Pharmacological Agents Targeting Drug-Tolerant Persister Cells in Cancer

Abstract

Current cancer therapy can be effective, but the development of drug resistant disease is the usual outcome. These drugs can eliminate most of the tumor burden but often fail to eliminate the rare, “Drug Tolerant Persister” (DTP) cell subpopulations in residual tumors, which can be referred to as “Persister” cells. Therefore, novel therapeutic agents specifically targeting or preventing the development of drug-resistant tumors mediated by the remaining persister cells subpopulations are needed. Since approximately ninety percent of cancer-related deaths occur because of the eventual development of drug resistance, identifying, and dissecting the biology of the persister cells is essential for the creation of drugs to target them. While there remains uncertainty surrounding all the markers identifying DTP cells in the literature, this review summarizes the drugs and therapeutic approaches that are available to target the persister cell subpopulations expressing the cellular markers ATP-binding cassette sub-family B member 5 (ABCB5), CD133, CD271, Lysine-specific histone demethylase 5 (KDM5), and aldehyde dehydrogenase (ALDH). Persister cells expressing these markers were selected as the focus of this review because they have been found on cells surviving following drug treatments that promote recurrent drug resistant cancer and are associated with stem cell-like properties, including self-renewal, differentiation, and resistance to therapy. The limitations and obstacles facing the development of agents targeting these DTP cell subpopulations are detailed, with discussion of potential solutions and current research areas needing further exploration.

INTRODUCTION

Drug-Tolerant Persister (DTP) cells which remain after treatments play a key role in disease relapse (1). While approximately 90% of cancer-related deaths are attributed to metastatic disease (2,3), it is crucial not to disregard the non-metastatic niche. DTPs in the primary tumor site, even in the absence of metastases, can contribute to treatment resistance (1), and disease recurrence (4). DTPs from both exhibit similar resilience mechanisms, including slow cycling (1,5), stemness (6), and resistance to therapy (1). Therefore, targeting DTPs in both metastatic and non-metastatic niches is essential for preventing disease recurrence, improving patient outcomes, and ultimately reducing cancer mortality. By addressing the heterogeneity and plasticity of DTPs within both niches, more effective therapeutic strategies that encompass the entire spectrum of the disease, from localized tumors to metastatic spread should be considered.

This review focuses on the putative cancer DTP cell markers, ATP-binding cassette sub-family B member 5 (ABCB5), CD133, CD271, Lysine-specific histone demethylase 5 (KDM5), and aldehyde dehydrogenase (ALDH). The selection of these markers is based on the reported roles of these DTP cell markers in stem cell-like properties, including self-renewal, differentiation, and resistance to therapy. However, it’s essential to note that the landscape of cancer DTP cell markers is continually evolving, and additional markers will likely be identified in the future that also play critical roles in identifying and targeting these resilient cell populations. Marker selection should therefore be guided by robust experimental evidence, including functional studies demonstrating the relevance of these markers in DTP cell biology, as well as validation in preclinical and clinical settings. Hence, while the selected markers have demonstrated significance in certain contexts, it’s prudent to consider the inclusion of other markers based on emerging research findings and the potential utility in improving the identification and targeting of cancer DTP cells.

HALLMARKS OF DTP CELLS

DTP cells have hallmark-characteristics that aid the development of drug resistance and it is important to take them into account when designing strategies to target these unique cells to better manage cancer (7). These important “hallmarks of DTP cells” are first, primary (innate) or acquired persistence or drug resistance (8,9). The distinction between innate and acquired persistence highlights the inherent resilience of DTP cells from initial tumor formation versus those that develop resistance over time due to treatment pressure, necessitating tailored therapeutic strategies for each scenario (9). Second, metastatic potential (10), where the DTP cells can disseminate from the primary tumor site to distant locations, fueling disease progression and posing challenges for effective treatment management by necessitating interventions that target both primary and metastatic lesions. Third, heterogeneity (8), where intracellular tumor heterogeneity within DTP cell populations contributes to varying levels of resistance, complicating treatment approaches as different subpopulations of cancer cells may necessitate diverse therapeutic interventions to effectively eradicate the disease. Fourth, stem cell-like characteristics (8) where the presence of stem cell-like traits in DTP cells, including self-renewal and pluripotency (9), render them particularly resilient to therapy and necessitating innovative treatment strategies to overcome resistance mechanisms. Fifth, alterations in signaling pathways (8) where DTP cells often exhibit changes in signaling pathways related to apoptosis (1), DNA repair (4), and cell cycle regulation (4), enhancing survival and resistance to treatment (6), thus highlighting the importance of targeting these pathways to overcome therapeutic resistance and improve patient outcomes. Sixth, the epithelial-to-mesenchymal transition (EMT) (8), which can modulate the profile of DTP cells, impacting response to treatment and potentially enhancing invasive and metastatic capabilities, highlighting the importance of targeting EMT-associated pathways to prevent disease progression and metastasis. Seventh, tumor microenvironment influences on the behavior of DTP cells (8), where the tumor microenvironment can modulate the behavior of DTP cells, shaping survival mechanisms and resistance strategies, necessitating a comprehensive understanding of tumor-host interactions to develop effective therapeutic interventions targeting both cancer cells and the microenvironment. The reverse is also possible where DTPs may also develop mechanisms to evade the immune response (11), creating an immunosuppressive environment (12) that facilitates survival and proliferation.

DTP MARKERS FOCUSED ON IN THIS REVIEW

This review prioritizes DTP markers associated with stemness, drug efflux, epigenetic regulation, and/or metabolism, which are particularly relevant to the biology as well as hallmarks of DTPs and have been extensively studied in the context of treatment resistance. However, a more comprehensive listing of these and other DTP markers is provided in Table 1, which lists the DTP marker, the tumor types in which it has been found and techniques that led to identification. While the listing in the table is comprehensive, the focus of this review is only on the ABCB5, CD133, CD271, KDM5, and ALDH DTPs.

Table 1.

Comprehensive listing of putative DTP markers including the tumor types where they have been identified and techniques that have been used for identification. The DTP shown in blue are putative and excluded by many due to lack of functions observed with the other five listed in the table, which include conferring drug resistance, promoting cancer stemness, and facilitating the survival of cancer cells following treatment.

MARKER	TUMOUR TYPE	IDENTIFICATION TECHNIQUES
ABCB5	Melanoma, colon, liver and breast, leukemia, head and neck, oral squamous cell, and Merkel cell carcinoma (13)	Immunohistochemistry (IHC), gene expression profiling, flow cytometry, functional assays (14)
CD133	Brain tumors (glioblastoma), colorectal cancer, pancreatic cancer, liver cancer, lung cancer, prostate cancer (15)	Flow cytometry, IHC, gene expression analysis, functional assays (16)
CD271	Melanoma (17), hypopharyngeal cancer (18), esophageal (18), oral squamous cell carcinoma (18), and ovarian cancers (19)	Flow cytometry, IHC, gene expression analysis, functional assays (20)
KDM5	AML, glioblastoma, neuroblastoma, melanoma, breast, lung, gastric, ovarian and prostate cancers (21)	Flow cytometry, IHC, gene expression analysis, functional assays (22)
ALDH	Melanoma, liver, prostate, colon, brain, lung, breast, pancreas, stomach, ovary, and esophagus cancers (18)	Flow cytometry, enzymatic activity assays, gene expression analysis, functional assays (23)
IGF1R	Most cancers exhibit increased IGF1R (24)	Immunohistochemistry (IHC), gene expression analysis, functional assays (25)
Axl	Breast cancer, lung cancer, ovarian cancer, colorectal cancer, and renal cancer (26)	Gene expression analysis, functional assays (27)
Gli1	Medulloblastoma, basal cell carcinoma, glioma, osteosarcoma, rhabdomyosarcoma, melanoma, esophagus cancer, bladder cancer, colorectal cancer, pancreatic cancer, and prostate cancer (28)	Gene expression analysis, IHC, functional assays (29)
Gli2	Medulloblastoma, basal cell carcinoma, lung cancer, breast cancer, and glioma (30)	Gene expression analysis, IHC, functional assays (31)
STK11	Lung cancer, pancreatic cancer, colorectal cancer, cervical cancer, breast cancer (32)	Clinical patient profiling and assessment, IHC (33)

The presumed DTP cell markers that are listed in the blue cells in Table 1 have functional differences when compared to the ABCB5, CD133, CD271, KDM5, and ALDH, DTP cell markers, which leads to controversy as to whether they true markers. For example, the SHH family members, such as Gli1 (28) and Gli2 (34), along with the tumor suppressor gene STK11 (33,35), are considered by some but not others as DTP markers. For the review, they are not considered as DTP markers because of significant differences in molecular functions and roles in cancer biology compared to DTP cell markers ABCB5, CD133, CD271, KDM5, and ALDH. Specifically, SHH family members and STK11 are primarily involved in signaling pathways regulating cell proliferation (28), and survival (36), rather than promoting cancer stemness to facilitating survival following drug treatment.

