Eradicating the tumor “seeds”: nanomedicines-based therapies against cancer stem cells

Abstract

Objectives

Cancer stem cells (CSCs), a small subpopulation of cells with high tumorigenesis and strong intrinsic drug resistance, exhibit self-renewal and differentiation abilities. CSCs play a crucial role in tumor progression, drug resistance, recurrence and metastasis,and conventional therapy is not enough to eradicate them. Therefore, developing novel therapies targeting CSCs to increase drug sensitivity and preventing relapse is essential. The objective of this review is to present nanotherapies that target and eradicate the tumor “seeds”.

Evidence acquisition

Evidence was collected and sorted from the literature ranging from 2000 to 2022, using appropriate keywords and key phrases as search terms within scientific databases such as Web of Science, PubMed and Google Scholar.

Results

Nanoparticle drug delivery systems have been successfully applied to gain longer circulation time, more precise targeting capability and better stability during cancer treatment. Nanotechnology-based strategies that have been used to target CSCs, include (1) encapsulating small molecular drugs and genes by nanotechnology, (2) targeting CSC signaling pathways, (3) utilizing nanocarriers targeting for specific markers of CSCs, (4) improving photothermal/ photodynamic therapy (PTT/PDT), 5)targeting the metabolism of CSCs and 6) enhancing nanomedicine-aided immunotherapy.

Conclusion

This review summarizes the biological hallmarks and markers of CSCs, and the nanotechnology-based therapies to kill them. Nanoparticle drug delivery systems are appropriate means for delivering drugs to tumors through enhanced permeability and retention (EPR) effect. Furthermore, surface modification with special ligands or antibodies improves the recognition and uptake of tumor cells or CSCs. It is expected that this review can offer insights into features of CSCs and the exploration of targeting nanodrug delivery systems.

Graphical abstract

Introduction

A total of 19.3 million new cancer cases and approximately 10.0 million cancer deaths occurred in 2020 [1], suggesting that anticancer treatment requires further improvement. Cancer stem cells (CSCs) are a subset of cancer cells characterized by their ability to self-renew, differentiate and drive tumor initiation, metastasis, drug resistance and relapse. CSCs were first discovered in 1994 in a study of human acute myeloid leukemia (AML), in which a population of CD34⁺CD38⁻ AML-initiating cells were identified in AML patients by transplanting into severe combined immune-deficient mice.These cells were engrafted in the mice and produced a large number of colony-forming progenitors [2]. In 2003, CD44⁺CD24⁻ lineage(-) tumorigenic breast cancer cells were isolated. As few as 100 CSCs with this phenotype were able to form tumors in NOD/SCID mice [3]. Additionally, CSCs have been successfully identified in brain tumors [4], prostate cancers [5], ovarian cancers [6] and colon cancers [7]. CSCs can drive tumor initiation, but their origin of is currently unknown. Several hypotheses for this issue have been summarized: (1) genetic or epigenetic changes in stem cells, progenitor cells and differentiated somatic cells; (2) dedifferentiation of normal cancer cells; (3) cell fusion; and (4) the effects of the tumor microenvironment (TME) [8–10].

CSCs are often quiescent and intrinsically drug resistant through the expression of drug efflux transporters, and the activation of antiapoptotic signaling pathways [11, 12]. Thus, conventional chemotherapy and radiotherapy usually fail to kill CSCs, often resulting in an increased CSC fraction in tumors. Thus, effective approaches to eliminate CSCs are quite necessary. With the development of nanotechnology, there are an increasing number of nanomedicines approved by FDA or in clinical or preclinical trials for cancer treatment. In regard to targeting CSCs, nanomedicines demonstrate many advantages. This review summarizes the hallmarks and surface markers of CSCs, and overviews the nanomedicine strategies to kill CSCs.

