Field

[0001] Cross - reference to Related Applications This application claims the benefit of U.S. Provisional Patent Application No. 62 / 900,139, filed on September 13, 2019, which is hereby incorporated by reference in its entirety.

[0002] This disclosure relates to ATP - based cell sorting for identifying, isolating, and treating aggressive hyperproliferative cancer stem cell ( "CSC") phenotypes that are metabolically hyperactive and for preventing or reducing the potential for metastasis.

Background Art

[0003] Researchers have been making efforts to develop new cancer treatments. Conventional cancer therapies (e.g., radiation therapy, alkylating agents such as cyclophosphamide, and antimetabolites such as 5 - fluorouracil) attempt to selectively detect rapidly proliferating cancer cells and disrupt the cellular mechanisms involved in cell growth and DNA replication to eradicate them. Additionally, there are cancer therapies that use immunotherapy to selectively bind mutant tumor antigens to rapidly proliferating cancer cells (e.g., monoclonal antibodies). Unfortunately, it is not uncommon for tumors to recur at the same or different sites after such treatments, suggesting that not all cancer cells have been eradicated. The cause of recurrence may be insufficient chemotherapy dosage and / or the emergence of treatment - resistant cancer clones. Therefore, new cancer treatment strategies are needed.

[0004] Advances in mutation analysis have made it possible to study in detail the genetic mutations that occur during the development of cancer. In modern oncology, despite knowledge of the genomic situation, it has become difficult to identify the major driver mutations that span cancer subtypes. It seems that each patient's tumor is unique, and there is the harsh reality that a single tumor may contain multiple different clonal cells. What is needed, therefore, is a new approach that emphasizes the commonalities between different cancer types. Targeting the metabolic differences between tumors and normal cells is promising as a new cancer treatment strategy. Analysis of transcription profiling data of human breast cancer samples revealed an increase in more than 95 mRNA transcripts related to mitochondrial biogenesis and / or mitochondrial translation. Furthermore, more than 35 of the 95 upregulated mRNA transcripts encoding mitochondrial ribosomal proteins (MRPs). Similarly, proteome analysis of human breast cancer stem cells also revealed significant overexpression of several mitochondrial ribosomal proteins and other proteins related to mitochondrial biogenesis.

[0005] Mitochondria are extremely active organelles that constantly divide, elongate, and bind to each other to form tubular networks or fragmented microsomes in order to meet the demands of the cell and adapt to the cell's microenvironment. The balance between mitochondrial fusion and fission determines the morphology, quantity, function, and spatial distribution of mitochondria, thus affecting many important biological processes that depend on mitochondria, such as ATP production, mitophagy, apoptosis, and calcium homeostasis. On the other hand, mitochondrial dynamics may be controlled by mitochondrial metabolism, respiration, and oxidative stress. Therefore, it is not surprising that an imbalance in fission and fusion activities has an adverse effect on several pathological conditions, including cancer. Fragmented mitochondria are frequently seen in cancer cells, and an increase in fission and a decrease in fusion are often observed with cancer, but a comprehensive understanding of how mitochondrial dynamics affect tumorigenesis is still needed.

[0006] In particular, as tumor growth and metastatic seeding progress, it becomes necessary to maintain and enhance metabolic functions to meet the increased bioenergetic and biosynthetic demands of cancer cells. Naturally, mitochondrial-dependent metabolic pathways provide an essential biochemical platform for cancer cells by extracting energy from several fuel sources.

[0007] Cancer stem-like cells are a relatively small subpopulation of tumor cells that share characteristic features with normal adult stem cells and embryonic stem cells. Therefore, CSCs are thought to be the "major biological cause" of tumor recurrence and distant metastasis, ultimately leading to treatment failure and premature death in cancer patients undergoing chemotherapy and radiotherapy. Evidence suggests that CSCs also function in tumor initiation, as isolated CSCs behave experimentally as tumor-initiating cells (TICs) in preclinical animal models. Given that approximately 90% of cancer patients worldwide die prematurely from metastatic disease, the development of novel therapies that effectively target and eradicate CSCs is an urgent and unmet clinical need. Most conventional therapies do not target CSCs and often increase the frequency of CSC emergence in primary tumors and distant sites.

[0008] In recent years, energy metabolism and mitochondrial function have been associated with specific dynamics involved in the maintenance and proliferation of cancer stem-like cells (CSCs), a distinct cell subpopulation within tumor masses that is involved in tumor initiation, metastatic dissemination, and resistance to anticancer therapies. For example, CSCs exhibit a specific and unique increase in mitochondrial mass, enhanced mitochondrial biogenesis, and improved activity of mitochondrial protein translation. These behaviors suggest a strict dependence on mitochondrial function. Consistent with these observations, enhanced mitochondrial metabolic function and OXPHOS have been detected in CSCs of multiple tumor types.

[0009] One of the new strategies for eliminating CSCs is to utilize cell metabolism. CSCs are the most energy-rich cancer cells. In this method, metabolic inhibitors are used to induce ATP depletion and starve the CSCs. To date, the inventors have identified a number of FDA-approved drugs with anti-CSC properties and off-target mitochondrial side effects that induce ATP depletion, such as the antibiotic doxycycline, which functions as an inhibitor of mitochondrial protein translation. Doxycycline is a long-acting tetracycline analog and is currently used in the treatment of various forms of infectious diseases, such as acne, rosacea, and malaria prevention. In a recent Phase II clinical trial, preoperative oral administration of doxycycline (200 mg / day for 14 days) reduced the CSC load in early breast cancer patients by 17.65 - 66.67% with a positive response rate of nearly 90%.

[0010] However, there are certain limitations to the sole use of anti-mitochondrial agents in cancer therapy because tumor masses may be equipped with adaptive mechanisms to overcome the lack of mitochondrial function. These adaptive mechanisms include, for example, the ability of CSCs to shift from oxidative metabolism to alternative energy pathways in a multi-directional process of metabolic plasticity caused by both intrinsic and extrinsic factors in the tumor cell and its surrounding niche. In particular, manipulation of such metabolic flexibility in CSCs may be advantageous from a therapeutic perspective. What is needed, therefore, is a treatment method that prevents such metabolic shifts or utilizes the shifts to suppress the growth of cancer cells.

[0011] Adenosine 5'-triphosphate (ATP) is the biological energy "currency" of all living cells and organisms. Chemically, ATP is a nucleoside triphosphate containing adenine, ribose sugar, and a triphosphate group. When ATP is cleaved by the terminal phosphate group, two major reaction products, ADP and inorganic phosphate (Pi), are formed, thereby releasing a large amount of stored energy. In eukaryotic cells, mitochondria produce a large amount of ATP through the TCA cycle and oxidative phosphorylation (OXPHOS), while the amount of ATP obtained through glycolysis is minimal. Mitochondrial dysfunction causes ATP depletion and leads to apoptosis (cell death) by mitochondria.

[0012] In MCF7 breast cancer cells, OXPHOS by mitochondria is responsible for 80% of ATP production, and glycolysis is responsible for the remaining 20%. That is, like normal cells, cancer cells also greatly depend on mitochondrial ATP production. However, little is known about how the ATP level in cancer cells is related to the ability to cause anchorage-independent growth, which is a characteristic feature of "stemness", cell cycle progression, and metastatic spread.

[0013] Since ATP is very important as a barometer of cell metabolism, numerous luminescent probes and fluorescent probes have been developed to measure and track ATP levels in response to various cell stimuli. For example, ATP-Red1 (CAS number: 1847485-97-5, IUPAC name: [2-[3’,6’-bis(diethylamino)-3-oxospiro[isoindole-1,9’-xanthene]-2-yl]phenyl]boronic acid) becomes fluorescent only when bound to ATP and does not recognize other nutrients such as ADP. ATP-Red1 can dynamically visualize the ATP level in living cells and tissues.

[0014] The objective of the present disclosure is to describe an executable ATP depletion strategy that targets and eradicates up to "optimal" cancer cells.

[0015] Another object of the present disclosure is to explain the specific composition of cells with a particular hyperproliferative phenotype.

[0016] Another object of the present disclosure is to identify a new anti-cancer treatment approach that targets mitochondria and causes ATP depletion to metabolically starve CSCs.

Prior Art Documents

Patent Documents

[0017]

Patent Document 1

Patent Document 2

Patent Document 3

Patent Document 4

Patent Document 5

Patent Document 6

Patent Document 7

Patent Document 8

Summary of the Invention

[0018] This method describes the use of a fluorescent ATP imaging probe that metabolically fractions a cancer cell population and separates the hyperproliferative cell subpopulation. The resulting compositions can be used for many advantageous purposes, from rapid drug development and screening to predicting and preventing metastasis and drug resistance. The method also provides a method for metabolic fractionation of cancer cells, as well as a method for diagnosing and preventing metastasis, in addition to five gene signatures that are precursors of cancer metastasis.

[0019] Using the "stratification" of bioenergetic cells with ATP-based biomarkers, "optimal" cancer cells can be isolated for identification, diagnosis, treatment, and therapeutic drug development. Specifically, a cell population can be stained using a fluorescent ATP imaging probe such as Biotracker ATP-Red1, and the resulting ATP-based fluorescence can be used to metabolically fractionate the population into ATP-high, and further into bulk and ATP-low subpopulations as needed. The data disclosed herein using this novel method includes the first evidence that large amounts of mitochondrial ATP are a major determinant of the behavior of aggressive cancer cells, including spontaneous metastasis.

[0020] Cancer cell populations have a very high degree of phenotypic diversity and metabolic heterogeneity. This heterogeneity allows "optimal" cancer cells to evade current therapies, resulting in tumor recurrence and distant metastasis due to drug resistance.

[0021] High intracellular ATP levels can be used as a metabolic biomarker indicative of an aggressive and proliferative cancer cell phenotype. In this method, fluorescence ATP markers such as the vital staining dye BioTracker® ATP-Red1 (EMD Millipore Corporation, Billerica, Massachusetts, USA) are used to quantify the mitochondrial ATP levels in a cancer cell population, and ATP-high and ATP-low cancer cell subpopulations can be isolated by flow cytometry. From the phenotypic analysis of these subpopulations, high mitochondrial ATP levels have been shown to be a metabolic trait that confers hyperproliferation, stemness, anchorage independence, antioxidant capacity, and multidrug resistance to cancer cells. Quantitatively similar results were obtained in four human breast cancer cell lines, MCF7, T47D, MDA-MB-231, and MDA-MB-468.

[0022] By combining ATP with other CSC markers, such as CD44 or ALDH activity, the CSC population can advantageously be fractionated into two subpopulations. The CD44-high / ATP-high subpopulation has a level of anchorage-independent growth that is approximately two-fold higher than that of the CD44-high / ATP-low subpopulation. Thus, CD44-high / ATP-low cancer cells are a highly quiescent CSC population. Importantly, these results indicate that ATP levels can be a functional regulator of quiescence in CSCs.

[0023] This method also includes complementary bioinformatics data related to mitochondrial ATP synthesis in the stemness, metastasis, and detection of circulating tumor cells (CTCs). Disclosed herein is an ATP-related metastasis gene signature of five members, including ABCA2, ATP5F1C, COX20, NDUFA2, and UQCRB. According to the clinical findings based on these metastases, ATP-high MDA-MB-231 cells showed a dramatic increase in both cell migration and cell invasion in vitro and the ability to undergo spontaneous metastasis in vivo.

[0024] Thus, the present method systematically identifies, investigates, and targets the stemness, multidrug resistance, and metastasis of cancer cells, providing a new cell platform. The present disclosure also mechanistically explains the beneficial therapeutic benefits of i) nutrient deprivation and ii) calorie restriction mimetics for improving cancer therapy by inducing ATP depletion.

[0025] In embodiments of the present method, the vital dye ATP-Red1 is used as a molecular probe to identify and isolate ATP-high and ATP-low subpopulations of cells, more specifically, cancer cells and CSCs. The ATP-high subpopulation of cancer cells is large, energy-rich, hyperproliferative, and capable of anchorage-independent growth, thus corresponding to a more "stem-like" phenotype. Such ATP-rich cells can eradicate the energetically "optimal" CSCs, reduce drug resistance, and prevent metastasis using ATP depletion therapy.