Other putative marker not included in this review are the SHH family members and STK11. They may contribute to treatment resistance in certain contexts but are not typically considered specific markers of DTP cells based on distinct biological functions and mechanisms of action promoting cancer progression. For example, in medulloblastoma, aberrant activation of Gli1 and Gli2 in the SHH pathway contributes to treatment resistance (37) and tumor recurrence (38). Additionally, loss of STK11 function occurs in lung, breast and colon cancers(32), where it promotes resistance to targeted therapies(39) and chemotherapy(40). These examples underscore the importance of understanding the role of SHH family members and STK11 in driving drug resistance and tumor persistence across diverse cancer types.

Another excluded set of putative DTPs are those expressing the receptor tyrosine kinases (RTKs), including insulin-like growth factor 1 receptor (IGF1R) (6) and Axl receptor tyrosine kinase (Axl) (6). They play critical roles in cancer progression, therapy resistance and are frequently associated with DTPs (6). Briefly, IGF1R is a cell surface receptor involved in various cellular processes, including cell proliferation, survival, and differentiation (41). Activation of IGF1R signaling has been linked to cancer progression and resistance to therapy (42). In DTPs, upregulation of IGF1R signaling pathway components can promote cell survival and evasion of treatment-induced cell death (42). Additionally, IGF1R signaling has been implicated in the maintenance of cancer stem cell-like properties (43), contributing to therapy resistance (42) and disease recurrence (44). Similarly, Axl is a receptor tyrosine kinase that plays a role in regulating cell survival, proliferation, and migration (45). Activation of the Axl signaling pathway has been associated with metastasis (45), epithelial-to-mesenchymal transition (EMT) (45), and therapeutic resistance (45). In DTPs, increased expression of Axl has been observed and linked to enhanced survival (27) and metastatic potential (46), as well as resistance to various anticancer agents (46). Future research may further elucidate the interplay between RTK signaling pathways and other resistance mechanisms in DTPs, offering new insights into targeted therapeutic strategies.

DRUG RESISTANT DTP CELLS AND THE AGENTS TARGETING THEM

For decades, targeted cancer therapies, such as tyrosine kinase inhibitors (TKI) in non-small lung cancer (NSCLC) (47), MAPK pathway inhibitors and mutant BRAF inhibitors in melanoma and other tumors (48,49) have initially been effective cancer treatments. However, the effectiveness of these therapies is often shortened by the expansion of a DTP cell population leading to drug resistance.

Resistance to these targeted cancer therapies can be categorized as innate and acquired (50,51). Innate or primary drug resistance refers to the intrinsic ability of cancer cells to evade therapy, which can be accomplished through either pre-existing gene mutations or rapidly adapting to therapies (52). On the other hand, acquired drug resistance happens after a period of initially effective therapeutic treatment (52). In a well-characterized EGFR addicted NSCLC cell line, PC9, drug resistance against the prolonged treatment of a EGFR inhibitor, erlotinib, arose from a variety of mechanisms including secondary mutation of EGFR, EGFR^T790M, and MET amplification (53), which suggests that the drug-tolerant persister state can function as a cell reservoir of cells from which heterogeneous drug resistance mechanisms can evolve.

Overall, four common mechanisms can lead to drug resistance and involve epigenomic modification (54), transcriptional dysregulation (55), metabolic remodeling (56), and tumor microenvironment interaction (57). Epigenomic modification, involving DNA hypermethylation and histone modifications (54) of tumor suppressor gene promoters can silence expression, leading to resistance to chemotherapy drugs like cisplatin in various cancers (58). Transcriptional dysregulation of gene expression through alterations in transcriptional control mechanisms can confer drug resistance(55). For example, upregulation of drug efflux transporters like ATP-binding cassette (ABC) transporters, such as ABCB1, can enhance the efflux of chemotherapeutic agents from cancer cells, reducing their intracellular concentration and efficacy(59). Metabolic remodeling can aid drug resistance by enabling cancer cells to adapt cellular metabolism to support rapid proliferation and survival under stress conditions, including exposure to anticancer drugs (56). For example, increased expression of enzymes involved in glycolysis, such as lactate dehydrogenase (LDH), and altered mitochondrial metabolism can promote resistance to therapies targeting metabolic vulnerabilities in cancer cells (56). Tumor microenvironment interaction can change to enable drug resistance (57). The tumor microenvironment, comprising stromal cells, immune cells, and extracellular matrix components, plays a crucial role in modulating drug response. Immunosuppressive factors within the tumor microenvironment, such as regulatory T cells and myeloid-derived suppressor cells (57), can dampen antitumor immune responses and promote resistance to immunotherapy drugs like immune checkpoint inhibitors (60,61). Additionally, hypoxia-induced signaling pathways (62) and the secretion of growth factors by stromal cells (63) can confer resistance to targeted therapies by promoting tumor cell survival and proliferation (62,63).

The processes leading to innate and acquired drug resistance occur in DTPs through several pathways, including DTP mediated involvement due to high levels of marker expression, drug efflux to reduce effective agent concentrations, drug inactivation to inhibit cell death, changes in DNA damage repair to circumvent drug effects, altered proliferation patterns to overcome drug inhibition, modifications in EMT to mitigate drug killing, epigenetic alterations in gene expression patterns to overcome drug inhibition antioxidative stress, mitochondrial oxidative phosphorylation, alterations in the WNT/b-catenin pathway, altered signaling in the YAP-TEAD pathways, and autophagy (6,64). The innate chemoresistance and acquired resistance to specific anticancer agents mediated by DTPs is shown in Table 2. The summary in the table shows that the five DTP markers (ABCB5, CD133, CD271, KDM5, and ALDH) are associated with innate and acquired resistance mechanisms. The innate resistance for ABCB5(65), CD133(15), CD271(66), KDM5(67), and ALDH(68) are associated with the upregulated expression of these markers in cancer stem cells (CSCs) or progenitor cells, which inherently possess stem cell-like properties such as self-renewal and resistance to therapy (69). Under these circumstances, elevated expression before therapy would significantly contribute to the survival and resistance to drugs, making them less susceptible to treatment leading to general chemoresistance. The acquired resistance of the DTP cells to specific agents is listed in Table 2. The listed references provide details of the mechanisms though which the acquired resistance develops. Understanding the innate and acquired mechanisms through which DTP markers contribute to resistance is crucial for developing effective strategies to overcome resistance and improve cancer treatment outcomes.

Table 2.

Innate and acquired chemoresistance to specific drugs that have been identified for each persister cell marker.