Hallmarks and surface markers of CSCs

In recent decades, CSCshave been defined as a small and unique subset of cells that exhibitself-renewal, differentiation potential, and intrinsic chemotherapy and radiotherapy resistance [13]. CSCs have acquired chemotherapy and radiotherapy resistance through multiple mechanisms, such as the upregulation of drug-efflux pumps (P-gp, ABCG2, MRP1, MDR1), a superior DNA-repair capacity, several activated survival signaling pathways,and enhanced protection against reactive oxygen species (ROS) [14–16]. In addition the plasticity of CSCs, especially their adoptionof quiescence, has been an important element of their drug resistance [17]. CSCs can release hypoxia-inducible factor 1 to induce the release of proangiogenic factors for tumor angiogenesis [18]. In addition, CSCs demonstrate epigenetic alterations such as histone modifications and DNA methylations, that contribute to tumor heterogeneity [19].

CSC surfaces, such as CD133, CD44, CD166 and EpCAM, play an important roles in the adhesion of the cells to their niche, [20]. ALDH1 is responsible for the oxidation of aldehydes into carboxylic acids and is related to tumorigenesis and metastasis in CSCs [21]. These cellular markers are usually applied to isolate CSCs from heterogeneous tumor cell populations (Table 1). CSCs share many of the same signaling pathways as normal stem cells, including Wnt, Notch and Hedgehog. Signaling pathways, such as TGF β along with PIK3/AKT, STAT or EGFR, are oncogenic cascades in CSCs [22–24]. CSCs are in a quiescent state and can efflux drugs out by highly expressed drug pumps, resulting in resistance to conventional treatment. However, the key CSC markers and signaling pathways demonstrate promise for targeted treatment. Due to in-depth investigations on the self-renewal, drug resistance, abnormal metabolism and embryonic signaling pathways in CSCs, much effort has been put forward for targeting CSCs. Based on the characteristics of CSCs, the possible therapies and strategies that can eliminate CSCs are summarized in Fig. 1.

Table 1.

Cancer stem cells markers [25–30]

Tumor types	CSCs markers
Breast cancer	CD44, CD24, ALDH1, CD133, ESA, PROCR, CD29, CD61,CD70, CD90, LGR5, CXCR4, EpCam, ProC-R, Sox-2, Nanog, OCT3/4
Colon cancer	ALDH1, CD44, CD24, CD133, LGR5, EpCam, Letm1, Sall4, Sox-2, Nanog
Glioblastoma	CD133, CD15, A2B5, ALDH, Sox-2, BMI1, Nestin, MUSASHI1, Nanog
Lung cancer	CD133, ALDH1, CD166, CD90, EpCam, Nanog, OCT4+
Liver cancer	CD24, CD90, CD13, ESA, OV6, CD133, ALDH, IB50-1, EpCam, Sall, AFP, Nanog, OCT3/4, Sox-2
Gastric cancer	CD44, Lgr5,CD24, CD90, CXCR4, LINGO2, ALDH, Letm1, Musashi2, Nanog, OCT3/4
Ovarian cancer	CD44, CD117, CD133, ALDH1, CD24, ROR1
Prostate cancer	CD44, integrin α2β1,CD133+, ALDH1+, FAM65B+ MF12 ,LEF1, Trop2
Leukemia	CD34, CD38, CD117
Melanoma	CD133, CD20,CD271

Fig. 1.

Possible therapies and strategies that can eliminate CSCs

Drugs targeting CSCs

Inhibitors targeting the signaling pathways in CSCs

The pathways affecting CSCs mainly include the Wnt, Hedgehog, Notch, BMP, Bmi, PI3K/Akt, integrin and STAT pathways [31]. Among them, the Wnt, Hedgehog and Notch signaling pathways have been thoroughly studied in CSCs and many inhibitors have entered clinical trials andhave been approved [32]. The Wnt signaling pathway regulates cell proliferation, differentiation, apoptosis and stemness. Many studies have found that tumor cells with high Wnt activity are functionally related to CSCs [33]. For example, as one of the markers of CSCs in cancers, CD44 expression is promoted by Wnt/β-catenin signaling and cytokines secreted from tumor-associated cells [34]. In addition to Wnt signaling, the Notch signaling pathway also plays an important role in CSCs. Wang R. et al. reported that Notch components (Notch1, Hes1 and Hey1) were upregulated in sphere-forming liver CSCs [35]. The Hedgehog pathway has also been shown to regulate the properties of CSCs in glioma, chronic myeloid leukemia (CML) and multiple myeloma, especially through the SMO gene [36]. Therefore, these signaling pathways are potential targets for CSCs. Several inhibitors in clinical trials targeting the Wnt, Hedgehog and Notch signaling pathways are summarized in Fig. 2; Table 2.