[0026] Some embodiments of the present method can take the form of a purified composition of hyperproliferative cancer stem cells in the form of a subpopulation of cells from a human cancer cell population, where the cancer cell population expresses a range of fluorescence signals in response to a fluorescent adenosine triphosphate (ATP) imaging probe, and the subpopulation of cells expresses above a range of ATP-based fluorescence signals. The fluorescent ATP imaging probe can be, for example, BioTracker ATP-Red1. The upper portion, i.e., the ATP-high subpopulation, can be the top 10%, 5%, or 1% of the ATP-based fluorescence signal, depending on the embodiment. Other portions can also be used. In some embodiments, the composition is positive for the CD44 marker. In some embodiments, the composition is positive for the ALDH marker. In some embodiments, the composition is frozen.

[0027] In some embodiments, the method can take the form of a purified cell composition that includes a subpopulation of cancer stem cells stained with a fluorescent ATP imaging probe and that expresses a target portion of the ATP-based fluorescence signal range of the cancer cell population. The cancer cell population expresses a range of ATP-based fluorescence signals, and the target portion of the ATP-based fluorescence signal range can be the upper portion (e.g., the ATP-high subpopulation) and / or the lower portion (e.g., the ATP-low subpopulation) of the ATP-based fluorescence signal. The target portion can be the upper or lower 10%, 5%, or 1% of the ATP-based fluorescence signal, or other portion selected.

[0028] Some embodiments can take the form of a purified composition of cells obtained by staining a human cancer cell population with a fluorescent ATP imaging probe, separating a fraction of the human cancer cell population having a target portion of the ATP-based fluorescence signal, and purifying the separated cells. The target portion can be, for example, the upper 10% of the ATP-based fluorescence signal, the upper 5% of the ATP-based fluorescence signal, the lower 10% of the ATP-based fluorescence signal, the lower 5% of the ATP-based fluorescence signal, and the like. The separated cells are positive for one of the CD44 marker and the ALDH marker.

[0029] Some embodiments may take the form of an ATP-based method of cell fractionation. Cells of a cell population can be stained with a fluorescent ATP imaging probe that fluoresces when bound to ATP. The ATP-based fluorescence signal of the stained cells in the cell population can be measured. The stained cells can be separated based on a target portion of the ATP-based fluorescence signal. Fluorescence-activated cell sorting (FACS) and gating of a target portion of the ATP-based fluorescence signal can be used to separate the stained cells. Gates can be set and stained cells having the top 10% of the measured fluorescence signal, and / or stained cells having the bottom 10% of the measured fluorescence signal can be recovered. Of course, other percentages can also be used. The cell population can be derived from, for example, blood, urine, saliva, tumor tissue, non-cancerous tissue, or metastases. Some embodiments may further include measuring the ALDH activity of the separated cells, measuring the anchorage-independent growth of the separated cells, measuring the mitochondrial mass of the separated cells, measuring the glycolytic and oxidative mitochondrial metabolism of the separated cells, measuring the cell cycle progression and proliferation rate of the separated cells, and measuring the ploidy of the separated cells.

[0030] Embodiments of this method can take the form of a method for separating and recovering metabolically active cells from a cell population. The cells of the cell population can be stained with an ATP-labeling dye that emits fluorescence when bound to ATP. The fluorescence signal of the stained cells can be measured in the cell population, and then the stained cells can be measured based on the measured fluorescence signal. At least a portion of the separated cells, where the measured fluorescence signal is either above a predetermined threshold or below a predetermined threshold, can be recovered, for example, using a FACS apparatus. The predetermined threshold includes the upper part of a certain percentage of the measured fluorescence signal, such as the top 25%, top 20%, top 15%, top 10%, top 5%, top 2%, and top 1%. Without departing from this method, other percentages can also be used. In some embodiments, the separated cells can be further separated based on a second marker, such as CD44(+), CD133(+), ESA(+), ALDEFLOUR(+), MitoTracker-High, EpCAM(+), CD90(+), CD34(+), CD29(+), CD73(+), CD90(+), CD105(+), CD106(+), CD166(+), and Stro-1(+). Without departing from this method, other markers can also be used. The second marker can, in some embodiments, take the form of magnetic beads coated with an antibody.

[0031] This method can take the form of a method for identifying and treating cancer stem cells in a biological sample. The biological sample can be obtained from a patient, and then the cells in the biological sample can be stained with an ATP-labeled dye, which fluoresces when it binds to ATP. The fluorescence signal of the stained cells in the cell population can be measured and then compared with a predetermined threshold indicating the presence of cancer stem cells. When the measured fluorescence signal exceeds the predetermined threshold, an ATP depletion therapeutic agent can be administered to the patient. The ATP depletion therapeutic agent can be, for example, doxicycline, tigecycline, azithromycin, pyrvinium pamoate, atovaquone, bedaquiline, niclosamide, irinotecan, actinonin, CAPE, berberine, bruceellidine, meridianin, oligomycin, AR-C155858, Mitoriboscin, Mitoketoscin, Mitoflavoscin, a TPP derivative, dodecyl TPP, 2-butene-1,4-bis-TPP, or a combination of doxicycline, azithromycin, and ascorbic acid.

[0032] In some embodiments, the method may take the form of a method for testing candidate compounds for anti-cancer activity. The cancer cell population can be stained with an ATP-labeling dye that emits fluorescence when bound to ATP, such as BioTracker ATP-Red1. The ATP-based fluorescence signal of the stained cells can be measured, and the stained cells can be separated based on a target portion of the ATP-based fluorescence signal to create a hyperactive cancer cell subpopulation. The candidate compound can be administered to the hyperactive cancer cell subpopulation, and the effect of the candidate compound on the hyperactive cancer cell subpopulation can be measured. The ATP-labeling dye can be BioTracker ATP-Red1. The target portion of the ATP-based fluorescence signal can be, for example, the top 25%, top 20%, top 15%, top 10%, top 5%, top 2%, and top 1%. In some embodiments, the hyperactive cancer cell subpopulation is positive for one of the CD44 marker and the ALDH marker. Embodiments may also include measuring the ALDH activity of the hyperactive cancer cell subpopulation, measuring the anchorage-independent growth of hyperactive cancer cell subpopulation cells, measuring the mitochondrial mass of the hyperactive cancer cell subpopulation, measuring the glycolytic and oxidative mitochondrial metabolism of the hyperactive cancer cell subpopulation, measuring the cell cycle progression and proliferation rate of the hyperactive cancer cell subpopulation, and measuring the ploidy of the hyperactive cancer cell subpopulation.

[0033] This method can take the form of a method for diagnosing and preventing the risk of metastasis in cancer patients. Determine the expression levels of the gene signatures of five members, ABCA2, ATP5F1C, COX20, NDUFA2, and UQCRB, in a biological sample of the patient's cancer, and then it can be compared with the baseline expression levels of ABCA2, ATP5F1C, COX20, NDUFA2, and UQCRB in the patient's non-cancerous biological sample. When the detected expression level exceeds the baseline expression level, an ATP depletion compound can be administered to the patient. The ATP depletion compound can be, for example, doxicycline, tigecycline, azithromycin, pyrvinium pamoate, atovaquone, bedaquiline, niclosamide, irinotecan, actinonin, CAPE, berberine, bruceine, melittidin, oligomycin, AR-C155858, mitriboskin, mitoketoskin, mitoflavoskin, a TPP derivative, dodecyl TPP, 2-butene-1,4-bis-TPP, or a combination of doxicycline, azithromycin, and ascorbic acid.

[0034] Some embodiments can take the form of a kit for identifying peripheral circulating tumor cells in a biological sample. The kit can include reagents for identifying the upregulation of ABCA2, ATP5F1C, COX20, NDUFA2, and UQCRB in a biological sample, such as antibodies targeting the proteins encoded by these genes. The kit can be used, for example, for a liquid biopsy method for detecting CTCs.

[0035] This method can also take the form of a method for detecting peripheral circulating tumor cells (CTCs) in a biological sample. The expression levels of ABCA2, ATP5F1C, COX20, NDUFA2, and UQCRB in the biological sample can be determined, and then, when the determined expression level is upregulated relative to the control, CTCs are identified as present. The biological sample can be, for example, blood, urine, saliva, tumor tissue, non-cancerous tissue, or a metastasis. The sample can be further processed using the methods described herein to isolate ATP-high cells.

[0036] Such embodiments will be apparent to those skilled in the art in view of the present specification, the claims appended hereto, and applications incorporated herein by reference.

Brief Description of the Drawings

[0037]

Figure 1A

Figure 1B

Figure 1C

Figure 1D

Figure 2A

Figure 2B

Figure 2C

Figure 2D

Figure 2E

Figure 2F

Figure 2G

Figure 2H

Figure 2I

Figure 2J

Figure 2K

Figure 2L

Figure 2M

Figure 2N

Figure 2O

Figure 2P

Figure 2Q

Figure 3A

Figure 3B

Figure 3C

Figure 4A

Figure 4B

Figure 4C

Figure 5A

Figure 5B

Figure 6A

Figure 6B

Figure 7A

Figure 7B

Figure 7C

Figure 7D

Figure 7E

Figure 8A

Figure 8B

Figure 9A

Figure 9B

Figure 10A

Figure 10B

Figure 10C

Figure 10D

Figure 10E

Figure 11A

Figure 11B

Figure 11C

Figure 11D

Figure 12A

Figure 12B

Figure 13A

Figure 13B

Figure 13C

Figure 13D

Figure 14A

Figure 14B

Figure 15

Figure 16A

Figure 16B

Figure 16C

Mode for Carrying Out the Invention

[0038] The following description shows embodiments of the present method in sufficient detail to enable the implementation of the present method. The present method is described with reference to these specific embodiments. However, of course, the present method can be embodied in different forms, and this description should not be construed as limiting any appended claims to the specific embodiments described herein. Rather, by providing these embodiments, the present disclosure is made complete and full, and the scope of the present method is fully conveyed to those skilled in the art.

[0039] In this description, various terms that should be understood by those skilled in the art are used. Clarifications are made below to avoid misunderstandings. The terms "treat", "treated", and "treatment" include the alleviation or mitigation of a condition, disorder, or disease being treated, particularly at least one symptom associated with or caused by cancer. In certain embodiments, treatment includes alleviating and / or mitigating at least one symptom associated with or caused by cancer treated by a compound of the invention. In some embodiments, treatment includes causing the death of certain cancer cells, such as CSCs, in a host, preventing the further proliferation of cancer cells, and / or inhibiting CSC function by, for example, depriving such cells of mechanisms that generate energy. For example, treatment may be the alleviation of one or more symptoms of cancer or the eradication of cancer. As another example, this approach can be used to inhibit mitochondrial metabolism in cancer, eradicate CSCs in cancer (e.g., kill at a faster rate of proliferation), eradicate TICs in cancer, eradicate peripheral circulating tumor cells in cancer, inhibit cancer growth, target and inhibit CSCs, target and inhibit TICs, target and inhibit peripheral circulating tumor cells, prevent metastasis (i.e., reduce the likelihood of metastasis), prevent recurrence, enhance cancer sensitivity to chemotherapy, enhance cancer sensitivity to radiotherapy, and enhance cancer sensitivity to phototherapy.

[0040] The terms "cancer stem cell" and "CSC" refer to a subpopulation of cancer cells within a tumor that, when transplanted into an animal host, have the ability to self-renew, differentiate, and form tumors. CSCs have an increased mitochondrial mass, enhanced mitochondrial biogenesis, and improved activity of mitochondrial protein translation compared to "bulk" cancer cells. As used herein, "peripheral circulating tumor cells" are cancer cells that have flowed from a primary tumor into the vasculature or lymphatic system and are carried throughout the body via the bloodstream. Peripheral circulating tumor cells can be detected using the CellSearch peripheral circulating tumor cell assay.

[0041] The terms "ATP-high" and "ATP-low" refer to subpopulations of cells having an ATP-based fluorescence signal, representing the upper and lower portions of the ATP-based fluorescence signal in the starting cell population, respectively. The upper portion may represent the top 25%, or top 20%, or top 15%, or top 10%, or top 5%, or top 2%, or top 1% of the ATP-based fluorescence signal of the starting cell population. The lower portion may represent the bottom 25%, or bottom 20%, or bottom 15%, or bottom 10%, or bottom 5%, or bottom 2%, or bottom 1% of the ATP-based fluorescence signal of the starting cell population.

[0042] As used herein, the term "pharmaceutically effective amount" refers to the amount necessary to be administered to a host, or to a cell, tissue, or organ of a host, to achieve a therapeutic result such as modulation, regulation, or inhibition of protein kinase activity, e.g., inhibition of the activity of a protein kinase, or treatment of cancer. A physician or veterinarian having ordinary skill in the art can readily determine and prescribe the effective amount of the pharmaceutical composition required. For example, a physician or veterinarian could start with a dose of the compound of the invention used in the pharmaceutical composition at less than the amount required to achieve the desired therapeutic effect and gradually increase the dose until the desired effect is obtained.

