Biomarkers for CSC Populations in Solid Cancers In solid cancers, the clinical use of CSC specific biomarkers is very limited, besides the use of the carcinoembryonic antigen (CAE), fragments of the cytokeratin 19 (YFRA 21-1) (58) and the alpha-fetoprotein (AFP) that is expressed by cancer stem cells (58, 59). Importantly, most markers expressed in CSCs can also be found in adult tissue resident stem cell populations, human embryonic stem cells (hESC) or adult tissues (60). Additionally, most markers label heterogeneous stem cell populations pointing to the fact that their characterization and isolation has to be based on marker combinations using several surface markers or combinations of extracellular as well as intracellular markers; to identify and isolate cells that promote tumor initiation, resistance and relapse. Below, a short summary of the most prominent markers is provided. CSC markers that could have potential usefulness within therapeutic, diagnostic, and prognostic approaches are pointed out (compare Tables 1–7) and focus on most deadliest tumors of lung, liver, breast, stomach, and colorectal as well as AML and CML. Tables 1–7 provide an extensive list of markers expressed in CSCs. A comparison shows that several markers are expressed in several tumor types. Table 1 Examples of lung cancer stem cell markers and their use as diagnostic, predictive, or therapeutic biomarkers. Marker Stem cell marker Biomarker diagnostic Biomarker therapeutic Biomarker prognostic Surface markers, CD CD44 (and its variants) (61–66) (39, 67–69)* (70)** (71) (71–80) (61, 64, 70, 81) (39, 69)* CD87 (82) CD90 (83) (39, 67)* CD133 (84–99) (39, 67–69)* (70, 100)** (74, 101–104) (69)* (91, 105–112) (39, 67, 68)* (70)** CD166 (62, 66, 113) (39, 68)* (113) Surface markers, not CD EpCAM (62, 66, 86, 114, 115) (116–120) (121) (117, 122–124) Intracellular markers ALDH (65, 84, 114, 125–129) (39, 68, 69, 130)* (131) (132–134) (62, 128, 135) (39, 69, 130)* (70, 126)** Nanog (70) (70, 126) (69)* Oct-3/4 (96) (67, 69)* (67)* (136) (69)* The table lists examples of cancer stem cell markers and indicates those which have been tested as biomarkers within a therapeutic (metastasis, tumor stage, size), diagnostic, or prognostic (survival, resistance etc.) approach. Starsindicate reviews (*) and contradictory results (**). Table 2 Examples of breast cancer stem cell markers and their use as diagnostic, predictive, or therapeutic biomarkers. Marker Stem cell marker Biomarker diagnostic Biomarker therapeutic Biomarker prognostic Surface markers, CD CD24 (137) CD29 (ß1 integrin) (137, 138) CD44 (and its variants) (139–149) (150–154) (76, 150, 152, 154–166) (166–171) (172, 173)** CD49f (174–176) (177)* (178) (175, 178, 179) CD61 (137, 180) CD70 (181) CD90 (182) CD133 (183) (184)* (185–187) (188–190) (184)* (191–193) (184)* Surface markers, not CD CXCR4 (194) EpCAM (186) (186) LGR5 (195) (195) ProC-R (196) Intracellular markers ALDH (147, 148, 197, 198) (199, 200)* (198, 201, 202) (199)* (171, 192, 197, 203–208) (200)* (209, 210)** BMI-1 (143, 211–218) (219)* Nanog (142) (220, 221) Notch (222–224) (222, 225) (187, 212, 222, 224, 226–230) (222, 226, 231–234) (235)* Oct-3/4 (142) (220, 221) Sox2 (142) Signaling pathways Wnt/ß-Catenin (195, 236, 237) (236) (237) The table lists examples of cancer stem cell markers and indicates those which have been tested as biomarkers within a therapeutic (metastasis, tumor stage, size), diagnostic, or prognostic (survival, resistance etc.) approach. Stars indicate reviews (*) and contradictory results (**). Table 3 Examples of gastric cancer stem cell markers and their use as diagnostic, predictive, or therapeutic biomarkers. Marker Stem cell marker Biomarker diagnostic Biomarker therapeutic Biomarker prognostic Surface markers, CD CD24 (238) (239)* (240)* (241) (242–244) (239, 245)* CD44 (and its variants) (246–251) (239, 240, 245, 252)* (247, 251, 253, 254) (240)* (255–257) (239, 240, 245)* (247, 251, 254, 258–260) (239, 240, 245, 252)* CD90 (251) (239, 245)* CD133 (247, 249, 250) (239, 240, 252)* (254, 261) (240)* (257) (239, 240)* (254, 262–265) (239, 240, 252)* Surface markers, no CD CXCR4 (266) (267)* (268–271) EpCAM (248, 249, 272) (239, 240, 252)* (273) (265, 272) LGR5 (274) (252)* (240)* (275, 276) (252)* (275, 277–279) LINGO2 (280) (280) Intracellular markers ALDH (249, 281, 282) (239, 240, 252)* (260, 281, 282) Letm1 (283) (283) Musashi2 (284) (284) Nanog (285) (239, 