Fig. 2.

Chemical structures of several compounds targeting Wnt and Notch signaling pathways of CSCs

Table 2.

Several inhibitors targeting the Wnt, Hedgehog and Notch signaling pathways in clinical trials

Compound	Target	Tumor type	Phase state	Efficacy outcomes
LGK974 [37]	Wnt	Metastatic colorectal cancer	Phase I (NCT02278133)	N/A
ETC-159 [38]	Wnt	Solid tumor	Phase I (NCT02521844)	Recruiting
PRI-724 [39]	Wnt	Advanced pancreatic cancer	Phase I (NCT01764477)	Stable disease rate: 40%
MK-0752 [40]	Notch	T cell acute lymphoblastic leukemia	Phase I (NCT00100152)	Terminated
LY-900,009 [41]	Notch	Advanced cancer	Phase I (NCT01158404)	Stable disease in 5 patients from dose escalation group
RO4929097 [42]	Notch	Advanced-stage solid tumors	Phase I	Partial response in 1 of 96 evaluable patients, Stable disease in 28 of 96 patients
BMS-906,024 [43]	Notch	Advanced cancer	Phase I (NCT01986218)	Terminated
Nirogacestat [44]	Notch	1.Metastatic Cancer Pancreas 2.Desmoid tumors	Phase II (NCT02109445)	1.Terminated 2.Confirmed partial response in 5 patients, stable disease in anoth -er patients
Patidegib [45]	Hedgehog	Basal cell nevus syndrome; skin cancer	Phase II (NCT02762084)	Terminated
BMS-833,923 [46]	Hedgehog	Leukemia	Phase I/II (NCT01218477)	N/A
Taladegib [47]	Hedgehog	Advanced cancer	Phase I/II (NCT01226485)	Stable disease rate: 30.9%,Objective respo -nse rate: 26.2%

Classical drugs

Studies have found that some classical drugs possess anti CSC abilities. Metformin, a conventional drug for the treatment of type 2 diabetes, has been proven to exert potential antitumor effects and also appears to act on CSCs. A landmark study of the effect of metformin on CSCs was conducted by Hirsch et al. In their reserch, CD44⁺/CD24^−/low CSCs were eliminated after metformin treatment, and a synergistic antitumor effect was found with doorubicin [48]. Salinomycin, an antibiotic applied in veterinary medicine, has been isolated from the bacterium Streptomyces albus. It has been shown to eliminate CSCs in different types of tumors, such as human ovarian CSCs and breast CSCs. An increasing number of derivatives of salinomycin have been synthesized to have a higher selectivity for CSCs. The salinomycin C20-propargylamine derivative has shown a strikingly low IC₅₀ of 23 nM against HMLER CD24 ^low/CD44 ^high breast CSCs by sequestering iron in lysosomes, providing a highly selective agent to target CSCs [49]. Puromycin is another antibiotic that can suppress CSCs. Puromycin effectively eliminates CSCs through ribosomal protein translation inhibition in tumor spheres and monolayer cultures [50]. Nelfinavir, a human immunodeficiency virus (HIV) protease inhibitor, lowers self-renewal and induces apoptosis in CSCs,ultimately impairing CSC-induced allograft formation in vivo [51].