DTP CELL TYPE
	ABCB5	CD133	CD271	KDM5	ALDH
Resistance to Drug	Innate Chemoresistance	Innate Chemoresistance	Innate Chemoresistance	Innate Chemoresistance	Innate Chemoresistance
	Acquired Chemoresistance: Rodamine 123 (13) Doxorubicin (13) 5-Fluorouracil (13) Daunorubicin (13) Clorgylin (13) Taxane (13) Anthracyclines (13)	Acquired Chemoresistance: Doxorubicin (79) 5- Fluorouracil (78, 79)		Acquired Chemoresistance: Cisplatin (22) Temozolomide (22) Bortezomib (22) Vemurafenib (22)	Acquired Chemoresistance: Paclitaxel (122) Cisplatin (123) Cytarabine (136)

Drugs have been developed that can target the various DTP subpopulations, which has required unraveling the biology of these cellular markers, and then developing drugs to target them. Strategies targeting each of the currently identified DTP subpopulations in particular cancer subtypes have been developed and evaluated in preclinical studies, which are summarized in Table 3. However, these studies are just scratching the surface and revealing the complex problems faced when using drugs to target these cells. For example, drugs targeting specific DTPs can be hindered when they express multiple DTP markers in a single cell or tumors consist of heterologous subpopulations (70), such as occurs for many of the cancers and is illustrated in Table 4 with the references cited for each. Under these circumstances, it is not yet clear how the expression of multiple markers might lead to resistance and therefore remains a major unanswered question in the field. However, the current thought is to target the DTP with available agents regardless of how many other markers might be present.

Table 3.

Drugs evaluated preclinically to target specific DTPs in different cancer types.

DRUGS TARGETING PERSITER CELLS EXPRESSIING SPECIFIC MARKERS IN CANCER TYPES
DTP CELL TYPE
	NON-SPECIFIC		CANCER TYPE
ABCB5	Etoposide Carboplatin Rodamine 123, Doxorubicin, 5-fluorouracil, Daunorubicin, Clorgyline, Taxanes, Anthracyclines		Merkel Cell Carcinoma (18)
	SPECIFIC		CANCER TYPE
	Blocking Antibodies		Merkel Cell Carcinoma (72)
	1,2,4-trioxolanes		Liver Cancer (73)
	RNAi		Colorectal Cancer (14)
	Monoclonal Antibodies		Melanoma (74, 75)
CD133	Redirected T cells
	Bispecific Antibody
	CD133-OKT3		Colorectal Cancer (81)
	CD133-CD3		Pancreatic Cancer (82)
	NK Cell Mediated Toxicity
	Bispecific Antibody
	CD133-CD16		Colorectal Cancer (83)
			Burkitt’s Lymphoma (83)
	Antibody Conjugated Nanoparticles
	αCD133-paclitaxel		Colorectal Cancer (84)
			Breast Cancer (84, 85)
	Immunotoxin
	CD133KDel		Head and Neck Cancer (87)
	αCD133-pseudomonas endotoxin		Ovarian Cancer (89)
			Triple Negative Breast Cancer (90)
	Aptamers
	Aptamers Binding- CD133		Colorectal Cancer (92)
	Fungal Immunomodulatory Protein
	GM1		Lung Cancer (80)
	3-phenylthiazolo [3,2-a] benzimidazoles		Colon Cancer (93) Breast Cancer (93)
CD271	αCD271		Melanoma (18)
	CD271 Small Molecule Ligand (LM11A-31)		Ovarian Cancer (19)
	Small Molecule Drugs: PD90780		Stomach Cancer (97)
KDM5	N-Oxalylglycine (NOG)		Not evaluated in cancer
	2,4-Pyridinedicarboxylic Acid (2,4-PDCA)		Osteosarcoma (103)
	KDOAM-25		Myeloma (104) Uveal Melanoma (115)
	KDM5-C49		Myeloma (116)
	Non-Carboxylate Inhibitors
	GSK- 467		Liver Cancer (117)
	GSK-J1		Head & Neck Squamous Cell Carcinoma (118)
	4-[pyridine-2-yl]thiazol-2-amino analog		Not evaluated in cancer
	CPI-455		Melanoma (110) Breast Cancer (110) Non-small Cell Lung Cancer (110)
	Pyrazolyl pyridine analogs		Breast cancer (111)
	Pyrazole analog 1-[4-methoxyphenyl] -n- [2-methyl-2-morpholinopropyl]-3-Phenyl-1H-Pyrazole-4-Carboxamide analogs		Gastric Cancer (112)
ALDH	Isoform Specific Inhibitors
	NCT-506	Targets ALDH1A1	Ovarian Cancer (122, 123)
	Pyrazolopyrimidinone CM39 analog	Targets ALDH1A1
	CVT-10216	Targets ALDH2	Leukemia (AML) (136)
	CB7	Targets ALDH3A1	Lung Adenocarcinoma (126) Glioblastoma (126)
	Broad Spectrum Inhibitors
	DIMATE		Prostate Cancer (128–130) Melanoma (131) Non-small Cell Lung Cancer (138) Leukemia (AML) (139)
	DEAB		Colon (140) Ovarian Cancer (141, 143) Pancreatic Cancer (142) Endometrial Cancer (144) Triple-negative Breast Cancer (132)
	Aldi-6		Head & Neck Squamous Cell Carcinoma (145)
	KS100		Melanoma (3)

Table 4.

Sets of persister cell markers ABCB5, CD133, CD271, KDM5 and ALDH that are present in the same cancer and sets present in different cancer types.

DTP Markers	ABCB5	CD133	CD271	KDM5	ALDH
Cancer Type
Brain		✓ (15)			✓ (126)
Breast	✓ (13)	✓ (84, 85)		✓ (111)	✓ (132)
Burkitt’s Lymphoma		✓ (83)
Colon/Colorectal	✓ (13)	✓ (15, 83)			✓ (137)
Esophageal			✓ (18)
Head and Neck Squamous Cell carcinoma	✓ (13)		✓ (18)	✓ (118)	✓ (133)
Hypopharyngeal			✓ (18)
Leukemia	✓ (13)
Leukemia (AML)					✓ (136)
Liver	✓ (13)	✓ (15, 79)		✓ (117)
Lung		✓ (15)		✓ (110)	✓ (126, 138)
Melanoma			✓ (17)	✓ (99)	✓ (51, 131)
Merkel Cell Carcinoma	✓ (13)
Myeloma				✓ (105, 116)
Ovarian			✓ (19)		✓ (122)
Pancreas		✓ (15)			✓ (142)
Prostate		✓ (15)			✓ (128–130)
Stomach		✓ (78)		✓ (112)

Another obstacle for the use of agents targeting persister cells in the clinic is that DTP drugs generally target a small subpopulation of cells, while the majority would not be affected. Therefore, drug combinations will likely be needed for successful incorporation into the clinic, which will require solving typical issues of drug combination toxicity, drug ratios for optimal potency, and delivery issues when agents are not administered through the same route. These are critically important areas needing substantial development before the fully clinical potential of agents targeting DTPs can be fully realized.

PERSISTER CELLS IDENTIFIED BY SINGLE MARKERS AND THE DRUGS AVAILABLE TO TARGET THEM.

Persister cells expressing ABCB5.

The ABCB5 transmembrane transporter facilitates drug resistance by pumping structurally diverse drugs out of the cells, thereby enabling cell survival (18). ABCB5 is predominantly expressed in normal pigmented cells, such as melanocytes and retinal epithelial cells (13). Elevated ABCB5 can be found in numerous cancers, including those of the colon, liver and breast, as well as, leukemia, head and neck, oral squamous cell, and Merkel cell carcinoma (13). By utilizing the energy of ATP hydrolysis, ABCB5 translocates binding substrates across the cellular membrane to expel them from cells (71). ABCB5 increased the non-specific drug efflux of anticancer agents such as rodamine 123, doxorubicin, 5-fluorouracil, daunorubicin, clorgyline, taxanes and anthracyclines, as listed in Table 3 (13).

Drugs to target ABCB5 expressing persister cells.

Nonspecific agents such as etoposide and carboplatin, shown in Table 3, have been used to overcome drug resistance mediated by ABCB5^+ve persister cells in Merkel cell carcinoma cell lines MKL-1 and WaGa (18). These approaches can be successful even though the drugs do not specifically block the activity of the transporter but kill persister and non-persister cells alike. Perhaps this might be a future approach explored to target all DTP subpopulations, which would enable more rapid movement to clinical utility.