[0043] Bioinformatics analysis reveals the role of mitochondrial ATP synthesis in three-dimensional matrix-independent growth, stemness, and distant metastasis. In particular, mitochondrial ATP synthesis is an important determinant of three-dimensional matrix-independent growth and metastasis in bioinformatics approaches. Existing proteome profiling data were queried, and two different ER(+) breast cancer cell lines (MCF7 and T47D) were compared in two-dimensional monolayers and three-dimensional mammospheres. Overall, in three-dimensional mammospheres, 22 ATP-related proteins out of 1,519 proteins common to both cell lines were found to be upregulated in both datasets. Table 1 below shows the accession numbers of these proteins and the fold changes in expression levels in MCF7 and T47D cells (spheres and two-dimensional adherent cultures). Among the 21 ATP-related proteins, seven subunits of mitochondrial ATP synthase, including ATP5F1B, ATP5F1C, ATP5IF1, ATP5MG, ATP5PB, ATP5PD, and ATP5PO, were detected. Using Ingenuity Pathway Analysis (IPA) software, predicted upstream regulators of three-dimensional matrix-independent growth were highly conserved between the two cell lines and were observed to be specifically associated with existing IPA datasets related to tumor growth and tumor cell proliferation.

[0044] [Table 1]

[0045] In addition, the GEO transcription profiling dataset was re-analyzed to compare two-dimensional culture, three-dimensional culture, and in vivo tumor growth of MDA-MB-231 cells (triple-negative breast cancer cell line). Figure 1A shows a heatmap of ATP-related genes that were transcriptionally upregulated relative to all two-dimensional adherent cultures under both three-dimensional culture conditions (anchorage-independent tumors and in vivo tumors). The first column shows gene names, the second column shows the expression profile of two-dimensional MDA-MB-231 cells, the third column shows the expression profile of three-dimensional MDA-MB-231 cells, and the fourth column shows the expression profile of MDA-MB-231 cells in xenografts. Dark cells indicate a small fold change, and bright cells indicate a large fold change. The heatmap shows the log of the fold change, for example, the brightest cells are + / -4. In the two-dimensional MDA-MB-231 column, bright cells indicate a negative change (e.g., the fold change of ATP11A-AS1 is -4log), whereas in the three-dimensional and xenograft columns, bright cells indicate a positive change (e.g., the fold change of ATP12A is 4log).

[0046] The transcriptional expression of ATP-related genes (OXPHOS and ATP-related transporters) in two different GEO DataSets related to human breast cancer metastasis was useful for identifying ATP-related genes associated with metastasis. Figures 1B and 1C show the volcano plots of the GSE2034 and GSE59000 GEO DataSets. Specifically, in Figure 1B, gene expression in scenarios with and without metastasis was compared (GSE2034), and in Figure 1C, gene expression in metastatic and primary tumor scenarios was compared (GSE59000). The volcano plots were created by examining the annotations provided by OncoLand's Metastatic Cancer (QIAGEN OmicSoft Suite) and performing a functional "core analysis" using Ingenuity Pathway Analysis software (IPA; QIAGEN) on genes annotated with an uncorrected p-value cutoff <0.05. The transcriptional profiles of ATP-related genes (OXPHOS and ATP-related transporters) were specifically associated with an increase in both GEO DataSets related to metastasis.

[0047] Figure 1D shows a Venn diagram that intersects two breast cancer metastasis GEO DataSets (GSE2034 and GSE59000) used to identify ATP-related genes highly upregulated in both data sets as prognostic biomarkers for metastasis. The intersection of the two GEO DataSets was performed using IPA software as described in relation to Figures 1B and 1C. There were only 5 ATP-related genes included in the overlapping set of 1,055 genes. These ATP-related genes, ABCA2, ATP5F1C, COX20, NDUFA2, and UQCRB were highly upregulated in both metastatic GEO DataSets and thus have prognostic significance for predicting cancer metastasis. These 5 ATP-related genes could be used as an ATP-related metastasis gene signature indicative of metastasis. Among them, ATP5F1C (also known as ATP5C1) encodes the γ subunit of the soluble catalytic core F1 of mitochondrial ATP synthase. UQCRB is an essential component of mitochondrial complex III, functionally binds to ubiquinone, and is involved in electron transport. COX20 is a chaperone essential for the construction of mitochondrial complex IV. NDUFA2 is essential for the function of mitochondrial complex I. Also, ABCA2 is a member of the ATP-binding cassette transporter gene family.

[0048] In bona fide breast cancer metastases, ATP5F1C transcriptional expression is positively correlated with the co-expression of i) 5 metastasis marker genes (EPCAM, MKI67, RRP1B, VCAM1, CXCR4); ii) 4 cell cycle control genes (CDK1, CDK2, CDK4, CDK6); and iii) 11 CSC marker genes (CDH1, ALDH2, ALDH1BA1, ALDH9A1, SOX2, VIM, CDH2, ALDH7A1, ALDH1B1, CD44, ALDH3B2, listed in order of statistical significance). Figures 2A - 2N are data plots showing the positive correlation of ATP5F1C with each of the genes CDH1, ALDH2, SOX2, VIM, CD44, EPCAM, MKI67, RRP1B, CXCR4, VCAM1, CDK1, CDK2, CDK4, and CDK6 in that order.

[0049] Furthermore, the transcriptional expression of ATP5F1C is also positively correlated with the co-expression of three of the other five members of mitochondrial complexes I-V, transcripts encoded by mt-DNA, and the metastasis gene signature, namely UQCRB, COX20, and NDUFA2. Figures 2O-2Q are data plots showing the positive correlations of ATP5F1C with UQCRB, COX20, and NDUFA2, respectively. The expression of two members of this metastasis gene signature, ATP5F1C and UQCRB, is functionally correlated with the high levels and function of maximal oxygen uptake (V 02max ) and the proportion of type 1 fibers (rich in mitochondria).

[0050] The expression of ATP5F1C in skeletal muscle also significantly increases in response to the improvement of the patient's muscle strength after exercise training. Conversely, the expression of ATP5F1C decreases with age and is reduced in patients with progeria syndrome. These results strongly suggest that the high expression of ATP5F1C is a biomarker for increased mitochondrial ATP production at the cellular level.

[0051] In particular, in ER(+) patients who were lymph node negative at diagnosis and received tamoxifen treatment (hazard ratio (recurrence-free survival period)=2.77; P=3.4E-06; N=471), it was confirmed using Kaplan-Meier (KM) analysis that ATP5F1C serves as a prognostic biomarker for distant metastasis and tumor recurrence. Figure 3A shows the Kaplan-Meier curve of ER(+), recurrence-free survival period (N=3,082), Figure 3B shows the Kaplan-Meier curve of ER(+), distant metastasis-free survival period (N=1,395), and Figure 3C shows the Kaplan-Meier curve of ER(+), lymph node negative, with a history of tamoxifen treatment, recurrence-free survival period (N=471).

[0052] These results are consistent with previous studies showing that mitochondrial activity is functionally upregulated in breast cancer metastases within surgically resected lymph nodes, using histochemical activity staining to detect mitochondrial complex IV. Furthermore, the inventors previously pointed out that 16 members of the ATP5 gene family, including ATP5F1C (4.64-fold; p = 1.14E-05), are transcriptionally upregulated in human breast cancer cells relative to adjacent stromal cells, using samples from N = 28 breast cancer patients.

[0053] Using existing GEO DataSets (GSE55470), the use of ATP-related genes and OXPHOS genes as transcriptional biomarkers for breast cancer peripheral blood circulating tumor cells (CTCs) was evaluated in patients. Figure 4A shows a heatmap of the ATP-ABC gene expression profiles in the dataset, including a legend. Figure 4B shows a heatmap of the OXPHOS gene expression profiles based on the same legend as in Figure 4A. Generally, the lighter the color of the cell, the higher the absolute value of the expression. In the black and white used in connection with this application, it is difficult to distinguish between positive and negative fold changes. Most of the first 5 columns showing control blood are green in the original heatmap, indicating a negative fold change in expression. Most of the remaining cells are red, indicating a positive fold change in expression. Generally, a high ATP content in CTCs may be useful as a biomarker for identifying and tracking CTCs in whole blood, thereby potentially improving cancer diagnosis and preventing metastatic spread.

[0054] Figure 4C shows the results of Western blot analysis of MDA-MB-231 cells in the ATP-high and ATP-low subpopulations. The results indicate that both mitochondrial markers and CTC markers are upregulated in ATP-high MDA-MB-231 cells. Mitochondrial markers of Complexes I-V, including ATP5F1C, were all overexpressed in ATP-high subpopulation MDA-MB-231 cells compared to ATP-low subpopulation MDA-MB-231 cells. Furthermore, two known markers of CTC and metastasis (VCAM-1 and Ep-CAM) were overexpressed in the ATP-high subpopulation. β-actin and β-tubulin were used as markers for equal protein loading.

[0055] Collectively, bioinformatics data and analysis reveal that increased mitochondrial ATP synthesis may be an important driver of three-dimensional matrix-independent growth and metastasis. Based on this analysis, cancer cells with the highest ATP levels are thought to be more proliferative, more stem-like, grow independently of three-dimensional matrices, and exhibit other aggressive behaviors compared to cancer cells with low ATP levels. Similarly, cells with the lowest ATP levels are considered to be highly quiescent. Both subpopulations are of great value, especially for use in cancer research and drug screening. This approach provides a method for separating these subpopulations from a cancer cell population by metabolic fractionation. There are numerous subpopulations within a cancer cell population. CSCs are a small subpopulation of cancer cells that have self-renewal ability, are capable of differentiation, and show tumor-forming ability when transplanted. However, as described herein, not all CSCs are generated in the same way. CSCs separated and purified based on ATP levels have unique phenotypic characteristics not seen even in naturally occurring cancer cell populations or CSCs separated and purified using conventional markers such as CD44, CD24, and CD133.

[0056] In this method, cells, preferably cancer cells, can be fractionated based on metabolic conditions using a fluorescent ATP-labeling dye such as ATP-Red1 and flow cytometry. Ultimately, the ATP level determines phenotypic traits of cancer cells such as "stemness" and proliferative ability. Therefore, using an ATP-labeling dye, "optimal" cancer cells in terms of energy can be identified and purified from the whole cell population. The inventors selected the fluorescent biostaining dye BioTracker ATP-Red1 and labeled the ATP of live cancer cells. Naturally, other fluorescent ATP imaging probes, including those developed later, can also be used without departing from this method. Preferably, the fluorescent ATP imaging probe targets mitochondrial ATP.

[0057] ATP-Red1 is usually non-fluorescent but becomes fluorescent when bound to ATP and does not bind to other related nucleotides and metabolites including ADP. More specifically, BioTracker ATP-Red1 does not recognize sugars (arabinose, galactose, glucose, fructose, ribose, sorbose, sucrose, or xylose) or other nucleotides (AMP, ADP, CMP, CDP, CTP, UMP, UDP, UTP, GMP, GDP, or GTP). Importantly, this fluorescent ATP imaging probe shows a "turn-on" fluorescent reaction to ATP, with fluorescence enhanced by about six-fold. Using a fluorescence microscope, ATP-Red1 is mainly detected within mitochondria, which are the main source of ATP production in cells. Therefore, ATP-Red1 is preferable as a fluorescent probe for metabolically fractionating cancer cell populations by flow cytometry.

[0058] After separation or fractionation into ATP-high and ATP-low cell subpopulations, the cancer cell population can be phenotypically characterized. The subpopulations can be defined with respect to the ratio of upper and lower fluorescence signals (e.g., upper and lower 20%, upper and lower 1%, etc.), and the FACS gating cut-off values for cell selection and recovery are determined based on the selected ratio. The data disclosed herein mainly relied on upper and lower 5% and upper and lower 10% as gating cut-off values, but of course other ratios can be used without departing from the present method. Needless to say, the ratio should be less than 50%, and it would be expected that the larger the ratio, the lower the specificity of the phenotypic characterization for a given cell population.