286)* (287) (240)* (287, 288) (286)* Oct-3/4 (239, 252)* (289)** (247, 265, 288) (289)** Sox2 (247) (239, 240, 252, 290)* (240)* (291)** (292) (247, 288, 293) (265)** The table lists examples of cancer stem cell markers and indicates those which have been tested as biomarkers within a therapeutic (metastasis, tumor stage, size, resistance), diagnostic (i.e., resistance), or prognostic (survival, resistance etc.) approach. Stars indicate reviews (*) and contradictory results (**). Table 4 Examples of liver cancer stem cell markers and their use as diagnostic, predictive, or therapeutic biomarkers. Marker Stem cell marker Biomarker diagnostic Biomarker therapeutic Biomarker prognostic Surface markers, CD CD24 (294–296) (297, 298)* (298)* (295) CD44 (299, 300) (298)* (300–303) (298)*(304)** CD90 (295, 300, 305–308) (297, 298)* (295, 300, 304, 309) (298)* CD133 (295, 296, 300, 310–313), (297, 298)* (314) (295, 300, 304, 311, 314–319), (320)**, (298)* Surface markers, not CD EpCAM (297, 298)* (294, 300, 304, 311, 321) (322) (298)* (300, 301, 304, 311, 319, 321–327) (298)* Intracellular markers and pathways AFP (311, 321) (328) (311, 321, 329), (330)* Nanog (312, 313, 331), (298)* (298)* (331) (298)* Notch (295, 296, 305) (295) Oct-3/4 (313, 331), (298)* (309, 331), (298)* Sox2 (313) (298)* Wnt/ ß-catenin (295, 313) (295) (313) (330)*, ** The table lists examples of cancer stem cell markers and indicates those which have been tested as biomarkers within a therapeutic (metastasis, tumor stage, size, resistance), diagnostic (i.e., resistance), or prognostic (survival, resistance etc.) approach. Stars indicate reviews (*) and contradictory results (**). Table 5 Examples of colorectal cancer stem cell markers and their use as diagnostic, predictive, or therapeutic biomarkers. Marker Stem cell marker Biomarker diagnostic Biomarker therapeutic Biomarker prognostic Surface markers, CD CD24 (332) CD44 (333–335) (336)* (337, 338) (339) CD133 (334, 340) (336)* (340) (338, 341–343) (340, 344) CD166 (333) (336)* (333) Surface markers, not CD EpCAM (335) (336)* (345, 346) (347)* LGR5 (335, 348–350) (336)* (351) (352) (353, 354) Intracellular markers ALDH (335, 355, 356) (336)* (355) (357)* Letm1 (358) (358) Nanog (359, 360) (336)* (361) (361, 362) Oct-3/4 (363, 364) (336)* (363, 365) Sall4 (366) (366) Sox2 (359, 367, 368) (336)* (367–369) The table lists examples of cancer stem cell markers and indicates those which have been tested as biomarkers within a therapeutic (metastasis, tumor stage, size, resistance), diagnostic (i.e., resistance), or prognostic (survival, resistance etc.) approach. Starsindicate reviews (*). Table 6 Examples of AML cancer stem cell markers and their use as diagnostic, predictive, or therapeutic biomarkers. Marker Stem cell marker Biomarker diagnostic Biomarker therapeutic Biomarker prognostic Surface markers, CD CD33 (370) (371–392) (393) CD123 (370, 394–396) (395, 397–399) (373–376, 397, 400–412) (394, 399, 403, 413) Surface markers, not CD CLL-1 (414–416) (370) (414, 417–419) (415, 420) TIM3 (421) (422) (420, 423) Intracellular markers ALDH (424) (424, 425) Nanog (426) (427) (426) Oct-3/4 (428) (429) (429–431) Sox2 (431, 432) The table lists examples of cancer stem cell markers and indicates those which have been tested as biomarkers within a therapeutic (metastasis, tumor stage, size, resistance), diagnostic (i.e., resistance), or prognostic (survival, resistance etc.) approach. Table 7 Examples of CML cancer stem cell markers and their use as diagnostic, predictive, or therapeutic biomarkers. Marker Stem cell marker Biomarker diagnostic Biomarker therapeutic Biomarker prognostic Surface markers, CD CD25 (433–437) (438–440)* (439)* (441) CD26 (433–437, 442–445) (438–440)* (443, 446) (439)* (434, 447, 448) (443) CD33 (433, 434) (438–440)* CD36 (434, 435) (438)* (435) CD117 (433, 434, 437) (439, 440)* CD123 (434, 449–451) (439, 440)* (449, 450) Surface markers, not CD IL1RAP (433–437, 452, 453) (438–440)* (439)* (452, 453) (437) Intracellular markers JAK/STAT (433) (438)* Wnt/β-catenin (454–456) (438, 457)* (454, 458, 459) (457)* FOXO (460) (438)* (460) Hedgehog/Smo/Gli2 (461) (438)* (461) The table lists examples of cancer stem cell markers and indicates those which have been tested as biomarkers within a therapeutic (metastasis, tumor stage, size, resistance), diagnostic (i.e., resistance), or prognostic (survival, resistance etc.) approach. Stars indicate reviews (*). CD44 CD44 is a biomarker which is not only expressed in solid but also in hematological cancers (see below). Its expression is associated with increased proliferation, self-renewal and metastasis (61, 149, 462, 463). For example, in colorectal cancers, expression of CD44/CD166 characterizes a cell population able to form tumor spheres, suggesting anchorage-independent proliferation of these cells (333). In other studies, CD44high/CD133high cells showed increased tumorigenic capabilities (334). Also in breast cancers, the percentage of CD44+/CD24−/CK+/CD45− cells was shown to be increased in malignant lesions compared to non-malignant lesions (139). A significant decrease in proliferation and migration of breast cancer cells was observed after the knock-down of CD44 (140). In gastric cancers, the knock-down of CD44 reduced sphere formation and caused decreased tumor growth in severe combined immunodeficiency mice (246). In many tumors (e.g., breast and liver), CD44 is expressed as isoform and its expression has been associated with increased cancer stem cell properties (141). In lung cancers, CD44v9 expression correlates significantly with early-stage lung adenocarcinoma and epidermal growth factor receptor (EGFR) mutations (464). Variants of CD44 are also expressed in gastric cancers and promote tumor initiation (248). The CSC marker CD44 has been indicated as a biomarker for diagnostic, therapeutic, and prognostic approaches (compare Tables 1–5). In gastric cancer patients, CD44+ circulating tumor cells correlated with a poor prognosis (465). In colorectal cancers, a prognostic quantitative real-time PCR was established to analyze the expression of CD44v2 showing that a high expression correlated with a worse prognosis (339). In gastric cancers, the expression of CD44 and CD90 correlated with distant metastasis and could therefore be used as a diagnostic biomarker (251) and was suggested as a biomarker for treatment response (253). Therapeutic approaches targeting CD44 have been made using e.g. adenoviral delivery of siRNA in vitro (337). Furthermore, CD44-targeting drug conjugated aptamers (76) or hyaluronic acid coated onto nanoparticles have been in the focus of research (155). Antibody-based photosensitizer conjugates for combined fluorescent detection and photo-immunotherapy (PIT) of CD44-expressing cells in triple-negative breast cancers (TNBC) (150) or other antibody-based approaches tested in safety studies (466–468). CD133 The biomarker CD133 (Prominin-1) is expressed on hESCs and rarely found on normal tissue cells (60). The marker has been additionally identified in tumors of breast, liver, stomach, and colon (compare Tables 1–5) and has also been described as a marker that characterizes cells with high tumorigenicity and a high ability to form spheroids (184, 469). In breast cancers, its expression correlates with N-cadherin expression that was found to be significantly higher in patients with metastasis (191). In lung cancers, the expression of CD133 has been correlated to epithelial to mesenchymal transitions (EMT), in combination with other additional stem cell markers, such as BMI1 (84). The expression of CD44 and CD133 in colorectal cancers can predict metastasis (470), however, no correlation to patient outcome could be detected (471). In breast cancers, CD133 mRNA was suggested to be suitable for prognosis prediction (193, 472) and CD133 protein has been correlated to a poor prognosis (193). Pre-clinical therapeutic approaches cover antibody-based targeting of colorectal (341, 342) as well as breast cancers (188) (compare Tables 1–5). EpCAM The epithelial cell adhesion molecule (EpCAM, CD326) is expressed on CSCs in various tumor types including colon and hepatocellular cancers (473–476). Furthermore, it is expressed in non-transformed tissues such as epithelial cells (476), and various stem and progenitor cells (477, 478). EpCAM is involved in proliferation and differentiation as well as in cell signaling and formation and maintenance of organ morphology (479). In cancer tissue, EpCAM is homogeneously expressed on the cell surface, while in epithelia it is expressed on the basolateral side (476). In breast cancers, the expression of EpCAM is correlated to CSC-like phenotypes that promote formation of bone metastases in mice (480). In lung cancers, the expression of EpCAM is often associated with the expression of CD44 and CD166. Triple positive cells show