Specific agents have also been developed to target ABCB5. ABCB5 blocking antibodies that disrupt the drug pumping function of ABCB5 have improved the efficacy of etoposide or carboplatin for Merkel cell carcinoma (72). Relatively few small molecule ABCB5 specific inhibitors have been developed. One approach used a structure-activity drug development strategy focusing on 1,2,4-trioxolanes to identify lead compounds (Figure 1, compounds 1 and 2) that strongly inhibit ABCB5 P-glycoprotein-overexpressing HepG2 liver cancer cells (73). Other specific approaches to target ABCB5 have used monoclonal antibodies (74,75) and short hairpin [sh] RNA-mediated knockdown (14), to reverse melanoma resistance to doxorubicin (75) and to inhibit the colorectal tumor growth (14). Gene silencing of ABCB5 sensitized melanoma cells to 5-fluorouracil and camptothecin (76). Although, these findings were encouraging and strongly suggest that targeting of ABCB5 in conjunction with chemo or immune-therapy may have additive and potentially even synergistic effects, most of the reports are more than a decade old and the field remains largely unexplored. Therefore, drugs targeting ABCB5 cells nonspecifically or its function are one potential strategy for reducing drug resistant cancer recurrence (72).

Figure 1.

Drugs targeting ABCB5 [1 and 2], CD133 [3 and 4] and CD277 [LM11A-31, 5]. PD 90780 [6] and Ro 08-2750 [7] are small molecule inhibitors preventing nerve growth factor [NGF] from binding to CD271.

Persister cells expressing CD133.

CD133 is a transmembrane cell surface glycoprotein that has frequently been used to isolate cancer stem cells from tumor of the brain, colon, pancreas, prostate, lung, and liver (15). CD133 regulates the stem cell-status and fate of cells (15). In addition, CD133 can promote resistance to chemotherapy (18) and radiotherapy (77). Recently, CD133-expressing gastric cancer cells were shown to be more resistant to 5-fluorouracil (5-FU) than cells lacking CD133 (78). CD133 overexpression in gastric cancer cells significantly enhances 5-FU resistance while CD133 reduction promotes 5-FU cytotoxicity and apoptosis (78). In hepatocellular carcinoma (HCC), a CSC subpopulation lacking CD133 expression was found to be sensitive to chemotherapeutic agents such as 5-FU or doxorubicin while CD133+ HCC cells tolerated higher doses of chemotherapy (79). Interestingly, treatment with 5-FU and doxorubicin significantly enriched the CD133+ subpopulation which expresses high levels of survival proteins involved in the Akt/PKB and Bcl-2 pathways as a potential mechanism by which CD133 promotes drug resistance.

Drugs to target CD133 expressing persister cells.

Several approaches to target CD133 have been reported, outlined in Table 3, including those that function through redirected T cells (18), NK cell-mediated cytotoxicity (18), antibody-conjugated nanoparticles (18), immunotoxins (18), aptamers (18) and a fungal immunomodulatory protein, called GMI (80). None of these CD133-targeted approaches have been evaluated clinically but show intriguing results in preclinical studies as detailed below:

T cell redirection.

An asymmetric bispecific antibody (BsAbl) made to the AC133 epitope of CD133 (mouse anti-human CD133 monoclonal antibody) and a single chain of humanized OKT3 (anti-human CD3) has been developed to arm activated T cells against CD133+ cells (81). This BsAbl showed a dual-high specificity for CD3 and CD133, which allowed for longer and more potent cytotoxicity in a colorectal cancer (CRC) model. This approach selectively targets CD133^high compared to CD133^low CRC cells, resulting in T cell production of cytokines [interferon-γ and GM-CSF], and inhibiting tumor growth and development in nonobese diabetic-severe combined immunodeficient mice with neglectable toxicity (81). Using the same approach, a BsAbl targeting CD3 and CD133 has been developed (82). When bound to cytokine-induced killer cells, it significantly reduced pancreatic tumor development mediated by CD133^high cells (82). Thus, simultaneously targeting CD3 and CD133 is a promising approach to eliminate this DTP subpopulation.

NK cell-mediated cytotoxicity.

A BsAbl has been engineered which can simultaneously bind CD16 on NK cells and CD133 on DTPs, which is listed in Table 3 (83). These immune engagers stimulate an immunological synapse between NK cells and targeted cancer cells to induce antibody-dependent cell-mediated cytotoxicity. This biological agent increased NK cell mediated killing of human colorectal Caco-2 cells (high CD133 expression) as well as NK-resistant human Burkitt’s lymphoma (low CD133 expression). These agents have demonstrated that innate immune cells can be recruited to kill DTPs.

Antibody-conjugated nanoparticles.

Table 3 shows a nanoparticle containing paclitaxel and conjugated with an anti-CD133 antibody (clone 7) (84) having high efficacy for eliminating colorectal Caco-2 cells and reducing DTPs in breast cancer (85). In addition, this nanoparticle therapy reduced tumor growth significantly in an orthotopic breast cancer model (84).

Immunotoxins.

Table 3 illustrates that CD133 has been targeted using an immunotoxin (86) in which a deimmunized targeted toxin called CD133KDEL, which has been synthesized by combining an anti-CD133 scFv reactive against the extracellular domain of CD133, and a truncated, deimmunized form of pseudomonas exotoxin A (PE38) (87). Once internalized, PE38 catalyzed ADP-ribosylation of elongation factor 2 (eF-2) and inhibited protein synthesis causing cell death (88). This immunotoxin was effective in multiple cancer types, including head and neck (87), ovarian (89), and triple negative breast cancers (90). These findings suggest targeting CD133 with an immunotoxin can effectively eliminate DTPs and prevent cancer relapse.

Aptamers.

Aptamers are short bands of either DNA or RNA whose 3D-structure provide them with the capacity to bind to specific cell surface molecules (86). Aptamers have several advantages over antibodies with higher stability, lower immunogenic potential, easier synthesis and lower cost (91). Aptamers can enable better tumor penetration due to small size and efficient epitope binding (86). Two aptamers capable of binding the AC133 epitope of CD133 have been developed (92). Using a 3D cell culture colorectal tumor sphere model, these aptamers displayed superior efficiency in tumor penetration, retention, and internalization when compared to an AC133 antibody. Overall, aptamers can be utilized efficiently to target cancer cells, but to date have not been utilized to eliminate CD133+ cells, which might be a promising area of future research.

Fungal immunomodulatory protein, GMI.

An immunomodulatory protein identified from fungus, called GMI, has been identified that inhibits CD133 expression in pemetrexed-resistant lung cancer cells. It led to autophagy, which decreased DTP cell survival and proliferation (80). The inhibitory activity of four 3-phenylthiazolo[3,2-a]benzimidazoles related fungal compounds for reducing cell surface expression of CD133 were tested (93). Two of the four compounds [Figure 1, compounds 3 and 4] inhibited CD133 expression and retarded the growth of colon cancer HT-29 and triple negative breast cancer MDA-MB-468 cells. However, the binding and selectivity for CD133 and the detailed mechanism behind the anti-proliferation effect were not demonstrated in this study but may provide a scaffold on which to build more specific CD133 inhibitors in the future.

In summary, significant progress has been made to target CD133, which suggest that this strategy has the potential to enhance immune recognition, promote drug delivery, and reduce cancer growth in several cancer types. The underlying mechanisms of CD133-mediated tumor inhibition still requires in-depth investigation, and this knowledge could be used to design new therapies to eliminate DTP cells.

Persister cells expressing CD271.