Targeting the abnormal metabolism in CSCs

CSCs may have highly glycolytic or oxidative phosphorylation phenotypes, depending on their surface markers andTME. CD44 variant isoforms (CD44v), a marker of CSCs, can interact with and stabilize cysteine-glutamate transporter, and promote cysteine uptake [52]. Breast CSCs shift glucose metabolism from mitochondrial oxidative phosphorylation to anaerobic by increasing the pyruvate kinase M2 isoform along with lactate dehydrogenase and glucose 6-phosphate dehydrogenase; additionally, 2-deoxyglucose combined with doxorubicin has been used to inhibit breast CSC proliferation [53]. CD9 promotes the plasma membrane localization of ASCT2, the major glutamine transporter, enhancing glutamine uptake in pancreatic ductal adenocarcinoma cells [54]. Glutamine deprivation significantly inhibits self-renewal, decreases the expression of stemness markers, increases ROS levels in pancreatic cancer stem cells, and finally enhances radiosensitivity [55]. CSCs are also highly dependent on lipid metabolism, especially fatty acid synthesis, lipid desaturation and β-oxidation. Therefore, there are different therapeutic methods for targeting CSCs based on lipid metabolism, mainly by targeting key enzymes including fatty acid synthase (FASN), acetyl‐CoA carboxylase‐1 (ACC), stearoyl CoA desaturase 1 (SCD1), and carnitine palmitoyl-transferase 1 (CPT1). Upregulated FASN has been observed to maintain stemness in CSCs. Once FASN is inhibited by cerulenin, the sphere formation, invasive abilities of CSCs and expression of stemness markers such as SOX2, nestin, CD133, and FABP7 were inhibited [56]. Other therapeutic molecules, such as resveratrol (FASN inhibitor), sorafenib (ACC inhibitor), and etomoxir (CPT1 inhibitor) can also target CSCs.

Problems of conventional drugs targeting CSCs

Although significant progress has been made in drugs targeting CSCs, several problems remain to be resolved. First, the effects of conventional targeting drugs can be hampered by drug properties, such as poor solubility and stability, unfavorable pharmacokinetics, lack of selectivity, and dose-limiting toxicities. Second, CSCs and normal stem cells share many similar properties ; thus, conventional drugs targeting CSCs can also destroy normal stem cells. Third, CSCs are more resistant than other tumor cells, leading to inefficient therapeutic outcomes. Therefore, more precise and efficient CSC targeting strategies need to be developed.

Nanomedicine-based strategies for CSC therapy

Nanocarriers for delivering small molecule drugs and nucleic acid agents

As mentioned above, some drugs have the ability to eliminate CSCs, including salinomycin (Sal), curcumin, disulfiram and chloroquine. However, these drug applications are limited by the physicochemical properties, pharmacokinetics, and stability of the drugs. Applications of Sal are limited by its poor water solubility. Sal-loaded nanomedicines have been developed to improve their pharmacokinetic properties, safety and tolerability. Wang Q et al. designed Sal-loaded gelatinase-responsive nanoparticles for the elimination of cervical cancer stem cells. The CD44-positive HeLa cell level was lower in Sal-loaded nanoparticle group than in the control and Sal solution group. In addition the nanoparticles-treated group significantly suppressed the regenerative ability of HeLa cells [57]. CD24 ^low/CD44 ^high breast CSCshave shown high sensitivity to Sal-loaded Gold nanoparticles by ferroptosis-induced cell death [58]. Sal is also used in combination with some classic anticancer drugs to eradicate both bulky tumor cells and CSCs. Sal and docetaxel have been encapsulated in PLGA/TPGS nanoparticles in a synergistic 1:1 ratio of to target both breast cancer cells and stem cells. These coloaded nanoparticles demonstratebetter inhibition of MCF-7 cells in vitro and in vivo. In addtion, the combination is more effective in inhibiting mammosphere formation of MCF-7 CSCs [59]. Kun Y et al. developed curcumin-loaded PLGA nanoparticles modified with sialic acid and anti-ALDH to permeate the blood-brain barrier (BBB) and target brain CSCs (Fig. 3) [60]. Wedelolactone, mainly a liver disease treatment, has been promise as an effective anticancer drug, but its effects on CSCs are unknown. Das S et al. found that wedelolactone-encapsulated PLGA nanoparticles enhanced drug uptake in breast cancer cells and in breast CSCs. The ALDH⁺ or CD44⁺/CD24⁻ population decreased after treatment with wedelolactone. In addition, it could increase the chemosensitivity of paclitaxel by downregulating SOX2 and ABCG2 [61]. Doxorubicin and thymoquinone coloaded cocktail shell-derived aragonite CaCO3 nanoparticles significantly reduce ALDH activity, sphere-formation and the percentage of CD44⁺CD24⁻ in breast cancer cells [62].