[0059] Figure 5A shows an embodiment of the metabolic fractionation procedure according to this embodiment. The fluorescent ATP imaging probe can be dissolved and incubated in the medium (501). The results described herein were obtained by dissolving 5 μM of Biotracker ATP-Red as a fluorescent ATP imaging probe in the medium and incubating with cells for 30 minutes. Next, the cells were washed with PBS, trypsinized, resuspended in FACS buffer, and filtered through a 40-μm cell strainer (503). Cells obtained from 3D sphere or 2D adhesion conditions were analyzed using a FACS sorter (e.g., SH800 from SONY) (505). The cells were gated and sorted at a desired ATP content using an ATP-based fluorescent signal (e.g., top / bottom 1%, 2%, 3%, 4%, 5%, 10%, etc.) (507). Figure 5B shows an example of metabolic fractionation of MCF7 cells by ATP-Red1 to isolate ATP-high (top 5%) and ATP-low (bottom 5%) cell subpopulations. A bulk (5%) population was also selected for comparison. The right figure shows the number of cells at various fluorescence intensities, identifying the regions of the ATP-high (top 5%) subpopulation, ATP-low (bottom 5%) subpopulation, and bulk medium. The left figure of Figure 5B shows the average value of the ATP-based fluorescent signal for each subpopulation. Based on the average signal intensity, the ATP-high MCF7 cells are estimated to have an ATP level approximately 15-fold higher than the ATP-low population and 2-fold higher than the bulk cell population.

[0060] Figures 6A and 6B show the results of a continuous real-time assay system for cell proliferation in all three cell subpopulations (ATP-low 5%, bulk 5%, ATP-high 5%). Cell proliferation was evaluated using the xCELLigence® RTCA DP instrument. The cells were first sorted by ATP content, counted, and then seeded into an RTCA DP E-Plate (in common medium at 1×10 4) and performed real-time growth analysis. All three subpopulations (ATP-low 5%, bulk 5%, ATP-high 5%) are shown in the graph. From the results, it can be seen that after 120 hours, the growth of the ATP-high population is approximately 2-fold higher than that of the bulk cell population and approximately 5-fold higher than that of the ATP-low population. Data are shown as mean ± SD, n = 3. One-way ANOVA, Dunnett's multiple comparison test, **p < 0.001, ***p < 0.001, ****p < 0.0001. As can be seen from Figure 6A, the Cell Index of the ATP-high subpopulation was significantly higher compared to the other subpopulations, and the Cell Index of the ATP-low subpopulation was significantly lower compared to the other subpopulations. Figure 6B shows the time-course slope of the Cell Index for each subpopulation. At each time point (24, 48, 72, 96, and 120 hours), the slope of the ATP-high cell population was significantly higher compared to the other two subpopulations. Data are shown as mean ± SD, n = 3. Two-way ANOVA, corrected by Tukey's method, *p < 0.01. It is clear that the growth rate of the ATP-high subpopulation was the highest throughout the 120-hour assessment. From the results, it can be seen that the growth of the ATP-high population is at least approximately 2-fold higher than that of the bulk cell population and at least approximately 5-fold higher than that of the ATP-low population. This indicates that the mitochondrial ATP level is an important determinant of MCF7 cell growth, and that the metabolic fractionation by the fluorescent ATP imaging probe of this method is an effective technique for identifying the most proliferative cell population and the least proliferative cell subpopulation.

[0061] In a further assay, it was confirmed that the ATP-high MCF7 subpopulation had hyper-proliferated energetically, with significantly increased three-dimensional matrix-independent growth, cancer stem cell markers, and mitochondrial mass. To confirm the selectivity of ATP-Red1, after flow cytometry, the ATP levels of the cells were measured using Cell-Titer-Glo. However, since Cell-Titer-Glo is a luciferase-based assay, cell lysis is required to detect ATP levels, and thus it cannot be used for live cell sorting or imaging. After counting the cells, the same number of single cells were then used to evaluate the relative ATP content by luminescence using a Varioskan (trademark) LUX plate reader. The bar graph in Figure 7A shows that the cells of the ATP-high MCF7 subpopulation showed at least a 15-fold increase in ATP levels compared to the ATP-low cell population, while the bulk cells showed an approximately 7-fold increase. It can also be seen from this graph that the ATP levels of the ATP-high subpopulation in the MCF7 population are at least twice that of the bulk cells.

[0062] Using a three-dimensional mammosphere assay, matrix-independent growth, a functional readout of CSC activity and CSC proliferation, was measured. Figure 7B shows the results of the three-dimensional mammosphere assay of the ATP-high, bulk, and ATP-low subpopulations using 5% as the gating cut-off. Figure 7C shows a comparative image of the cell subpopulations after the assay. Images of the three-dimensional mammospheres were acquired using an EVOS FL Auto2 microscope. These are representative images of the cells of the three sorted cell populations respectively. A 4x objective lens was used. The scale bar is 1,000 μm. The ATP-high MCF7 cell subpopulation showed a 9-fold increase in three-dimensional spheroid formation compared to the ATP-low subpopulation, and almost twice the mammosphere formation of the bulk subpopulation. These data suggest that ATP-high cells have a higher ability to perform three-dimensional matrix-independent growth than the bulk CSC population.

[0063] Using the two established CSC markers, CD44 and ALDH activity, the "stemness" of the subpopulation was examined. Figure 7D shows that when the FACS gating cutoff value was set at 5%, the ATP-high MCF7 cell subpopulation (right bar) had approximately 4-fold higher cell surface expression of CD44 and approximately 5.5-fold higher ALDH activity compared to the ATP-low subpopulation (left bar). Similar results were obtained with MitoTracker-Deep-Red, an established marker for mitochondrial mass, with a 3-fold increase observed in the ATP-high MCF7 subpopulation compared to the ATP-low subpopulation. Mitochondrial mass is a specific marker of stemness in CSCs.

[0064] These indicate that the CSC activity of the ATP-high subpopulation is enhanced by the metabolic fractionation of this approach. Importantly, while high ALDH activity is considered a biomarker for EMT (epithelial-mesenchymal transition) of CSCs, CD44 is rather considered an epithelial CSC marker. That is, both epithelial and mesenchymal CSCs are significantly increased in the ATP-high cell subpopulation.

[0065] Fluorescent bioprobes indicating antioxidant capacity and multipotency also select the ATP-high MCF7 cell population. The effectiveness of the BioTracker ATP-Red1 imaging probe was compared with several other fluorescent biostaining dyes, particularly with respect to the selectivity for the ATP-high cell population. For this purpose, a two-dimensional monolayer of MCF7 cells was recovered with trypsin and live stained with a panel of five other fluorescent BioTracker probes for i) antioxidant capacity such as cysteine uptake ("cysteine FITC") and γ-glutamyltranspeptidase activity (GGT); ii) pluripotent stem cells; iii) hypoxia; and iv) senescence (β-galactosidase activity, β-Gal). Then, immediately after flow cytometry, the total ATP level was measured using Cell-Titer-Glo.

[0066] Figure 7E shows the results of Cell-Titer-Glo for this analysis, representing the fold change in luminescence of the upper 5% (bars to the right of each probe) relative to the lower 5% (bars to the left of each probe). Surprisingly, all probes showing antioxidant capacity (cysteine uptake and GGT activity) and multipotency selected the ATP-high subpopulation of MCF7 cells. However, among the additional fluorescent bioprobes tested, the BioTracker probe that directly measures cysteine FITC uptake was the most effective in selecting the ATP-high cell subpopulation, but not as effective as ATP-Red1 (3-fold vs 20-fold). Interestingly, high antioxidant capacity is known to be tightly associated with stemness and drug-resistant phenotypes. The hypoxia probe also actively selected the ATP-high cell subpopulation. This could be due to the association between hypoxia and increased mitochondrial biogenesis. However, the senescence probe (β-galactosidase activity) did not select either the ATP-high or ATP-low cell population.

[0067] Figures 8A and 8B show the results regarding the metabolic profiling of 3D mammospheres and ATP-high MCF7 cells. To deepen the understanding of the metabolism underlying 3D scaffold-independent growth, the intracellular ATP levels of MCF7 cells cultured in 2D monolayers or 3D spheroids were compared. The latter cell population is known to be highly enriched in CSCs. The levels of metabolites in MCF7 cells grown as 2D adherent monolayers or 3D mammospheres were compared using kits from Promega (Cell-Titer-Glo, GSH / GSSG-Glo, NADP-NADPH-Glo, NAD-NADH-Glo). The 2D monolayers and 3D mammospheres were first dissociated into single cells with trypsin, injected with a 25-gauge needle, and filtered through a 40-μm cell strainer. After counting the cells, the relative luminescence of these was then evaluated using the same number of single cells. Note that the cells obtained from 3D mammospheres all showed an increase in ATP level more than 2-fold, reduced glutathione level nearly 2-fold, NADP-NADPH level more than 2-fold, and NAD-NADH level nearly 1.5-fold compared to the 2D monolayers. The data represent the mean growth doubling rate ± SD, n = 4 for adherent cells. Unpaired t-test, **p < 0.005, ***p < 0.0005, ****p < 0.0001.

[0068] Figure 8A compares the changes in luminescence of 3D spheroids (right bar) and 2D monolayers (adherent, left bar) for probes targeting ATP, GSH / GSSG, NADP-NADPH, and NAD-NADH. In the quantitative analysis of MCF7 cells obtained from 3D mammospheres, the ATP level showed a 2.3-fold increase compared to 2D monolayer cells. An approximately 2-fold increase was observed in both the GSH / GSSG ratio and NADP / H level, and similar results were obtained for NAD / H. These data are consistent with the idea that the improvement of antioxidant capacity may also be required for 3D scaffold-independent growth.

[0069] From the above, it is expected that the ATP-high subpopulation of two-dimensional monolayer cells has the ability to undergo three-dimensional scaffold-independent growth. Under low adhesion conditions, usually more than 90% of MCF7 cells undergo anoikis, a special form of apoptotic cell death. If the ATP level is high, it is speculated that CSCs are more resistant to the great stress of growth in suspension caused by the absence of cell-matrix adhesion. However, if the energy storage amount is large, resistance to multiple stress factors is also conferred, which may result in multidrug resistance.

[0070] ATP-high and ATP-low MCF7 cells were metabolically profiled for NAD / H and two important antioxidants, GSH and NADP / H, using Promega kits (Cell-Titer-Glo, GSH / GSSG-Glo, NADP-NADPH-Glo, NAD-NADH-Glo). First, two-dimensional monolayer cells were stained with BioTracker ATP-Red1 and sorted by flow cytometry according to ATP content. After counting the cells, equal numbers of single cells were then used to evaluate their relative luminescence. Notably, all ATP-high cells showed an approximately 25-fold increase in ATP level, a 6-fold increase in reduced glutathione level, an approximately 8-fold increase in NADP-NADPH level, and more than a 2-fold increase in NAD-NADH level compared to ATP-low MCF7 cells. Data are presented as mean fold increase ± SD relative to ATP-low 5% cells, n = 4. Unpaired t-test, **p < 0.005, ***p < 0.0005. As shown in Figure 8B, all cells in the ATP-high subpopulation contained more than 1.5-fold NAD / H, more than 7.5-fold NADP / H, and more than 7-fold reduced glutathione (GSH / GSSG ratio) compared to cells in the ATP-low subpopulation. These data indicate that the ATP-high subpopulation of MCF7 cells has higher energy and, as a result, enhanced antioxidant capacity. High levels of antioxidants are known to be associated with drug resistance in cancer cells and exhibit a multidrug-resistant phenotype. Therefore, the cells in the ATP-high MCF7 subpopulation mimic a three-dimensional metabolic phenotype, demonstrating that the present method of separating ATP-high cells from the population generates a unique phenotype with many potential applications.

[0071] The phenotypes of the ATP-high and ATP-low subpopulations exist in many cancer tumors. All of the ATP-high subpopulations of MCF7, T47D, MDA-MB-231, and MDA-MB-468 cells show an increase in three-dimensional matrix-independent growth. From three other human breast cancer cell lines, T47D, MDA-MB-231, and MDA-MB-468 cells, compositions of the ATP-high and ATP-low cell subpopulations were created with a FACS gating cutoff value of 10%. The relative amounts of ATP in the ATP-high and ATP-low cell subpopulations were separately verified using Cell-Titer-Glo. First, two-dimensional monolayer cells were stained with BioTracker ATP-Red1 and sorted by ATP content using a flow cytometer. After counting the cells, the relative luminescence amounts of these were then evaluated using the same number of single cells. In this series of experiments, the ATP-high and ATP-low cell populations were defined with a cutoff value of 10%. Note that this metabolic fractionation method can be successfully applied to other breast cancer cell lines. The data represent the mean doubling rate ± SD, n = 3 for the ATP-low 10% cells. Unpaired t-test, *p < 0.05, **p < 0.005, ***p < 0.0005.