The cell surface protein CD271 is a low-affinity nerve growth factor (NGF) receptor, also named p75(NTR), which has been found on DTP cells (18). CD271 plays roles in cell survival, proliferation and apoptosis (94). Upregulated CD271 expression is clinically associated with several cancers, including melanoma (17), hypopharyngeal cancer (18), esophageal (18), oral squamous cell carcinoma (18), and ovarian cancers (19). CD271+ DTP cells exhibit high self-renewal capacity and chemoresistance (95). CD271 expression correlates with tumor stage and metastasis (95), with expression regulated epigenetically by DNA methylation (95). Epigenetic modulation of CD271 expression therefore provides one potential approach to target it, but current agents are non-specific, affecting the expression of many genes.

Drugs to target CD271 expressing persister cells.

Antibodies, ligands and drugs have been developed that bind to and target CD271, which are summarized in Table 3, but there is concern that broadly targeting cells expressing this protein could have negative side-effects because CD271 has important roles in the central nervous system (18). An anti-CD271 antibody-based approach has been used to decrease melanoma metastasis (18), suggesting that this DTP subpopulation can be therapeutically targeted to manage cancer. Concern related to the negative side-effects has significantly delayed the development of specific inhibitors targeting CD271. However, a small molecule ligand of CD271, called LM11A-31 (Figure 1 compound 5), has been developed that that inhibits CD271-mediated apoptosis of neurons for treating Alzheimer’s disease (NCT03069014; https://clinicaltrials.gov/ct2/show/NCT03069014). It has been shown to block NGF-stimulated changes of gene expression or migratory behavior of ovarian cancer cells (19). Other small molecule inhibitors called PD90780 (Figure 1, compound 6) and Ro 08-2750 (Figure 1, compound 7) have been shown to prevent NGF from binding to CD271 (96). PD90780 effectively reduces the expression of MMP9 and decreases the invasion potential of the IFT80-overexpressed SGC-7901 gastric cancer cells (97). However, these molecules remain to be tested in animal models (98), which might demonstrate that a proper delivery system in needed to specifically target CD271 in cancer cells without interfering the neural functions.

Persister cells expressing KDM5.

The KDM5 demethylase subfamily are members of the Jumonji (JmjC) KDMs (lysine demethylase) which share 2-oxoglutarate (2-OG) as a co-substrate (67). The KDM5 subfamily remove tri- and di-methyl marks from lysine 4 on histone H3 (H3K4) and depending on the methylation site, activate or suppress gene transcription. There are 4 members in the KDM5 subfamily called KDM5A-D, which share a high level of sequence homology and domain arrangement, leading to functional redundancy (67). KDM5A-D have all been linked to various cancers as oncogenic drivers (67). Particularly, KDM5A and KDM5B (also known as JARID1B) play roles in cell proliferation, regulation of tumor suppressor gene expression and enhancing drug tolerance. KDM5B has been used to isolate slow-cycling DTP cells with stem cell like features and when released from its existing microenvironment can produce rapid-proliferating offspring (99). In melanoma, KDM5B is important for xenografted tumor growth and metastatic progression (99). Depleting KDM5B from these slow-cycling melanoma cells can sensitize them to multiple anti-cancer drugs, such as cisplatin, temozolomide, bortezomib and vemurafenib (22).

Drugs to target KDM5 expressing persister cells.

Since all members of KDM5 share the same 2-OG co-substrate, many of the drugs, listed in Table 3, have been developed that target all family members, and fewer that only target particular ones (67). The identification of specific inhibitors directly targeting KDM5B/JARID1B required identifying the specific role of each KDM5 enzyme in cancer and normal cells (100). Representative examples of the structures of these drugs are shown in Table 5 and the unique properties for each are detailed below.

Table 5.

Drugs targeting KDM5 expressing persister cells.

N-Oxalylglycine [NOG].

NOG [compound 1, Table 5] was the first KDM5B inhibitor identified with an in vitro IC₅₀ of 24.5 μM. It was originally identified as a pan inhibitor of 2-OG oxygenase by binding in the pocket and chelating Fe2+ through its C-1 carboxylate oxygen and amido group (113,114). However, more potent, potentially better KDM5B inhibitors than NOG have now been identified, so there is little current interest in this compound.

2,4-Pyridinedicarboxylic acid [2,4-PDCA].

2,4-PDCA [compound 2, Table 5] that structurally mimics 2-OG [also known as α-ketoglutarate], is an inhibitor of histone lysine-specific demethylases that targets JMJD2A [KDM4A], KDM4C, KDM4E, KDM5B, KDM6A and other 2-oxogynases. It inhibits KDM5B with an IC₅₀ value of 3 ± 1 μM using the FDH-coupled assay (103). 2,4-PDCA inhibited the decrease of H3K4me3 in U2-OS osteosarcoma cells transfected with KDM5B (103) and also decreased proliferation and migration of high glucose treated vascular smooth muscle cells (102). Additional structural modifications on the 2,4-PDCA core have produced even more robust and selective inhibitors of KDM5B, which are KDM5-C49, and KDOAM-25.

KDOAM-25.

A range of compounds based on the 4-carboxypyridine core with a 2-aminomethyl group have been created (104). KDOAM-25, with an amide substitution at the pyridyl ring [compound 3, Table 5], has been identified as the most potent and highly selective inhibitor of KDM5. It acted as a pan-inhibitor with IC₅₀s of 71 nM, 19 nM, 69 nM, and 69 nM for KDM5A, KDM5B, KDM5C, and KDM5D, respectively. KDOAM-25 showed only moderate selectivity for KDM5B having an IC₅₀ of ~50 μM and for the other KDM5 family members at doses above 100μM. However, it had no cellular activity on any of the other JmjC family members. KDOAM-25 increased global H3K4 methylation at transcriptional start sites and impaired proliferation in multiple myeloma MM1S cells. Recently, Zhang et. al reported that KDOAM-25 effectively overcame the resistance to MEK inhibitors and decreased cell viability in uveal melanoma (115). Because of its selectivity for KDM5 enzymes, KDOAM-25 may be useful in elucidating the biological roles under different cellular conditions. Despite its good stability, high selectivity, and moderate cytotoxicity, the results with KDOAM-25 should be interpreted with caution since off-targets could not be excluded due to an IC₅₀ of ~50 μM. However, no other KDM sub-families were inhibited at this dose.

KDM5-C49.

KDM5-C49 [compound 4, Table 5] was also designed based on the 2,4-PDCA scaffold and found to be a potent and selective pan-KDM5 histone demethylase inhibitor (105,106). KDM5-C49 had >100-fold selectivity over KDM4 and KDM6 and showed 25-100-fold selectivity between KDM5B and KDM6B. The occupancy of the 2-OG-binding site was shown in the crystal structure of KDM5-C49 with KDM5B (PDB ID: 5a3t). Although, KDM5-C49 is potent and selective in vitro, the free polar carboxylic acid group in its structure has restricted its cellular permeability, thereby reducing its usefulness for inhibiting demethylation of H3K4 in cancer cells. This limitation led to the subsequent development of a cell permeable ethyl ester prodrug called KDM5-C70, that masks the polarity of the acid group of the KDM5-C49 but is hydrolyzed by an esterase within the cell to regenerate active KDM5-C49. As expected, the prodrug KDM5-C70 showed marked activity in cells and significantly increased global levels of H3K4me3, and thus could be a good candidate for in vivo use. JQKD82, a KDM5-C49 derivative, showed significant growth suppression of myeloma cells both in vitro and in vivo, suggesting a promising therapeutic application for targeting KDM5 (116).

NON-CARBOXYLATE INHIBITORS.

GSK467.

A series of non-carboxylate inhibitors of the KDM4 and KDM5 families have been designed and synthesized, identifying GSK467 (compound 5, Table 5) as a submicromolar inhibitor for the KDM4 family and KDM5C with activity [IC₅₀ < 10 μM] in cellular imaging assays (107). However, using an AlphaScreen assay, GSK467 was found to have a calculated Ki value of 10 nM for KDM5B with ~180-fold selectivity for KDM4C, while no inhibitory effects toward KDM6 or other family members were observed. GSK467 has also been co-crystalized with KDM5B, establishing the structural interactions. GSK467 has been showed to reduce spheroid formation, colony formation, invasion and migration of HCC cells in vitro and decrease the tumor burden in vivo (117). However, other studies indicated that GSK467 lacked efficacy for reducing myeloma DTPs (105). Despite of this disadvantage, it is thought that the structure of GSK467 may still be useful for creating a more selective KDMB5 inhibitor with cellular activity.