Fig. 3.

Anti-ALDH and Sialic acid co-modified PLGA nanoparticles for targeting brain CSCs. Adapted from [60]

In contrast to the combination of chemical drugs, the dual delivery of a hepatocyte nuclear factor 4 alpha (HNF4α)-encoding plasmid and cisplatin by mesoporous silica nanoparticles also exhibits excellent inhibition of pluripotency and tumorigenicity in CD133⁺ hepatoma cells [63] (Fig. 4).HNF4α has been proven to be impaired in hepatocellular carcinoma and is associated with poor differentiation, high metastatic potential, and poor prognosis. RNA nanotechnology has been widely applied to overcome the hurdles of gene delivery in vivo, such as rapid degradation, nonspecific distribution, and gene regulation. In the research of Yin H et al., the pRNA-3WJ motif was utilized as a core scaffold to deliver anti-miR21 and CD133 aptamers for targeting breast CSCs. The 3WJ/CD133_apt/anti-miR21 nanoparticles achieved a high accumulation in MDA-MB-231 and CSCs, and significantly increased PTEN and PDCD4 expression [64]. Carbamate-mannose modified PEI has been designed as a nonviral gene vector for NF-κB shRNA delivery. These nanocomplexes are able to decrease mammosphere formation, lower the ratio of ALDH⁺ breast CSCs, and inhibit migration and invasion. In addition, the above nanocomplexes sensitized doxorubicin-loaded micellar nanoparticles to treatment [65].

Fig. 4.

Mesoporous silica nanoparticles co-loaded cisplatin and HNF4α DNA plasmid to reduce the portion of liver CSCs. Adapted from [63]

Nanomedicinebased photothermal/ photodynamic therapy (PTT/PDT)

Photothermal or photodynamic therapy (PTT and PDT, respecitively) is a focus in cancer treatment. This treatment demonstrates higher selectivity, lower toxicity and better repeatability than traditional methods. PTT or PDT is also used to eradicate CSCs. Different types of photothermal agents are listed in Fig. 5. CD271 antibody-functionalized hollow gold nanospheres achieve specific targeting toward CD271⁺ osteosarcoma stem cells and induce the death of CSCs by apoptosis pathways and DNA injuries upon near-infrared laser irradiation [66]. Because CSCs preferentially reside in hypoxic regions and are centrally located in tumors, small nanoparticles more easily pass through the extracellular matrix for targeting CSCs. The Li YP group applied ferritin to form 26 nm nanocages for deep tumor penetration and binding to the transferrin receptor 1of CSCs. The DIR and epirubicin-loaded nanocages were preferentially accessible to CSCs in a 4T1 breast cancer model [67]. Fernandes S et al. revealed that monotherapy with iron oxide nanocubes could not eliminate quiescent CSCs, and could even restart division. A combination with doxorubicin-loaded thermally responsive nanoparticles caused a stop in quiescent-CSC activities [68]. In addition to chemotherapy combinations, immunotherapy, radiotherapy and PDT are also been reported (Fig. 6). According to Atkinson R, CSCs are more resistant to 6 grays of ionizing radiation, leading to an increased relative proportion after 48–72 h of ionizing radiation. However, a larger reduction in breast tumor size without an increase in the percentage of CSCs was achieved by local hyperthermia for 20 min at 42 °C after exposure to ionizing radiation using optically activated gold nanoshells [69]. Notably, PTT sensitized breast cancer stem cells to radiation therapy.

Fig. 5.

Different types of photothermal agents used for PTT

Fig. 6.