increased clonogenicity, spheroid formation, self-renewal capacity, and show increased resistance to both 5-fluouracil and cisplatin (62). As one of the first CSC markers, EpCAM has been evaluated as a therapeutic biomarker (compare Tables 1–5). Targeting EpCAM with different antibody formats has been performed in colorectal as well as breast cancers (347). In colorectal cancers, a therapeutic approach targeting EpCAM+ cells with aptamers has been performed in pre-clinical conditions (345, 346). Intracellular Biomarkers as Regulators of Stemness in Solid Cancers Both embryonic and CSCs show unlimited growth, invasive capacity and are characterized by an undifferentiated cellular state (481). This feature depends on transitions between epithelial and mesenchymal states, regulated by a network of intracellular pluripotency transcription factors. As reviewed by Hadjimichael et al. and also described by others pluripotency in ESC is regulated by a core-network of transcription factors, consisting amongst others of Oct-3/4, Sox2, Nanog, Klf4, and c-MYC as well as signaling pathways such as the Jak/Stat, Wnt/ß-catenin, Hedgehog/Notch, TGF-ß as well as FGF signaling pathways (367, 482, 483). The core-pluripotency network consisting of Nanog, Oct-3/4 and Sox2 (described in detail below) activates genes of self-renewal and suppresses genes involved in differentiation (482). Pluripotency factors as well as signaling pathways have been indicated as biomarkers for CSCs as shortly described below (compare Tables 1–5). Of note, the tables do not include all biomarkers, however describe the most abundant ones reported in the literature. Sox2 The transcription factor Sox2 belongs to the SRY-related HMG-box (SOX) family, and is involved in the maintenance of an undifferentiated cellular phenotype (367). Its aberrant expression in cancers often leads to increased chemotherapy resistance and asymmetric divisions, as observed in colorectal cancers (368). In those, Sox2 expression correlates with a stem cell state and with a decreased expression of the caudal-related homeobox transcription factor 2 (CDX2), which could serve as a prognostic marker for a poor prognosis (367, 368). In gastric cancers, expression of Sox2 correlates with the tumor stage as well as with a poor prognosis (247, 288). The formation of tumor spheroids in vitro also correlates to the overexpression of CD44 and CD133 as well as the transcription factors Sox2, Nanog and Oct-3/4 (247). However, in another study, Sox2 levels were downregulated in gastric cancers in comparison to normal tissue and high Sox2 expression correlated with decreased metastasis and a better prognosis for the patient due to increased p21 levels (293). Therefore, the oncogenic functions of Sox2 are controversially discussed in gastric cancers, in which Sox2 might also have tumor-suppressor functions. These different functions seem to depend on the cancer origin and cellular context (484). Oct-3/4 Oct-3/4, also known as POU5F1, belongs to the POU homeobox gene family and is also a regulator of pluripotency in mammalian stem cell population. Oct-3/4 is upregulated in several cancers and may support the neoplastic transformation and resistance (485). In colorectal cancer cells, Oct-3/4 causes increased migration and liver metastasis (363, 486) correlating with poor survival (365). As reviewed by Prabavathy et al. Oct-3/4 expression is correlated to increased self-renewal and metastasis in lung cancer cells (67). A meta-analysis showed that Oct-3/4 expression in lung cancer was associated with poor outcomes concerning the differentiation degree, the TNM Classification of Malignant Tumors (TNM) and lymphatic metastasis (136). In hepatocellular carcinoma (HCC) Oct-3/4 expression was correlating with tumor size and recurrence (309). Nanog Nanog is a homeobox domain transcription factor widely expressed in human cancers (487). In colorectal tumors its expression was significantly increased in CD133+ cells, and on the basis of a univariate analysis, Nanog expression correlated linearly to liver and lymph node metastasis and the TNM stage. It might therefore be useful as a prognostic biomarker in post-operative liver metastasis (362). In breast cancer, expression of Nanog and Oct-3/4 has been correlated to a poor prognosis of the patient as well as EMT (220, 221). In HCC cell lines Nanog expression drives selfrenewal and invasion, metastasis, and drug resistance (298).