GSK-J1.

GSK-J1, (compound 6, Table 5) was reported by Heinemann et al. in 2014 (108) as a KDM5 inhibitor. However, it was found to be a semi-selective inhibitor of the KDM6/KDM5 subfamilies and was ~8.5-fold more active against KDM5C than against KDM5B (106). The combination treatment of GSK-J1 and a lysine specific demethylase 1 inhibitor, TCP, impaired cell proliferation, caused apoptosis and senescence of HNSCC cells in vitro and retarded tumor growth and progression in vivo (118). More work is needed to evaluate the clinical value of GSK-J1 in cancer treatments.

4-[pyridin-2-yl]thiazol-2-amine analog.

4-[pyridin-2-yl]thiazol-2-amine was identified through a high throughput screen for KDM inhibitors. This lead compound underwent structure-based optimization to develop potent, cell permeable dual inhibitors of the KDM4 and KDM5 subfamilies. Incorporation of a rigidifying piperidine linker into the pyrazole C4 substituent, followed by the insertion of a meta-substituted phenyl ring at the piperidine 4-position, (compound 7, Table 5) led to the identification of a potentially useful inhibitor. The binding mode and favorable interactions of crystal structure of 4-[pyridin-2-yl]thiazol-2-amine in KDM4A was important to the balanced KDM4 and KDM5 activity profile. Incorporation of the conformationally constrained 4-phenylpiperidine resulted in selective inhibition of the KDM4 and KDM5 subfamily demethylases over KDM2, KDM3, and KDM6 subfamilies, with an IC₅₀ for KDMB5 of 0.014 nM.

CPI-455.

CPI-455 (compound 8, Table 5) is an example of a pan-KDM5 inhibitor, which targets the activity of all family members. A high-throughput screen of >100,000 commercial compound chemical library against the KDM4C JmjC domain followed by structural optimization, led to the development of CPI-455 (110). The survival of KDM5 expressing persister cells could be eradicated with CPI-455 and was dependent on the demethylase activity of KDM5, since there was an increase in global H3K4 trimethylation (H3K4me3) (119). Besides, CPI-455 decreased the number of DTPs in melanoma, breast and NSCLC cell lines treated with standard chemotherapy or targeted agents (110). CPI-455 is a potent pan-KDM5 inhibitor and had potency against KDM5A while demonstrating ∼200-fold selectivity for KDM5A over KDM4C (119). Given that drug-tolerant persister cells are a barrier to long-term treatment responses, inhibitors such as CPI-455 may have clinical utility.

Pyrazolyl pyridine analogs.

A series of pyrazolyl pyridines have been designed based on structural information from known KDM4 and KDM5 inhibitors, which led to the identification of an orally bioavailable (compound 9, Table 5) (111), as a potent and selective inhibitor of KDM5A/5B. Compound 9 increased H3K4me3 in the breast cancer cell line, ZR-75-1, and in an MCF-7 breast cancer xenograft PK/PD model. Due to difficulties in obtaining a KDM5B co-crystal structure, KDM4A was used as a surrogate to determine polar interactions with protein residues, to guide structural design. The co-crystal structure of (compound 9) bound to KDM5A has been solved (106), and could be useful to discover better KDM5A/5B inhibitors. Compound 9 could serve as a valuable tool compound to interrogate the biology of KDM5A/5B.

Pyrazole analog 1-[4-methoxyphenyl]-n-[2-methyl-2-morpholinopropyl]-3-Phenyl-1h-Pyrazole-4-Carboxamide analogs.

Employing a structure-based virtual and biochemical screening approach of 20 million molecules from the Enamine library, a pyrazole derivative KDM5B inhibitor was identified (112). The compound was refined by generating a series of pyrazole derivatives through structural based optimization and evaluation of inhibitory activities against KDM5B. This SAR study led to the identification of a potent pyrazole analog 1-[4-methoxyphenyl]-N-[2-methyl-2-morpholinopropyl]-3-phenyl-1H-pyrazole-4-carboxamide (compound 10, Table 5), as a selective KDM5B antagonist with an IC₅₀ of 0.0244 μM(112). Compound 10, Table 5, was shown to bind and stabilize KDM5B in MKN45 gastric cancer cells, causing accumulation of substrates of KDM5B, H3K4me2 and H3K4Me3, while not changing the amounts of H3K4me1, H3K9me2/3 and H3K27me2, suggesting that it was a potent and cellular active KDM5B inhibitor. Compound 10, Table 5 also inhibited MKN45 cell proliferation, wound healing, and colony formation, suggesting a cancer therapeutic potential. The in vivo efficacy and safety profile remain to be determined.

Persister cells expressing Aldehyde dehydrogenases (ALDHs).

The aldehyde dehydrogenase [ALDH] marker of DTP is considered by some to be a master tracker of persister cell subpopulations (18,120). The marker is a measure of the total aldehyde dehydrogenase activity for a group of 19 oxidoreductive enzymes of which 18 detoxify aldehydes and convert them to carboxylic acids, to enable vital cellular viability (120,121). ALDHs are needed for normal stem cell growth, differentiation, and maintenance (120). They are involved in the manufacture of retinoic acid (RA) (120) and protect cells against the negative effects of reactive oxygen species (ROS) (23). High ALDH activity has been linked to a poor clinical prognosis for cancer patients (18). Association of ALDHs biological activity with several diseases including cancer and the role of ALDH enzymes in DTP cells make it an attractive therapeutic target [34].

Drugs to target ALDH expressing persister cells.

Many drug development efforts have been undertaken to identify selective and broad spectrum ALDH inhibitors, which are summarized in Table 3. The structures of selected compounds targeting ALDH expressing persister cells, their target selectivity, and anticancer activities is provided in Table 6. Unfortunately, isoform specific inhibitors have failed clinically since targeting one family member can lead to its function being performed by another. Thus, the current realization in the field is that a broad-spectrum inhibitor will likely be needed for the clinic, but its efficacy would depend on the isoforms targeted and the resulting associated toxicity.

Table 6.

Drugs to target ALDH expressing persister cells.

SELECTED ALDH ISOFORM SPECIFIC INHIBITORS

NCT-506.

Quantitative high throughput screening led to the discovery of NCT-506, having a quinoline core (Compound 1; Table 6), (122) which displayed a potent ALDH1A1 inhibitory activity (IC₅₀ = 7 nM). NCT-506 was shown to inhibit the formation of 3D spheroid cultures of OV-90 ovarian cancer cells (IC₅₀ = 0.161 μM) and potentiate the cytotoxicity of paclitaxel in paclitaxel-resistant ovarian cancer cell line SKOV-3-TR, suggesting that this orally bioavailable compound could be combined with current existing anticancer drugs for efficacy. In addition, because of its selectivity for ALDH1A1, NCT-506 could serve as a probe for better understanding the physiological and pathophysiological actions of ALDH1A1.

Pyrazolopyrimidinone CM39 analog.

ALDH1A1 selective inhibitors (Compounds 2 and 3, Table 6) were identified by conducting a SAR study guided by a cocrystal structure of the high-throughput screening hit compound, a pyrazolopyrimidinone CM39, with ALDH1A1 (123), leading to inhibitors of the ALDH1A subfamily with excellent selectivity over ALDH2. Compounds 2 and 3, Table 6, depleted the CD133⁺ putative cancer stem cell pool, achieved efficacious concentrations in vivo following IP administration and synergized with cisplatin. Compound 3, Table 6, also synergized with cisplatin in a patient-derived ovarian cancer spheroid model. This study also discovered selective single isoform 1A2 and 1A3 inhibitors, however the selectivity was modest.

CVT-10216.