Combination of PTT with other antitumor therapies. Photodynamic therapy (PDT), Chemotherapy (CT), Immunotherapy (IT)

PDT kills tumor cells by generating ROS. PDT has been applied in the clinic for the treatment of various cancers over the past 40 years [70]. The effects of PDT during CSC treatment are under research. AlPcS4Cl-loaded anti-CD133 modified gold nanoparticles show a specific targeting toward lung CSCs. Adequate and continuous ROS exposure can eliminate CSCs [71]. Platinum metallacage-loaded nanoparticles have achieved chemo and photodynamic combination therapy. This platform can penetrate deeply into 3D tumor spheroids, and is able to eradicate liver CSCs, and decrease migration and spheroid formation [72]. In another report, salinomycin, photosensitizer chlorin e6 (Ce6) and vitamin E acetate were coencapsulated into a novel keratin-based nanoformulation to limit the self-renewal capacity and decrease the stemness of breast CSCs. Zebrafish embryo experiments in vivo have shown that the nanoformulations can interfere with the Wnt/β-catenin signaling pathway, which is an important pathway in CSCs [73]. Ginsenoside Rg3, a principal ginseng component, has presented significant anticancer activities and anti CSC stemness. Nanocarriers synthesized by graphene oxide (GO) linked with photosensitizer (PS) indocyanine green (ICG), folic acid, and polyethylene glycol (PEG) have been applied to encapsulate Rg3 to inhibit the progression and stemness of osteosarcoma [74].

Nanomedicines targeting the signaling pathways in CSCs

Nanomedicine-based CSC signaling pathway modulations include delivering small molecules to regulate signaling pathways, RNA interference nanotechnology and antibody-nanoparticle conjugates. For abnormal activation of the Wnt signaling pathway in various CSCs, Shamsian A et al. demonstrated that vorinostat (histone-deacetylase inhibitor) and PKF118-310 (Wnt-β-catenin pathway inhibitor) coloaded protein corona-capped gold nanoparticles reduced the stem cell population and Snail marker [75]. Cuprous oxide nanoparticles can also inhibit the stemness of prostate cancer cell lines and the Wnt signaling pathway [76]. Yallapu et al. developed curcumin-loaded PLGA nanoparticles to increase the chemo/radio sensitivity of cisplatinresistant ovarian cancer cells. These nanoparticles significantly reduce the β-catenin and c-Myc protein levels, which are important in CSCs [77]. The Notch signaling pathway is also highly activated in CSCs. In the research, α-mangostin-encapsulated PLGA nanoparticles have been used to inhibit the self-renewal ability, stem cell markers and pluripotency maintaining factors (Sox-2, KLF-4, c-Myc, Nanog) of CSCs by Notch signaling suppression [78]. SHP-1 siRNA-loaded ZnAs@SiO2 nanoparticles have been applied to eliminate hepatocellular carcinoma cells and impede recurrence by the SHP-1/JAK2/STAT3 signaling pathways. These nanoparticles inhibit tumor growth and induce changes in stemness markers (CD133, Sox-2 and OCT-4) and EMT markers (E-cadherin, Vimentin and Slug) [79]. Kaushik NK and colleagues found that cotreatment with low doses of PEG coated gold nanoparticles and cold plasma not only inhibited the proliferation of tumor cells but also inhibited EMT behavior, decreasing sphere formation and the self-renewal capacity of glioma stem cells by abolishing the PI3K/AKT signaling pathway [80].