[0072] Figures 9A and 9B show the Cell-Titer-Glo and three-dimensional mammosphere formation of the ATP-high and ATP-low subpopulations of MCF7, T47D, MDA-MB-231, and MDA-MB-468 cells using a 10% gate. Figure 9A shows that the ATP-high subpopulations of all of these cell lines had a 2- to 3-fold increase in total ATP levels, as confirmed by the Cell-Titer-Glo assay using luciferase, indicating an increase in ATP characteristic of the ATP-high subpopulation phenotype. As can be seen from Figure 9B, similar results were obtained in the three-dimensional spheroid assay, showing a 1.75- to 3-fold increase in CSC activity and proliferation depending on the cell lines examined. First, two-dimensional monolayer cells were stained with BioTracker ATP-Red1 and sorted by ATP content using a flow cytometer. After counting the cells, 5×10 3Individual cells were seeded into 6-well plates coated with poly-HEMA and counted after 5 days. Note that all of the ATP-high cell populations of MCF7, T47D, MDA-MB-231, and MDA-MB-468 cells showed enhanced three-dimensional matrix-independent growth ability. Data represent mean doubling rate ± SD relative to ATP-low 10% cells, n = 3. Unpaired t-test, ***p < 0.0005, ****p < 0.0001.

[0073] Cells of the ATP-high subpopulation show increased oxidative mitochondrial metabolism, glycolytic rate, and cell cycle progression. Figure 10A shows luminescence of the ATP-high and ATP-low subpopulations (10%) in the MCF7 cell population after 24 hours. After counting the cells, the same number of single cells were then used to evaluate the relative ATP content by luminescence using a Varioskan™ LUX plate reader 24 hours after plating. In this experiment, because a large number of cells were required, a cut-off value of 10% was used to define the ATP-high and ATP-low cell populations. Data represent mean doubling rate ± SD relative to ATP-low 10% cells, n = 3. Unpaired t-test, ****p < 0.0001. The observed increase in ATP levels decreased to one-third 24 hours after plating the ATP-high cells as a two-dimensional monolayer, indicating that the energy-rich ATP-high phenotype is relatively transient and consistent with a more stem-like phenotype.

[0074] Figures 10B to 10E show the results of metabolic flux analysis of the ATP-high and ATP-low subpopulations. Oxygen consumption rate (OCR) was determined by metabolic flux analysis using Seahorse XFe96. Note that the ATP-high MCF7 cell population shows an increase in mitochondrial ATP production in addition to basal respiration and maximal respiration. The cell population was analyzed 24 hours after plating. Data are shown as fold increase (%) ± SD relative to ATP-low 10% cells, n = 3. Unpaired t-test, *p < 0.05, **p < 0.005. Extracellular acidification rate (ECAR) was determined by metabolic flux analysis using Seahorse XFe96. Note that the ATP-high MCF7 cell population shows an increase in glycolysis. The cell population was analyzed 24 hours after plating. Data are shown as mean fold increase (%) ± SD relative to ATP-low 10% cells, n = 3. Unpaired t-test, ns = no significant difference, ***p < 0.0005. The energy profiles in Figures 10B and 10C indicate that the ATP-high subpopulation is metabolically active relative to the ATP-low population. 24 hours after cell adhesion, the ATP-high MCF7 monolayer cells showed a 2-fold increase in basal respiration, a 1.5-fold increase in maximal respiration, and a 3-fold increase in ATP production. Similarly, the ATP-high MCF7 monolayer cells also showed a 1.5-fold increase in basal glycolysis rate. The glycolysis rates in Figures 10D and 10E indicate that the bioenergy of the ATP-high subpopulation is significantly higher than that of the ATP-low subpopulation.

[0075] From the evaluation of the proliferative ability of the ATP-high cell subpopulation, it was revealed that the ATP-high cell subpopulation was significantly more proliferative than the ATP-low subpopulation in various cancer types. Figures 11A - 11D show the cell cycle progression of MCF7, T47D, MDA-MB-468, and MDA-MB-231 cells by FACS analysis using propidium iodide to detect DNA content. As can be seen, the ATP-high cell subpopulation transitioned from the G0 / G1 phase to the S phase and the G2 / M phase, and was significantly more proliferative than the ATP-low subpopulation. More specifically, while the G0 / G1 phase decreased from approximately 80 - 88% to 60 - 64%, the S phase increased from 4 - 8% to 9 - 21%. Similarly, the G2 / M phase increased from 7 - 12% to 17 - 30%. In all four cell lines, such a clear promotion of cell cycle progression was observed in the ATP-high cell subpopulation compared to the ATP-low subpopulation. Overall, this showed an increase of 1.9 - 3.6 times in the number of cells in the S phase and 1.8 - 3.8 times in the number of cells in the G2 / M phase in all four cell lines tested. Interestingly, the largest increase in the S phase was observed in MCF7 cells, while the largest increase in the G2 / M phase was observed in MDA-MB-231 cells.

[0076] On the other hand, the ATP-low population of each cell line was basically in a quiescent state, with 80 - 88% of the cells in the G0 / G1 phase of the cell cycle, showing the main phenotype of cell cycle arrest. That is, the ATP-low cell subpopulation well matches the current definition of the dormant state of cancer cells.

[0077] Therefore, high ATP levels are the main determinants of the "stemness" trait, anchorage-independent growth, and cell proliferation. Thus, by the practical methods described herein, it is possible to successfully isolate the most bioenergetically "optimal" and highly proliferative cancer cells from the whole cell population, forming a new composition of cells with unique phenotypic characteristics. These characteristics are closely related to drug resistance.

[0078] The ATP-high subpopulation of MCF7 cells exhibits a multidrug-resistant phenotype. A three-dimensional mammosphere assay was used as a functional readout of drug resistance to explore the sensitivity differences between the ATP-high and ATP-low MCF7 cell subpopulations to four drug classes. The drug classes included tamoxifen, doxorubicin, DPI, and palbociclib. Figure 12A shows the results for the ATP-low subpopulation (bottom 5%), and Figure 12B shows the results for the ATP-high subpopulation (top 5%). Figures 12A and 12B show two concentrations for each drug class.

[0079] Tamoxifen is an FDA-approved drug commonly used clinically to target ER(+) breast cancer cells. As a result of tamoxifen resistance and treatment failure, tumor recurrence and distant metastasis often occur. Interestingly, three-dimensional mammosphere formation by ATP-low MCF7 cells was highly sensitive to tamoxifen treatment, with a decrease of approximately 40% at 1 μM and over 90% at 5 μM. In contrast, from Figure 12B, it can be seen that three-dimensional mammosphere formation by ATP-high MCF7 cells remained above 80% of the solvent-treated control at 5 μM, indicating significant resistance to tamoxifen. Therefore, ATP-high MCF7 cells are clearly tamoxifen-resistant.

[0080] From Figure 12B, it can be seen that ATP-high MCF7 cells were also resistant to the mitochondrial OXPHOS inhibitor, diphenyleneiodonium (DPI). For example, in the DPI treatment of ATP-low cells, three-dimensional mammosphere formation decreased by over 90% at 100 nM. In contrast, in the DPI treatment (100 nM) of ATP-high cells, the decrease in three-dimensional mammosphere formation was only about 55%. Therefore, although both subpopulations were sensitive to the mitochondrial inhibitor, ATP-high cells were clearly more resistant.

[0081] Doxycycline is an FDA-approved antibiotic that inhibits mitochondrial ribosomal translation. Comparing FIGS. 12A and 12B, it can be seen that at concentrations that were very effective in ATP-low MCF7 cells, namely 25 μM and 50 μM, the ATP-high subpopulation showed significant resistance to doxycycline.

[0082] The effectiveness of palbociclib, an FDA-approved CDK4 / 6 inhibitor, is also evident from FIGS. 12A and 12B. Treatment of ATP-low cells with palbociclib reduced three-dimensional mammosphere formation by approximately 75% at 12.5 nM. However, in the treatment of ATP-high cells with palbociclib (12.5 nM), the reduction in three-dimensional spheroid formation was only about 50%. Thus, ATP-high cells were also highly resistant to the CDK4 / 6 inhibitor. Therefore, the ATP-high subpopulation is a phenotype that shows resistance to several drug classes.

[0083] CD44 and ALDH activity, established markers of stemness, were compared with BioTracker ATP-Red1. To directly compare the effectiveness of ATP-Red1 with other CSC markers, a dual-labeling method was applied to both MCF7 and MDA-MB-231 cells. Cells were dual-labeled for CD44 and ATP using different fluorescent channels for detection. This resulted in four experimental groups for CD44 and ATP, CD44-high / ATP-high, CD44-high / ATP-low, CD44-low / ATP-high, and CD44-low / ATP-low.

[0084] After cell sorting, a three-dimensional mammosphere assay was performed on the four resulting subpopulations as a functional readout of stemness. The cell surface expression of ATP and CD44 was compared. After cell sorting, both the MCF7 and MDA-MB-231 cell lines were used, and three-dimensional anchorage-independent growth was measured in different cell subpopulations as a functional readout of stemness. Briefly, first, the two-dimensional monolayer was co-stained with both BioTracker ATP (PE channel) and anti-CD44 (APC channel), and flow cytometry was performed using a Sony SH800 cell sorter. After counting the cells, 5×10 3 cells were seeded into a 6-well plate coated with poly-HEMA, and three-dimensional mammospheres were counted 5 days after plating. As shown in Figures 13A and 13B, CD44-low / ATP-low cells had the least anchorage-independent growth, as expected from the phenotypic characteristics of these subpopulations. Therefore, CD44-low / ATP-low cells were selected as the normalization point. Two cell subpopulations, CD44-high / ATP-high and CD44-low / ATP-high, showed the greatest anchorage-independent growth. Thus, a high level of ATP is a major determinant of stemness in both MCF7 and MDA-MB-231 cells compared to CD44.

[0085] We decided to examine only the CD44-high population and double-labeled the CD44-high population with ATP so that it could be classified into two subpopulations: a subpopulation with high proliferative capacity (CD44-high / ATP-high) and a subpopulation with low proliferative capacity (CD44-high / ATP-low). This showed that the CD44-high / ATP-low population clearly had significantly less anchorage-independent growth, making it a subpopulation with a high "quiescence" of CD44(+) CSCs.

[0086] Almost the same results were obtained with double labeling using ALDH activity and ATP. After cell sorting, both MCF7 and MDA-MB-231 cell lines were used, and three-dimensional anchorage-independent growth was measured in different cell subpopulations as a functional readout of stemness. Briefly, a two-dimensional monolayer was first co-stained with both BioTracker ATP (PE channel) and ALDH activity (APC channel), and flow cytometry was performed using a SONY SH800 cell sorter. After counting the cells, 5×10 3 cells were seeded into a 6-well plate coated with poly-HEMA, and three-dimensional mammospheres were counted 5 days after plating. Figures 13C and 13D show the results of the mammosphere formation assay of MCF7 and MDA-MB-231 cell lines double-labeled with ADLH activity and ATP. As expected, the two cell populations showing the greatest anchorage-independent growth were ALDH-high / ATP-high and ALDH-low / ATP-high. Thus, high levels of ATP are a major determinant of stemness in both MCF7 and MDA-MB-231 cells compared to ALDH. Similarly, double-labeling with ATP allows the ALDH-high population to be classified into two subpopulations: a subpopulation with high proliferative capacity (ALDH-high / ATP-high) and a subpopulation with low proliferative capacity (ALDH-high / ATP-low).

[0087] The above demonstrates that this method is a powerful and effective method for using ATP as a second marker for the dormant state and sub-fractionating CSCs into highly active hyperproliferative subpopulations and highly dormant subpopulations. This result also indicates that ATP levels are functional regulators of the dormant state of CSCs.

[0088] The role of mitochondrial ATP in cell migration, invasion, and spontaneous metastasis was also explored. The data demonstrate that mitochondrial ATP is a biomarker of the energy involved in the metastatic process of cancer cells. MDA-MB-231 cells are an established model for studying cell motility and metastasis both in vitro and in vivo. The ability of the ATP-high and ATP-low subpopulations of MDA-MB-231 cells to perform cell migration and invasion was evaluated by a modified Boyden chamber assay using Transwell. The bulk (5%) population was also selected for comparison. To examine invasion, the Transwell was coated with extracellular matrix, i.e., Matrigel, to prevent simple cell migration. Serum was used as a chemoattractant in both the cell migration and invasion assays. The parameters of migration and invasion were quantified separately using both crystal violet staining intensity and cell number.