CVT-10216 (Compound 4, Table 6) is a selective reversible inhibitor of ALDH2 with an IC₅₀ of 29 nM compared to 1300 nM for ALDH1 (124). Inhibition of ALDH2 led to aberrant dopamine metabolism, which resulted in the formation of a metabolite that decreases dopamine synthesis and cocaine self-administration (125). ALDH2 inhibition by CVT-10216 significantly sensitized AML cells to standard chemotherapy, cytarabine (136). Recently, Wei et.al reported that the CVT-10216-mediated ALDH2 inhibition targeted the β-catenin signaling and reduced migration and stemness of colorectal cancer cells (137). CVT-10216 has been shown to suppress alcohol and cocaine addiction, as well as anxiety. Its anxiolytic properties are well established in several rodent models. However, no studies have examined its efficacy in cancer animal models.

CB7.

Through high-throughput screen, a highly selective inhibitor for ALDH3A1 has been identified, called CB7 (compound 5, Table 6). CB7 (1-((4-fluorophenyl)sulfonyl)-2-methyl-1H benzimidazole) (IC₅₀ of 0.2 μM) binds to the aldehyde binding pocket of ALDH3A1 and does not inhibit the activities of ALDH1A1, ALDH1A2, ALDH1A3, ALDH1B1, or ALDH2 up to 250 μM (126,127). The use of CB7 in combination with mafosfamide enhanced the antiproliferative effects of mafosfamide in the ALDH3A1 expressing cell lines A549 (lung adenocarcinoma) and SF767 (glioblastoma), but not of the primary lung fibroblasts (CCD-13Lu), which do not express ALDH3A1. Selective small molecule inhibitors of ALDH3A1, such as CB7, thus may have application in potentiating the sensitivity of certain chemotherapeutic drugs.

SELECTED BROAD-SPECTRUM ALDH INHIBITORS.

Dimate.

DIMATE (Compound 6, Table 6) was discovered by optimization of ampal thiolester (ATE), in which the amine group was replaced with N,N-dimethyl. ATE had low potency for inducing apoptosis in BCL2 overexpressing BAF3 cells (IC₅₀ of 250 μM) and DU145, a human prostate epithelial cancer cell line (IC₅₀ of 650 μM) cells(128). A SAR study based on the derivatization of ATE structure by modulation of the N atom and the thiolester moiety led to the identification of two compounds: DIMATE, having a N,N-dimethylamino group and MATE, having a morpholino substituent, which demonstrated 70 to 90- fold improvement in growth-inhibitory activity on DU145 cells compared to ATE (128). DIMATE was subsequently shown to be an active-enzyme-dependent, competitive, irreversible inhibitor of ALDH1. It inhibited the activity of both ALDH1 and ALDH3 in human prostate epithelial cancer cells DU145 in culture and induced apoptosis (129,130). Furthermore, DIMATE, targeting both ALDH1A1 and ALDH1A3, reduced melanoma tumor development while also eliminating slow cycling cells, which contained the reservoir of chemoresistant tumor cells (131). This suggests that inhibiting ALDH1A1 and ALDH1A3 together may provide a therapeutic opportunity for melanoma treatment. DIMATE showed cytotoxicity in NSCLC cell lines and demonstrated antitumor activity in NSCLC orthotopic xenografts via hydroxynonenal-protein adduct accumulation, GSTO1-mediated depletion of glutathione and increased H₂O₂ (138). A first-in-human, phase I clinical trial using DIMATE is currently underway to characterize the Safety, Tolerability, Pharmacokinetic and Preliminary Signs of Activity in leukemia(139) (AML) patients (NCT05601726, https://www.clinicaltrials.gov/ct2/show/NCT05601726), however, potential toxicity might limit efficacy.

DEAB.

DEAB (N,N-diethylaminobenzaldehyde)(compound 7, Table 6) is a potent inhibitor of cytosolic (class 1) aldehyde dehydrogenase (ALDH) and possesses all the characteristics of a substrate for many ALDH isoenzymes. Morgan et al. (95) showed the selectivity of DEAB for ALDH1A1, ALDH1A2, ALDH1A3, ALDH1B1, ALDH1L1, ALDH2, ALDH3A1, ALDH4A1 and ALDH5A1 isoenzymes. DEAB also acted as a substrate for ALDH1A1, ALDH1A3, ALDH1B1, and ALDH5A1, a covalent inhibitor of ALDH1A2, and neither a substrate nor an inhibitor for either ALDH1L1 or ALDH4A1. The DEAB treatment effectively sensitized colon and ovarian cancer cells to chemo-therapeutic drugs (140,141). The anticancer activity of DEAB is also shown in pancreatic cancer cells with reduced cell viability, increased cell apoptosis and gemcitabine sensitization (142). Moreover, DEAB can preferentially depleted CD133+ ovarian CSCs (143) and inhibited the proliferation of spheroids generated from patient-derived endometrial CSCs (144). Interestingly, Matsunaga et al. (132) found that administering DEAB during a time of day when ALDH activity was higher in 4T1 triple-negative breast cancer tumor cells increased its anticancer and antimetastatic effects, showing a function for the circadian clock within the tumor microenvironment in regulating the circadian dynamics of CSC.

Aldi-6.

Aldi-6 93-(dimethylamino)-4′ bromopropiophenone) (compound 8, Table 6) was identified through a high throughput screen of small molecules (133), and contained the same aryl propanone core structure as Aldi-1, 2 and 3 (145). The authors showed an overall ALDH activity increase with cisplatin treatment of head and neck squamous cell carcinoma (HNSCC) cells, with ALDH3A1 protein expression being particularly enhanced. To investigate if the inhibition of ALDH3A1 can enhance cisplatin toxicity, the authors used Aldi-6, which inhibits ALDH3A1, as well as ALDH1A1 and 2, with an IC₅₀ of 600-1000 nM. Aldi-6 inhibited cisplatin-induced ALDH3A1 in HNSCC resulting in a significant reduction in cell viability, with a more pronounced effect when Aldi-6 was used in combination with cisplatin. In vivo, Aldi-6 was well tolerated, and reduced tumor burden more effectively than cisplatin; the combination of Aldi-6 and cisplatin showing even better inhibition than the individual agents. These findings suggest that ALDH3A1 may contribute to cisplatin resistance in HNSCC and that targeting it using ALDH1A1, 2 and 3A1 inhibitors such as Aldi-6 could be a feasible treatment.

KS100.

An isatin derivative called Cpd3 (compound 9; Table 6) was identified as an effective and selective ALDH1A1 inhibitor (IC₅₀ = 0.02 uM)(134). The selective inhibition of ALDH1A1 occurred through direct interactions between the 3-carbonyl group of the isatin ring and cysteine residues within the active site and was accomplished by altering substitutions at the C5 and N1 positions of the isatin ring system. While compound 9 was a selective ALDH1A1 inhibitor, it did not inhibit cancer cells viability.

The structure of compound 9; Table 6, was optimized by keeping the core isatin skeleton intact and adding an isothiourea functionality at the para position of the benzyl moiety attached to dibromoisatin, which led the development of KS100 (compound 10, Table 6) (3). KS100 was found to be a multi-ALDH isoform inhibitor targeting ALDH1A1, ALDH2, and ALDH3A1 and inhibited cultured melaoma cell survival and tumor growth. The toxicity of KS100 was overcome by encapsulating it in a nanoliposomal formulation, which might be a solution for all broad spectrum ALDH inhibitors, such as DIMATE, to decrease toxicity and improve clinical utility (3).

FUTURE AREAS OF DTP CELL RESEARCH.

DTP cells significantly contribute to the complexity of cancer treatment resistance; therefore, multifaceted approaches are needed to target these cells and the mechanisms though which they drive the development of drug resistance. It is important to consider topics such as tumor microenvironment analysis (6), epigenetic profiling (6), advanced imaging techniques (6), and metabolic studies (6), to unravel the underlying mechanisms driving DTP cell mediated drug resistance. Identification of biomarkers (146) predictive of DTP cell presence and treatment response, coupled with combination therapies (5) that address tumor heterogeneity (6), offers promising avenues for personalized therapeutic interventions. Furthermore, leveraging artificial intelligence (AI) and computational modeling could enable predictions of DTP cell behavior needed to optimize treatment strategies. If successful, these integrated system wide approaches could lead to more effective and tailored strategies to combat DTP cells and treatment resistance in cancer. A summary of the importance of these systems in DTP cells and drug resistance are provided below.