Nanocarriers targeting the specific markers of CSCs

Specific surface markers are not only separation markers for CSCs, but also therapeutic targets. CD44, CD13, CD36, CD24, CD133, CD20 and CD90 are commonly reported as nanomedicines-targeting biomarkers [81]. Cho JH et al. designed CBP4, a novel small peptide, to target CD133. CBP4 conjugated gold nanoparticles exhibited “signal on-off” properties by quenching depending on the presence of CD133⁺ glioma CSCs, which improved imaging effects when CSCs were present [82]. Salinomycin-loaded PEG-PLGA nanoparticles conjugated with CD133 aptamers have been applied to target CD133⁺ osteosarcoma CSCs. Results showed that CD133 aptamer-decorated nanoparticles are 4.92 or 2.33-times more effective than undecorated nanoparticles or salinomycin in CD133⁺ Saos-2 cells, respectively [83]. CD44 was firstly described in hematopoietic stem cells and then in cancer and leukemia initiating cells [84]. Hyaluronic acid (HA) is a common CD44 specific ligand and has been utilized for active targeting in tumor therapy. Su Z et al. constructed anti-CD44 antibody-modified superparamagnetic iron oxide nanoparticles to kill human oral squamous cell carcinoma stem cells. Results demonstrated that CD44 nanoparticles induced CD44 -overexpressed CSCs to undergo programmed death [85]. Albumin nanoparticles functionalized with HA exhibit high affinity to CD44-enriched B16F10 cells. Moreover, HA-eNPs/ATRA treatment significantly decrease the side population of B16F10 cells in vitro. Finally, tumor growth has been significantly inhibited by HA-eNPs/ATRA in lung metastasis tumors in mice [86]. It is recognized that CSCs also have different phenotypes; thus, it is more efficient to target two surface markers instead of one. CD44 and CD133 are both gastric CSC markers. Chen H and colleagues developed CD44 and CD133 antibody-conjugated all-trans retinoic acid-loaded PLA-lecithin-PEG nanoparticles for targeting gastric cancer stem cells. These dual antibody decorated nanoparticles showed better inhibitory effects than the single targeting and nontargeting ones [87]. Sal-loaded lipid-polymer nanoparticles decorated with anti-CD20 aptamers have shown selective cytotoxicity to CD20⁺ melanoma CSCs, as proven by suppressing the formation of tumor spheres and decreasing the proportion of CD20⁺ melanoma cells [88]. To reduce off-target effects and improve safety, several smart strategies to hide the target site during circulation and expose the tumor site for CSC targeting should be further investigated.

Exosome-based nanomedicines for CSCs

Exosomes are small extracellular vesicles secreted by mammalian cells with a diameter of approximately 100 nm. Exosomes deliver various bioactive molecules, such as nucleic acids, proteins and lipids to the parent or distant cells [89]. Compared to conventional synthetic nanoparticles and liposomes, exosome-based delivery systems have high biocompatibility and stability, long circulation time, easy modification, low immunogenicity, and low toxicity [90–92]. Exosomes can be fabricated to target CSCs. In one study, luminescent porous silicon nanoparticles (PSiNPs) loaded with doxorubicin (DOX@PSiNPs) were used. Human hepatocarcinoma Bel7402 cells were incubated with PSiNPs for 6 h and then incubated in nanoparticle-free medium. The results showed that PsiNPs were exocytosed from Bel7402 cells in a time-dependent manner to induce autophagy and could be obtained through centrifugation. Exosome-sheathed DOX-loaded PSiNPs (DOX@E-PSiNPs) were also obtained in a similar way. DOX@E-PSiNPs demonstrated excellent antitumor and anti-CSC efficacy [93] (Fig. 7).

Fig. 7.

Schematic illustration of exosome-sheathed porous silicon nanoparticles (PSiNPs) as nano-carriers for CSCs targeting delivery. Permission by Creative Commons CC BY [93]

Nano-immunotherapy for CSCs

Immune cells that infiltrate in tumors are expected to eradicate cancer cells. However, CSCs alone or with other noncancerous cells can lead to an immunosuppressive TME [94]. CSCs or stromal cells interfere with DC recruitment to the tumor site, impair maturation and improve the formation of immunoregulatory DCs (DCregs) [95]. The crosstalk between regulatory T cells and CSCs also provides an immunosuppressive TME [96]. Therefore, immunotherapy targeting CSCs is a key route to improve antitumor efficacy.