[0089] Figures 14A and 14B show the results of this migration and invasion analysis. The ATP-high MDA-MB-231 cells showed a 20- to 40-fold increase in their ability to cause cell migration compared to the ATP-low cells. As expected, the bulk (5%) cells showed an intermediate phenotype. The ATP-high MDA-MB-231 cells showed a 15- to 25-fold increase in their ability to cause invasion compared to the ATP-low cell population. Thus, the ATP-high MDA-MB-231 cells correspond to the pre-metastatic cell subpopulation in vivo.

[0090] For further evaluation, an established in vivo metastasis assay using the chorioallantoic membrane (CAM) of chicken eggs was used to quantitatively measure spontaneous metastasis. After isolating the ATP-high and ATP-low cell subpopulations by cell sorting, 30,000 cells (MDA-MB-231) were inoculated onto the CAM of each egg (day E9), and then randomly divided into several groups. On day E18, the lower CAM was harvested and analyzed by qPCR using specific primers for the human Alu sequence to evaluate the number of metastatic cells. Uninjected eggs were also evaluated in parallel as a negative control for specificity. More than 20 eggs were processed for each experimental condition.

[0091] Figure 15 shows the results of spontaneous metastasis in an in vivo CAM assay. From the data, it can be seen that the ATP-high subpopulation of MDA-MB-231 cells had 4.5-fold higher metastatic ability than the ATP-low cell subpopulation. These subpopulations were derived from the same cell line. Therefore, the ATP-high subpopulation of MDA-MB-231 cells corresponds to the pre-metastatic CSC subpopulation. As described above in relation to Figure 4C, the ATP-high subpopulation of MDA-MB-231 cells also overexpresses two CTCs and metastasis markers (VCAM-1 and Ep-CAM), indicating that the over-proliferating CSCs are the CTCs involved in seeding distant metastases.

[0092] Therefore, using this method, the possibility of cancer metastasis can be detected. For example, a biological sample of cancer can be metabolically fractionated to evaluate the content of the ATP-high subpopulation, and the content can be used to estimate the possibility of cancer metastasis. By early detecting and analyzing the cells of the ATP-high subpopulation in cancer patients, a very valuable opportunity is obtained to diagnose the risk of metastasis and identify appropriate treatments such as the ATP depletion therapeutic agents described herein.

[0093] The Tempo-ATP protein biosensor that purifies ATP-high MCF7 cells independently validates the use of ATP as a new biomarker of stemness in cancer cells. Tempo-ATP, a fluorescent protein biosensor, is a completely different probe for detecting ATP levels in living cells, and this was used to detect the high and low levels of ATP.

[0094] Tempo-ATP-MCF7 cells that recombinantly overexpress a cytoplasmic fluorescent protein ATP biosensor were specially produced by Tempo-Bioscience, Inc. (San Francisco, CA, USA) using a puromycin resistance marker for cell selection. This protein-based fluorescent ATP biosensor has an excitation wavelength of 517 - 519 nm and an emission of 535 nm. It consists of an ATP-binding peptide fused in-frame with a GFP-like fluorescent reporter protein. Tempo-ATP-MCF7 cells were classified by GFP content as a surrogate marker for cytoplasmic ATP content using a flow cytometer (excitation = 517 - 519 nm; emission = 535 nm). After counting the cells, the same number of single cells were then used to evaluate metabolic and phenotypic behavior.

[0095] The results are shown in FIGS. 16A - 16C. The relative increase in emission of the GFP-high subpopulation relative to the GFP-low subpopulation is shown in FIG. 16A. Data on cell cycle progression are shown in FIG. 16B, and the results of the mammosphere formation assay are shown in FIG. 16C. As expected from previous studies, ATP-high Tempo-MCF7 cells had a significant increase in ATP, a greater decrease in reduced glutathione, NADP / H, and NAD / H, and also showed increased cell cycle progression and three-dimensional matrix-independent growth. The Tempo-ATP data independently validate the results of BioTracker ATP-Red1, particularly that high ATP levels are important determinants of antioxidant capacity, cell proliferation, and three-dimensional matrix-independent growth. Although Tempo-ATP was also effective, BioTracker ATP-Red1 was significantly more effective because it localizes directly within mitochondria, the major ATP source in cells.

[0096] This approach has demonstrated that high levels of ATP production are a major driver of the "stemness" phenotype and proliferation in cancer cells. The observations disclosed herein can explain the molecular basis of metabolic heterogeneity observed in cancer cell populations and its relationship to the phenotypic behaviors required for in vivo metastatic seeding of CSCs, such as i) rapid cell cycle progression and ii) anchorage-independent growth.

[0097] As demonstrated, ATP can be used as a biomarker to metabolically fractionate cancer cell populations and identify subpopulations of overproliferating and quiescent cells. This indicates that ATP depletion therapy may be effective in treating overproliferating subpopulations and may reduce or eliminate the potential for tumor recurrence and metastasis.

[0098] Based on this approach, important fluorescent dyes that can measure the ATP levels of living cells, such as BioTracker ATP-Red1, can be used as imaging probes for metabolic fractionation. More specifically, BioTracker ATP-Red1 staining can be combined with a bioenergetic fractionation method that subjects the entire cell population to flow cytometry to isolate the ATP-high and ATP-low subpopulations of the population. In many of the examples considered above, the MCF7 cell line, an ER(+) human breast cancer cell line, has been used. Of course, this approach can be used with any cell line and any cancer type. The metabolic fractionation technique enables the isolation of the most "energy-rich" cancer cells within the entire cell population. Advantageously, the resulting ATP-high cancer cell subpopulation can be targeted for eradication by ATP depletion therapy and used as a basis for drug discovery and development. Considering the phenotypic characteristics, the ATP-high subpopulation can also be used to evaluate therapies that prevent or reduce the potential for recurrence and metastasis.

[0099] In a series of parallel studies, the inventors have identified over 20 mitochondrial-targeted therapeutic agents that may be useful for effectively implementing ATP depletion therapy. These therapeutic agent candidates include the following: FDA-approved drugs (doxycycline, tigecycline, azithromycin, pyrvinium pamoate, atovaquone, bedaquiline, niclosamide, irinotecan); natural products / functional foods (actinonin, CAPE, berberine, bruceellidine, meridianin); and experimental compounds (oligomycin, AR-C155858, mitribosocin (see International Patent Application PCT / US2018 / 022403, filed on March 14, 2018 (Patent Document 1). Also incorporated by reference in its entirety), mitoketosin (see International Patent Application PCT / US2018 / 039354, filed on June 25, 2018 (Patent Document 2). Also incorporated by reference in its entirety), mitoflavosin (see International Patent Application PCT / US2018 / 057093, filed on October 23, 2018 (Patent Document 3). Also incorporated by reference in its entirety), TPP derivatives (including dodecyl-TPP and 2-butene-1,4-bis-TPP. See International Patent Application PCT / US2018 / 062174, filed on November 21, 2018 (Patent Document 4). Also incorporated by reference in its entirety)). Three combinations of two antibiotics and vitamin C (doxycycline, azithromycin, and ascorbic acid) have been found to be particularly potent in targeting mitochondria, inducing ATP depletion and CSC proliferation at lower antibacterial levels (see International Patent Application PCT / US2019 / 066541, filed on December 16, 2019 (Patent Document 5). Also incorporated by reference in its entirety)).An ATP-depleting compound can be an existing compound that has been modified to increase efficacy, cell membrane permeability, and / or mitochondrial uptake, for which see, for example, International Patent Application PCT / US2018 / 033466, filed May 18, 2018 (Patent Document 6) (incorporated by reference in its entirety) and International Patent Application PCT / US2018 / 062956, filed November 29, 2018 (Patent Document 7) (incorporated by reference in its entirety). For example, doxycycline conjugated with a fatty acid such as myristate can be used as an ATP-depleting compound. Depending on the situation, such as when the ATP-high subpopulation shows resistance to the compound at the doses normally prescribed in the art, it may be appropriate to increase the dose of the compound for administration. The compound can be administered. Of course, any of the aforementioned compounds can be used as an ATP-depleting therapeutic agent to target the ATP-high subpopulation and prevent or reduce the likelihood of recurrence and metastasis. Of course, when it is recognized that the expression levels of the gene signatures of five members, ABCA2, ATP5F1C, COX20, NDUFA2, and UQCRB, in a patient's cancer biological sample are elevated relative to the expression levels in the patient's non-cancerous biological sample, any of the aforementioned compounds can be used as a therapeutic agent for administration to cancer patients.

[0100] Many compounds require Phase II clinical trials because they use FDA-approved antibiotics with excellent safety profiles for other purposes. For example, in the Phase II clinical preliminary trial of doxycycline, as demonstrated by using CD44 and ALDH1 as specific CSC markers, this antibiotic that has been used for over 50 years has already been shown to be actually effective in metabolically targeting the CSC population of early breast cancer patients. Mitochondrial ATP depletion therapy is expected to more effectively starve CSCs by functionally mimicking fasting and / or calorie restriction. In this method, fasting and / or calorie restriction can be implemented as part of cancer treatment to enhance the effect of ATP depletion therapy. For example, patients receiving ATP depletion therapy can fast for a period of time such as 12, 16, 24, 36, or 48 hours before receiving the therapeutic compound, and / or can fast for a period of time such as 12, 16, 24, 36, or 48 hours after receiving the therapeutic compound. In some embodiments, fasting can be performed before and after administration of the therapeutic compound to enhance the ATP depletion effect. This has important implications for cancer prevention and the possibility of extending human lifespan during aging.

[0101] Cells in the ATP-high subpopulation exhibit a multi-drug resistant phenotype with enhanced antioxidant capacity. Previous studies have shown that high antioxidant capacity greatly contributes to the expression of multi-drug resistance through an increase in the level of reduced glutathione, an increase in NADPH, and activation of the NRF2 signal. MCF7 cells in the ATP-high subpopulation have high antioxidant capacity with an increased level of reduced glutathione and are essentially resistant to four drugs (tamoxifen, palbociclib, doxycycline, and DPI). Therefore, the existence of the CSC phenotype of ATP-high can help mechanistically explain the cause of multi-drug resistance during cancer treatment. In this context, in current cancer therapies, only metabolically "optimal" cancer cells may survive. Then, these cells pose the greatest risk of recurrence and metastasis.

[0102] The data disclosed above also shows a direct causal relationship between mitochondrial "power" and tamoxifen resistance. For example, MCF7-TAMR cells generated by long-term exposure while increasing the concentration of tamoxifen and acquiring tamoxifen resistance showed an increase in the levels of mitochondrial OXPHOS and ATP production. In MCF7-TAMR cells, it was revealed from unbiased proteome analysis that the acquisition of tamoxifen resistance was due to the overexpression of two major antioxidant proteins (NQO1 and GCLC) and their favorable metabolic effects on mitochondrial metabolism. Also, when NQO1 or GCLC was recombinantly overexpressed in MCF7 cells, mitochondrial ATP production and tamoxifen resistance increased autonomously by about twofold. Furthermore, recombinant overexpression of a somatic mutation (Y537S) in the estrogen receptor (ERα; ESR1), which is clinically relevant to acquired tamoxifen resistance in breast cancer patients, genetically resulted in increased mitochondrial biogenesis, OXPHOS, and high ATP production. The proteome profiles of MCF7-TAMR cells and MCF7-ESR1(Y537S) cells also showed considerable overlap in functionally activated biological processes. Also, 60 gene products functionally related to mitochondrial ATP production were predictive of tamoxifen resistance in ER(+) / luminal A-type breast cancer patients. These predictive biomarkers included 18 mitochondrial ribosomal proteins (MRP) and more than 20 components of the mitochondrial OXPHOS complex. The data disclosed herein shows that "naive" MCF7 cells of the ATP-high subpopulation are essentially resistant to tamoxifen even without prior exposure to tamoxifen. This has clinically important implications in optimizing the effectiveness of hormonal therapy for breast cancer.