Tumor Microenvironment Analysis (6) is important for understanding the dynamic interplay between DTP cells and surrounding microenvironment. Components within the tumor microenvironment, such as cancer-associated fibroblasts (CAF), and tumor-associated macrophages (TAM), are known to influence DTP behavior and response to therapy (6). Therefore, studying the tumor microenvironment allows for the identification of key factors contributing to DTP survival and persistence. Strategies to modulate the tumor microenvironment to render it less conducive to DTP survival, such as targeting angiogenesis or immune checkpoint pathways, hold promise for improving treatment outcomes.

Epigenetic Profiling (6) to identify epigenetic alterations playing a significant role driving DTP formation and maintenance (6). By profiling epigenetic modifications, such as DNA methylation, histone modifications, and non-coding RNA expression, researchers can identify epigenetic signatures associated with DTP cells (6). Understanding these signatures would provide insights into the regulatory networks controlling DTP cell phenotypes and identifies potential epigenetic targets for therapeutic intervention. Epigenetic therapies (147), such as histone deacetylase inhibitors (147) or DNA methyltransferase inhibitors (147), offer promising avenues for disrupting DTP cells persistence and sensitizing tumors to therapy.

Imaging Techniques and Single-cell Analysis can contribute to a better understanding of DTP in the contexts of complex multicellular tumor environments. Advanced imaging techniques, such as multiphoton microscopy (148), positron emission tomography (PET) (149), and magnetic resonance imaging (MRI) (150), coupled with single-cell analysis methods (151), would enable the precise identification and characterization of DTP cell subpopulations within heterogeneous tumor tissues. These approaches could enable scientists to visualize spatial and temporal dynamics of DTP cells, uncovering distribution within the tumor microenvironment and interactions with neighboring cells. Single-cell analysis techniques, such as single-cell RNA sequencing, would provide valuable insights into the transcriptional profiles and functional states of individual DTP cells (151), revealing heterogeneity and potential therapeutic vulnerabilities.

Metabolic Profiling(6) focusing on cell reprogramming is a hallmark of cancer, and DTPs often exhibit distinct metabolic profiles that contribute to survival and persistence(6). Metabolic studies are needed that aim to elucidate the metabolic pathways and adaptations employed by DTP cells to tolerate therapeutic insults. Investigating metabolic vulnerabilities of DTP cells could offer opportunities for therapeutic intervention, such as targeting specific metabolic pathways or nutrient dependencies. Metabolomics approaches, including mass spectrometry and metabolic flux analysis, would provide comprehensive insights into the metabolic landscape of DTP cells and guide the development of metabolic-targeted therapies.

Identification of Biomarkers (146) predictive of DTP cell presence and treatment response are critical for guiding personalized therapeutic strategies (146). Biomarker discovery efforts focused on identifying molecular signatures, such as gene expression profiles, protein markers, or circulating tumor DNA, associated with DTP cells are needed in future studies. These biomarkers could serve as prognostic indicators or predictive markers of treatment response, enabling clinicians to tailor therapy based on individual patient characteristics. Integration of multi-omics data and machine learning algorithms could enhance biomarker discovery and validation, facilitating translation into clinical practice.

Combination Therapies (5), which target multiple pathways or mechanisms involved in DTP cell survival and resistance, represent a promising strategy to overcome treatment resistance (5). Rational combination approaches that exploit synergistic interactions between different therapeutic agents or modalities, could lead to maximizing treatment efficacy while minimizing toxicity. Combination regimens may include conventional chemotherapy agents, targeted therapies, immunotherapies, or novel agents targeting specific DTP cell vulnerabilities. Preclinical and clinical studies evaluating combination therapies provide valuable insights into optimal drug combinations, dosing schedules, and patient selection criteria.

Tumor Heterogeneity (6), poses a significant challenge in cancer therapy, contributing to treatment resistance and disease recurrence(6). DTP cells arise from genetically diverse tumor cell populations, each with distinct molecular profiles and therapeutic vulnerabilities. Dissecting the clonal composition and spatial organization of DTP cells within the tumor ecosystem is essential for devising effective treatment strategies. Strategies to address tumor heterogeneity include combination therapies targeting multiple subpopulations, adaptive treatment approaches based on real-time monitoring of tumor evolution, and spatially resolved therapeutic interventions targeting specific tumor regions.

AI and Computational Modeling (152) could play a crucial role in predicting DTP cell behavior and optimizing treatment strategies (152). Machine learning algorithms analyzing large-scale omics data sets could be useful for identifying patterns, correlations, and predictive biomarkers associated with DTP cells. These computational models could be designed to simulate tumor growth, evolution, and response to therapy, thereby guiding the design of rational treatment regimens. Integrating AI-driven predictive DTP cell modeling into clinical decision-making processes might enhance treatment efficacy, minimize treatment-associated toxicities, and improve patient outcomes. For this approach to be effective, it will require collaborative efforts between computational biologists, bioinformaticians, and clinicians for developing and validating robust AI-driven DTP cell models for personalized cancer therapy.

CONCLUSION

Approaches are needed to target DTP cells to prevent the development of drug resistance and disease recurrence, which would reduce the long-term lethality of cancer drug resistance for patients. The following are some of the urgent issues that need to be addressed in the DTP cell field:

1. It is not clear whether targeting persister cell populations requires identifying all the persister cell subpopulations that can lead to recurrent drug resistant cancer and then developing drugs to target each separately or simultaneously (issue is observable in Table 1).

2. It is likely that not all DTP cell subpopulations have been identified. So, they would need to be identified, their mechanism of action unraveled, and then drugs developed to target them (issue is apparent in Table 1 comparing accepted versus putative DTP cell subpopulations).

3. No one has demonstrated that treatment with a drug that has potential to kill only a small tumor DTP cell subpopulation might be effective for inhibiting long-term cancer progression (issue is highlighted in Table 2 showing the involvement of DTP cell subpopulations innate versus acquired resistance).

4. No one has developed a broad-spectrum drug killing persister and non-persister cells that might be successful for managing drug resistance (agents targeting individual DTP cell subpopulations is shown in Table 3).

5. It is unclear whether targeting one or several persister cell subpopulations in a tumor would be necessary to prevent drug resistance (Table 4 illustrates the presence of multiple DTP cell population in a single tumor type).

6. No one has dissected the biology of DTPs having multiple persister cell markers and developed a strategy to target them (Table 4 shows the presence of multiple DTP cell markers in a single cancer type).

7. It remains unclear as to whether there might be an optimal strategy to target DTP cells expressing multiple persister cell markers in a single cell or in tumors that consist of heterologous persister cells (Table 4).

8. For those cancers having multiple different persister cells subpopulations (Table 4), it is unclear whether multiple drugs might have to be combined to manage drug resistance, and then associated toxicities would need to be managed.

9. It remains unclear whether a master persister cell subpopulation exists, and if so, the properties it might have. Furthermore, no one has conclusively determined whether ALDH DTPs should be considered as a master persister cell (Tables 4 and 6 show the drug KS100 as a potential broad-spectrum ALDH DTP cell marker that could act in this way).

10. The role of phenotypic DTP EMT switching, and drug resistance needs investigation. Specifically, it is not clear whether treatment with a drug lead to an EMT switch enabling the evolution of a different persister cell subpopulation that is not affected by the agent.

Concluding Remark:

The persister cells field is at a very early stage, but many promising discoveries have already been made to develop drugs to target these cell subpopulations in cancer. This area of research has become increasingly important for resolving cancer drug resistance and will remain of central importance to providing better management of long-term treatments.