In recent years, extensive preclinical studies and initial clinical data have demonstrated that nanotechnology can overcome some of the challenges that currently limit cancer immunotherapy [97]. ALDH is a surface marker for isolating and identifying CSCs. Najafabadi A et al. developed a synthetic high-density lipoprotein nanodisk vaccination to generate ALDH-specific T-cell responses against ALDH⁺ CSCs based on ALDH1-A1 and ALDH1-A3 epitopes. The nanodisk vaccination exhibited potent antitumor efficacy in combination with anti-PD-L1 therapy in vivo (Fig. 8) [98]. In another study, sialic acid-modified doxorubicin-loaded liposomes were able to modulate the immunosuppressive TME in multiple ways, including inhibiting tumor-associated macrophage (TAM)-mediated immunosuppression and promoting the infiltration of CD8⁺ T cells. The combination of liposomes with anti-PD-1 therapy almost completely eliminated B16F10 tumors and effectively inhibited 4T1 tumors. To eradicate breast CSCs, metformin was incorporated into the above combination treatment, and the synergistic effects were beneficial in 4T1 tumors [99]. A cocktail nanodevice composed of paclitaxel, thioridazine (anti-CSC agent) and HY19991 (PD-1/PD-L1 inhibitor) coloaded enzyme/pH dual-sensitive micelles has been designed to reduce the proportion of CSCs and enhance T-cell infiltration [100]. As reported, nanoimmunotherapy can target CSCs and further improve the effects of immunotherapy.

Fig. 8.

Schematic illustration of nanodiscs against ALDH^high CSCs. Adapted from [98]

Nanomedicines targeting the abnormal metabolism in CSCs

The abnormal metabolism of CSCs provides an opportunity for CSC targeting. Shen Y et al. constructed a novel paclitaxel loaded liposome (Nano-Taxol) to investigate its effects on stemness and metabolic reprogramming in paclitaxel-resistant ovarian cancer. The results showed that the intraperitoneal delivery of Nano-Taxol effectively suppressed CSCs and redirected metabolism from glycolysis to oxidative phosphorylation [101]. Nano-realgar has been used for anticancer treatment. As reported, Nano-realgar inhibits lung cancer stem cell viability by inducing glucose metabolism, and downregulating the expression of the HIF-1α and PI3K/Akt/mTOR pathways [102].

Conclusion and perspectives

.CSCs are often quiescent and intrinsically drug-resistant through the expression of drug efflux transporters, and the activation of antiapoptotic signaling pathways. Conventional therapy often fails to eradicate them. Nanoparticle drug delivery systems (NDDSs) are very attractive carriers for the delivery of drugs to tumors. In our review, nanotechnology-based strategies for CSC targeting are proposed, include (1) encapsulating small molecular drugs and genes by nanotechnology, (2) targeting CSC signaling pathways, (3) utilizing nanocarriers targeting for specific markers of CSCs, (4) improving PTT/PDT, (5) targeting the metabolism of CSCs and (6) enhancing nanomedicine-aided immunotherapy.

Although nanomedicines have successfully been approved by the FDA for cancer treatment and more nanomedicines are in human clinical trials, the application of nanomedicines in anti-CSC therapies is at a relatively early stage. In designing rational nanomedicines for CSC targeting, the following issues could also be addressed: (1) CSCs and normal stem cells share many signaling pathways; therefore, the utilization of inhibitors or siRNA of signaling pathways can also cause the destruction of normal stem cells. Although loading by nanotechnology can reduce toxicity, it is more efficient to choose the abnormal activation of signaling pathways for targeted therapy of different CSCs. (2) A median of 0.7% of a nanoparticle injection dose arrives at solid tumors, and this low value contributed to the poor clinical translation of nanomedicines [103]. Ouyang B et al. found a dose threshold for effective nanoparticle delivery. There are a small number of CSCs in tumors. Thus, for CSC recognition and effective drug delivery, there may be a minimum dose threshold that is worth exploring. (3) CSCs are “cold” seeds, namely in a resting state with low proliferation ability and inherent therapy resistance. Remolding these “cold” CSCs to “hot” tumor epithelial cells with high proliferation ability and active telomerase activity before or during nanomedicine treatment may be a good strategy to enhance the killing effects.

Author contributions

Lin Li, Rui Ni and Lin Chen contributed to the idea for this review article. The literature search and analysis were performed by Dan Zheng. Further evidences collection, analysis and arrangement were accomplished by Lin Li and Rui Ni. The first draft of the manuscript was written by Lin Li and Rui Ni. Then, the manuscript was critically revised by Lin Chen (Corresponding Author). All authors read and approved the final manuscript.

Declarations

Conflict of interest

The authors confirm that this article content has no conflicts of interest.