[0103] Treatment with conventional chemotherapy regimens has been reported to selectively kill "bulk" cancer cells while actually increasing the number of CSCs. Prior to the present disclosure, no metabolic hypothesis has been presented to explain this phenomenon. Chan et al. (Genentech, Inc.) examined the effects of gemcitabine and etoposide on the entire cancer cell population. Surprisingly, they observed that after treatment with gemcitabine and etoposide, in the population of surviving cells, mitochondrial respiration increased, and ATP content and mitochondrial mass showed an increase. However, they did not provide a mechanistic explanation for these observations and did not consider the CSC population either. Instead, they only concluded that measuring ATP would not provide useful information for evaluating the effectiveness of chemotherapeutic agents. Considering the data disclosed herein, their results can be alternatively interpreted as gemcitabine and etoposide selectively killing the ATP-low and bulk subpopulations of cancer cells, thus increasing the more stem-like and drug-resistant "energy-rich" ATP-high subpopulation. Therefore, it is necessary to initiate new drug discovery efforts to eradicate the ATP-high subpopulation of cancer cells.

[0104] In various colorectal and ovarian cancer cell lines (HT29, HCT116, A2780) treated for a long time, it has been suggested that high intracellular ATP levels are the main cause of acquired drug resistance to oxaliplatin and cisplatin, although various mechanisms have been proposed, such as increased glycolysis and / or mitochondrial metabolism. However, in previous studies, ATP levels were only measured after selecting the drug-resistant cell population over a long period. Therefore, a direct causal relationship between ATP production and drug resistance could not be established.

[0105] Previously, the inventors used a more indirect method of detecting intracellular FAD, FMN, and riboflavin content using autofluorescence to isolate "energy-rich" cancer stem cells (e-CSCS). See International Patent Application PCT / US2019 / 037860, filed June 19, 2019 (Patent Document 8). Also, the entire application is incorporated herein by reference. However, the use of ATP-Red1 is a direct method and has been significantly improved. For example, fractionating a two-dimensional monolayer of MCF7 using high autofluorescence (AF; top 5%) yielded a cell AF-high population with a 1.5-fold increase in anchorage-independent growth and almost a 2-fold increase in ATP production. In contrast, using ATP-Red1 (top 5%) in this method yielded a subpopulation of ATP-high MCF7 monolayer cells with a 9-fold increase in anchorage-independent growth and more than a 15-fold increase in ATP content. The ATP-high subpopulations of other cancer cell lines (T47D, MDA-MB-231, and MDA-MB-468) also showed similar overgrowth characteristics. Therefore, using ATP as a direct energy biomarker is far superior to autofluorescence. Furthermore, ATP-Red1 was also effective in metabolically fractionating the other three breast cancer cell lines tested.

[0106] According to the conventional view of tumor dormancy, dormant cancer cells avoid cell death by treatment and develop multidrug resistance by slowing the rate of cell proliferation and / or arresting the cell cycle (quiescence). The data disclosed herein show the opposite. Using a 3D mammosphere assay as the readout, more than 87% of the MCF7 cells in the ATP-low subpopulation were in the G0 / G1 phase of the cell cycle and had low proliferative capacity, but were highly sensitive to four drugs. Conversely, more than 38% of the MCF7 cells in the ATP-high subpopulation were in the S or G2 / M phase, had significantly higher proliferative capacity, and exhibited an obvious multidrug-resistant phenotype. Therefore, high levels of mitochondrial ATP can be said to be an "optimal" population of cancer cells in terms of energy and an important driver of both cell proliferation and drug resistance.

[0107] The inventors have shown, using an in vivo xenograft animal model, that treatment with a series of different anti-mitochondrial therapies is sufficient to i) metabolically induce ATP depletion and ii) potently inhibit cancer cell metastasis. These results indicate that high ATP levels are important in the metastatic process of CSCs, and are consistent with the data disclosed herein showing that ATP-high CSCs are hyperproliferative, stem-like, anchorage-independent, and have increased antioxidant capacity and intrinsic multidrug resistance. Thus, ATP-high CSCs that can be isolated using this method are thought to be involved in tumor recurrence and metastasis in vivo.

[0108] The above bioinformatics analysis has shown that ATP-related genes are closely associated with stemness, proliferation, and metastasis, and in particular, ATP5F1C has been shown to encode the γ subunit of the catalytic core of mitochondrial ATP synthase. Furthermore, ATP5F1C is a prognostic biomarker for tumor recurrence and distant metastasis in ER(+) patients receiving tamoxifen treatment, and is also a marker of treatment failure. Also, ATP-high MDA-MB-231 cells showed a dramatic improvement in both cell migration and cell invasion in vitro and the ability to undergo spontaneous metastasis in vivo. Thus, mitochondrial ATP plays an important role in metastatic seeding. Therefore, mitochondrial ATP synthase inhibitors would be effective as therapeutic candidates for preventing metastasis, i.e., eradicating the CSC ATP-high subpopulation.

[0109] The pharmaceutical composition of this method contains the ATP depletion compound (identified above) in any pharmaceutically acceptable carrier. If a solution is desired, the carrier selected for the water-soluble compound or salt can be water. With respect to water solubility, organic vehicles such as glycerol, propylene glycol, polyethylene glycol, or mixtures thereof may be suitable. Further, without departing from the method, methods for improving water solubility can be used. In the latter example, the organic vehicle can contain a substantial amount of water. In any example, the solution can be sterilized by filtration through a 0.22 micron filter, by way of example, in a suitable manner known to those skilled in the art. After sterilization, the solution can be dispensed into a suitable container such as a depyrogenated glass vial. Dispensing is optionally carried out by aseptic methods. Then, a sterilized closure is attached to the vial, and if desired, the contents of the vial can be lyophilized. Embodiments containing a second inhibitor compound such as a glycolysis inhibitor or an OXPHOS inhibitor can co-administer the form of the second inhibitor available in the art. This method is not intended to be limited to a particular dosage form unless otherwise described.

[0110] In addition to the ATP depletion compound, the pharmaceutical formulation of this method can contain other additives known in the art. For example, in some embodiments, it can contain a pH adjuster such as an acid (e.g., hydrochloric acid), and a base or buffer (e.g., sodium acetate, sodium borate, sodium citrate, sodium gluconate, sodium lactate, and sodium phosphate). In some embodiments, it can contain antibacterial preservatives such as methylparaben, propylparaben, and benzyl alcohol. Antibacterial preservatives are often included when formulating the drug into vials designed for multiple administrations. The pharmaceutical formulations described herein can be lyophilized using techniques well known in the art.

[0111] In embodiments that include oral administration of an ATP depletion compound, the pharmaceutical composition can be administered in forms such as capsules, tablets, pills, powders, solutions, suspensions, etc. Tablets containing various excipients such as sodium citrate, calcium carbonate, and calcium phosphate can be used with various tablet disintegrating substances such as starch (e.g., potato starch or tapioca starch) and certain complex silicates, along with binders such as polyvinylpyrrolidone, sucrose, gelatin, and acacia. Further, for tableting, lubricants such as magnesium stearate, sodium lauryl sulfate, and talc can be included. Solid compositions of a similar type can be used as fillers for soft and hard filled gelatin capsules. Related materials include lactose and high molecular weight polyethylene glycol. When an aqueous suspension and / or elixir is desired for oral administration, the compounds of the present disclosure can be combined with diluents such as water, ethanol, propylene glycol, glycerin, and various combinations thereof, in addition to various sweeteners, flavorants, colorants, emulsifiers, and / or suspending agents. In embodiments having a carbocyanine compound and a second inhibitor compound, the second inhibitor compound can be administered in a form other than the form of the carbocyanine compound, without being limited thereto.

[0112] Additional embodiments provided herein include liposomal formulations of the ATP depletion compounds disclosed herein. Techniques for making liposomal suspensions are well known in the art. When the compound is a water-soluble salt, it can be incorporated into lipid vesicles using conventional liposome technology. In such cases, since the active compound is water-soluble, it can be substantially incorporated into the hydrophilic center or core of the liposome. The lipid layer used can be of any conventional composition and may or may not contain cholesterol. When the active compound of interest is water-insoluble, the salt can also be substantially incorporated into the hydrophobic lipid bilayer that forms the structure of the liposome using conventional liposome-forming techniques. In either example, the size of the liposomes produced can be reduced, such as by using standard sonication and homogenization techniques. Liposomal formulations containing the active compounds disclosed herein can be lyophilized to produce a lyophilizate, which can be reconstituted with a pharmaceutically acceptable carrier such as water to regenerate a liposomal suspension.

[0113] Regarding pharmaceutical compositions, the pharmaceutically effective amount of the ATP-depleting compounds herein is to be determined by medical practitioners and will vary depending on the patient's condition, physique, and age, as well as the route of delivery. In one non-limiting embodiment, a dosage of about 0.1 to about 200 mg / kg has a therapeutic effect, and the weight ratio is the weight of the ATP-depleting compound relative to the weight of the subject, including when using salts. In some embodiments, the dosage can be the amount of the compound necessary to obtain a serum concentration of the active compound between about 1 and 5, 10, 20, 30, or 40 μM. In some embodiments, a dosage of about 1 mg / kg to about 10 mg / kg can be employed for oral administration, and in some embodiments, a dosage of about 10 to about 50 mg / kg can be used. For intramuscular administration, a dosage of about 0.5 to 5 mg / kg can typically be employed. In some embodiments, for intravenous or oral administration, the dosage can be about 1 to about 50 μmol / kg, or optionally about 22 to about 33 μmol / kg of the compound. Oral dosage forms can contain any suitable amount of the active substance, for example, from 5 mg to 50, 100, 200, or 500 mg per tablet or other solid dosage form.

[0114] In the following paragraphs, the materials and methods used in connection with the data and embodiments described herein will be explained. Of course, those skilled in the art may use alternative materials and methods generally accepted in the art without departing from this approach.

[0115] Cell lines and reagents: ER(+)[MCF7 and T47D] and triple-negative [MDA-MB-231 and MDA-MB-468] human breast cancer cell lines were purchased from the American Type Culture Collection (ATCC). ATP-Red1 (BioTracker™ ATP-Red Live Cell Dye; also known as #SCT045) was purchased from Sigma-Aldrich, Inc.

[0116] The heatmap in Figure 1A was generated using the GSE36953 GEO DataSet previously registered in the NCBI database. Total RNA was prepared from the MDA-MB-231 cell line, a TNBC cell line, under three growth conditions: two-dimensional adherent culture, three-dimensional scaffold-independent growth, and in vivo tumor growth. The analysis was performed using the Affymetrix Human Genome U133 Plus 2.0 Array. The heatmap was created using QIAGEN OmicSoft Suite software. All ATP-related genes were transcriptionally upregulated relative to two-dimensional adherent culture under both three-dimensional culture conditions (scaffold-independent tumors and in vivo tumors).

[0117] Flow cytometry after live staining with ATP-Red1: Human breast cancer cell lines were first grown as two-dimensional monolayers or three-dimensional spheroids. The cells were then harvested, dissociated into single-cell suspensions, and analyzed or sorted by flow cytometry using a SONY SH800 cell sorter. Briefly, the ATP-high and ATP-low subpopulations of cells were isolated after live staining with the probe ATP-Red1. The ATP-high and ATP-low cell subpopulations were selected by gating within the ATP-Red1 signal. Unless otherwise stated, cells with the weakest fluorescence signal (bottom 5% or 10%) or the strongest fluorescence signal (top 5% or 10%) were collected as ATP-low and ATP-high, respectively. Cells outside the gate were discarded during sorting by gate setting. However, such settings are often necessary to ensure high purity during sorting. The data were analyzed using FlowJo 10.1 software.

[0118] ATP assay using Cell-Titer-Glo: Cell-Titer-Glo (#G7570) was obtained from Promega, Inc. and used according to the manufacturer's instructions to measure the ATP level in lysed cells. Cell-Titer-Glo is a luciferase-based assay system.

[0119] 3D matrix-independent growth assay: Single cell suspensions were prepared by enzymatic dissociation (Trypsin EDTA×1, Sigma Aldrich, catalog number T3924) and manual dissociation (25 gauge needle). 5000 cells were seeded into 6-well plates coated with hydroxyethyl methacrylate (poly-HEMA, Sigma, catalog number P3932) under non-adherent conditions in Mammosphere medium (DMEM-F12 / B27 / 20 ng / mL EGF / PenStrep). Cells were maintained by growing for 5 days in a humidified incubator at 37 °C, atmospheric pressure, 5% (v / v) carbon dioxide / air. After 5 days, 3D spheroids with a diameter greater than 50 μm were counted using a microscope equipped with a graticule eyepiece, and the percentage of cells that formed spheroids was calculated and normalized to 1 (1 = 100% MFE; Mammosphere formation efficiency). The Mammosphere assay was performed in triplicate and independently repeated 3 times.

[0120] Metabolic flux analysis: The extracellular acidification rate and oxygen consumption rate were analyzed using a Seahorse XFe96 analyzer (Agilent / Seahorse Bioscience, USA). Cells were maintained in DMEM supplemented with 10% FBS (fetal bovine serum), 2 mM GlutaMAX, and 1% Pen-Strep. 20,000 breast cancer cells per well were seeded in an XFe96 well cell culture plate and incubated at 37 °C in a 5% CO2 humidified atmosphere for at least 12 hours to allow cell attachment. After approximately 24 hours, MCF7 cells were washed with pre-warmed XF assay medium or XF assay medium supplemented with 10 mM glucose, 1 mM pyruvate, 2 mM L-glutamine for OCR measurement and adjusted to pH 7.4. The cells were then maintained in 175 μL / well of XF assay medium in a non-CO2 incubator at 37 °C for 1 hour. During incubation, 25 μL of 80 mM glucose, 9 μM oligomycin, and 1 M 2-deoxyglucose (for ECAR measurement) or 10 μM oligomycin, 9 μM FCCP, 10 μM rotenone, 10 μM antimycin A (for OCR measurement) were added to the inlets of the XFe96 sensor cartridge in XF assay medium. The measured values were normalized by protein content (SRB assay) and Hoechst33342 content. The dataset was analyzed by one-way ANOVA and Student's t-test using XFe96 software and GraphPad Prism software. All experiments were performed in quintuplicate and repeated independently three times.

[0121] Cell cycle analysis by FACS: Cell cycle analysis was performed on ATP-high and ATP-low cell subpopulations by FACS analysis using an Attune NxT Flow Cytometer. Briefly, after trypsin treatment, the resuspended cells were incubated with 10 ng / mL of Hoescht solution (Thermo Fisher Scientific) for 40 minutes at 37 °C in the dark. After 40 minutes, the cells were washed and resuspended in PBS (Ca, Mg) for acquisition or in sorting buffer [PBS×1 containing 3% (v / v) FBS and 2 mM EDTA] for FACS. 50,000 events were analyzed for each condition. The gated cells were manually classified into cell cycle stages.

[0122] Statistical significance: All analyses were performed using GraphPad Prism6. Data were represented as mean ± SD (or ± SEM if indicated). All experiments were performed at least 3 times independently and used more than 4 technical replicates for each experimental condition tested (unless otherwise specified, such as when representative data are shown). Statistical significance was determined using Student's t-test or analysis of variance (ANOVA) test. For comparisons between multiple groups, one-way ANOVA was used to determine statistical significance. p < 0.05 was considered significant, and all statistical tests were performed two-sided: p* < 0.05; p** < 0.01; p*** < 0.005; p**** < 0.0001.

[0123] Bioinformatics analysis: Unbiased label-free proteomics comparing two-dimensional monolayers and three-dimensional mammospheres was performed as described above using MCF7 and T47D breast cancer cell lines. Informatic analysis was performed using various publicly available GEO DataSets (GSE36953; GSE2034; GSE59000; GSE55470) registered in the NCBI database and related to three-dimensional growth, metastasis, and circulating tumor cells (CTCs). Gene expression profiling data were extracted from these GEO DataSets. Heatmaps were created with QIAGEN OmicSoft Suite software. Volcano plots were created by examining the annotations provided in OncoLand's Metastatic Cancer (QIAGEN OmicSoft Suite). Furthermore, a functional "core analysis" was performed using Ingenuity Pathway Analysis software (IPA; QIAGEN) on the annotated genes. Gene co-expression profiles were extracted from The Metastatic Breast Cancer Project (Provisional, 2020) using cBioPortal (https: / / www.cbioportal.org / ), and mRNA expression profiling (RNA Seq V2 RSEM) was performed by RNA sequencing of metastatic breast cancer samples from 146 patients.

[0124] Kaplan–Meier (KM) analysis: To perform the KM analysis of ATP5F1C, publicly available microarray data of up to 3,951 breast cancer patients were queried using an open-access online survival analysis tool. For this purpose, data from mainly ER(+) patients were analyzed. Biased array data were excluded from the analysis. As a result, ATP5F1C (also known as ATP5C1) could be identified as a significant prognostic marker. The hazard ratio was calculated at the automatically selected optimal cut-off value, the p-value was calculated by the log-rank test, and plotted in R. The KM curve was created online using the KM plotter (high-resolution TIFF file) with univariate analysis: https: / / kmplot.com / analysis / index.php?p=service&cancer=breast.

[0125] By this method, the in silico validation of ATP5F1C as a marker for tumor recurrence (RFS, recurrence-free survival) and distant metastasis (DMFS, distant metastasis-free survival) could be directly performed. All of these analyses utilized the latest 2020 version of the database.

[0126] Cell migration assay: Briefly, 2.5×10 4 cells in 0.5 mL of serum-free DMEM containing 0.1% BSA were added to the wells of an uncoated membrane-modified Boyden chamber (Transwell) with an 8-μm pore size. The lower chamber was filled with 10% fetal bovine serum in DMEM as a chemoattractant. The cells were incubated at 37 °C and allowed to migrate for 6 hours. Non-invasive cells were removed from the upper surface of the membrane by rubbing with a cotton swab. The chambers were stained with 0.5% crystal violet diluted in 100% methanol for 30 - 60 minutes, rinsed with water, and then observed under a bright-field microscope. The values of invasion and migration were obtained by counting 5 fields per membrane (20× objective lens), and are the average of 3 independent experiments. Note that aggressive cell invasion was measured using Transwells pre-coated with extracellular matrix (i.e., Matrigel), preventing simple cell migration.

[0127] Transplantation assay: The chicken embryo transplantation assay was carried out by INOVOTION (Company: 811310127, France, La Tronche). According to French law, ethical review approval is not required for scientific experiments using oviparous embryos (Decree No. 2013-118, February 1, 2013; Article R-214-88). The animal experiment was carried out based on Animal Experiment Permit No. 381029 and B3851610001 for INOVOTION. Fertilized eggs of white Leghorn were incubated at 37.5 °C and 50% relative humidity for 9 days. More than 20 eggs were processed for each experimental condition. At that time point (E9), the chorioallantoic membrane (CAM) was lowered by making a small hole from the eggshell to the air sac, and a 1 cm 2 window was opened in the eggshell above the CAM. The MDA-MB-231 tumor cell line was cultured in DMEM medium supplemented with 10% FBS and 1% penicillin / streptomycin. On day E9, the cells were detached with trypsin, washed with complete medium, and suspended in transplantation medium. After ATP-based cell sorting by flow cytometry, 30,000 cells were inoculated onto the CAM of each egg (day E9), and then the eggs were randomly divided into several groups. On day E18, 1 cm 2 of the lower CAM was collected, and the number of metastatic cells was evaluated in 8 samples per group (n = 10). Genomic DNA was extracted from the CAM (commercial kit) and analyzed by qPCR using specific primers for the human Alu sequence. The calculation of the Cq, average Cq of each sample, and the relative amount of metastasis of each group were directly managed by Bio-Rad® CFX Maestro software. As a negative control for specificity, uninjected eggs were also evaluated in parallel. For all data, one-way analysis of variance with post hoc tests was performed.

[0128] The terms used in the description of the embodiments of this method are for the purpose of describing specific embodiments only and are not intended to be limiting. The singular forms "a", "an", and "the" used in the description and the appended claims are intended to include the plural forms as well, unless the context clearly dictates otherwise. This method includes numerous alternatives, modifications, and equivalents, as will become apparent upon consideration of the following detailed description.

[0129] Of course, in this specification, terms such as "first", "second", "third", "a)", "b)", and "c)" may be used to describe various elements of the present method. However, the claims should not be limited by these terms. These terms are only used to distinguish one element of the present method from another. Therefore, without departing from the teachings of the present method, the first element described below could equally well be referred to as an aspect of the element or, similarly, as the third element. Therefore, terms such as "first", "second", "third", "a)", "b)", and "c)" are not necessarily intended to impart a hierarchy such as order to the related elements and are only used for identification purposes. The order of operations (or steps) is not limited to the order shown in the claims.

[0130] Unless otherwise defined, all terms (including technical and scientific terms) used in this specification have the same meaning as commonly understood by one of ordinary skill in the art. Furthermore, of course, terms defined as in a commonly used dictionary should be interpreted as having a meaning consistent with the meaning in the context of the present application and the related art, and should not be interpreted in an idealized or overly formal sense unless explicitly defined in this specification. All publications, patent applications, patents, and other references cited in this specification are hereby incorporated by reference in their entirety. In case of any contradiction in terms, the present specification shall prevail.

[0131] Also, as used in this specification, "and / or" refers to any and all possible combinations of one or more of the associated listed items, as well as the absence of combinations (i.e., "or") when selectively interpreted, and encompasses these.

[0132] Unless the context dictates otherwise, it is specifically intended that the various features of the present method described in this specification can be used in any combination. Furthermore, the present method is also contemplated to be able to exclude or omit any feature or combination of features described with respect to the exemplary embodiments in some embodiments.

[0133] As used herein, the transitional phrase "consisting essentially of" (and grammatical variations thereof) shall be construed to include the recited materials or steps "and those that do not materially affect the basic and novel one or more characteristics of the claims." Thus, the term "consisting essentially of" as used herein should not be construed as equivalent to "comprising."

[0134] As used herein, the term "about" when used in reference to a measurable value such as an amount or concentration means that it encompasses variations of ±20%, ±10%, ±5%, ±1%, ±0.5%, and even ±0.1% of the specified amount. The ranges presented herein for measurable values may include any other range and / or individual values therein.

[0135] Thus, while specific embodiments of the method have been described, it will be understood that the appended claims are not limited to the specific details shown in the above description, since many obvious variations are possible without departing from the spirit or scope claimed below.

Claims

1. A purified composition of hyperproliferative cancer stem cells consisting of a subpopulation of cancer stem cells from a human cancer cell population, wherein the cancer cell population expresses a fluorescence signal within a certain range in response to a fluorescent adenosine triphosphate (ATP) imaging probe, the subpopulation of cancer stem cells expresses above the upper limit of the fluorescence signal based on ATP within the certain range, and the composition is separated from the human cancer cell population.

2. The composition according to claim 1, wherein the upper limit includes the top 10% of the fluorescence signal based on ATP.

3. The composition according to claim 1, wherein the upper limit includes the top 5% of the fluorescence signal based on ATP.

4. The composition according to claim 1, wherein the composition is positive for one of the CD44 marker and the ALDH marker.

5. The composition according to claim 1, wherein the subpopulation of cells is stained with ATP-Red1.

6. The composition according to claim 1, wherein the composition is frozen.

7. A purified cell composition comprising a subpopulation of cancer stem cells stained with a fluorescent adenosine triphosphate (ATP) imaging probe and expressing a target portion within the range of the ATP-based fluorescence signal of the cancer cell population, wherein the subpopulation is separated from the cancer cell population, the cancer cell population expresses a certain range of ATP-based fluorescence signal, the target portion within the range of the ATP-based fluorescence signal includes one of the upper part of the ATP-based fluorescence signal and the lower part of the ATP-based fluorescence signal, and the target portion is the top 25% of the ATP-based fluorescence signal.

8. The composition according to claim 7, wherein the target portion includes one of the top 10% of the ATP-based fluorescence signal and the top 5% of the ATP-based fluorescence signal.

9. The composition according to claim 7, wherein the target portion includes one of the bottom 10% of the ATP-based fluorescence signal and the bottom 5% of the ATP-based fluorescence signal.

10. The composition according to claim 8, wherein the composition is positive for one of the CD44 marker and the ALDH marker.

11. The composition according to claim 1, wherein the subpopulation of cells is stained with a fluorescent dye for ATP imaging.

12. The composition according to claim 7, wherein the target portion comprises one of the top 10% of the ATP-based fluorescence signal, the top 5% of the ATP-based fluorescence signal, the bottom 10% of the ATP-based fluorescence signal, and the bottom 5% of the ATP-based fluorescence signal.

13. The composition according to claim 7, wherein the target portion comprises one of the top 10% of the ATP-based fluorescence signal and the top 5% of the ATP-based fluorescence signal, and the isolated cells are positive for one of the CD44 marker and the ALDH marker.

14. The composition according to claim 1 or 7, wherein the fluorescent ATP imaging probe comprises ATP-Red1.

Images