Nanoparticle delivery of anticancer drugs overcomes multidrug resistance in breast cancer

Abstract

Breast cancer is a serious threat to women's health, because multidrug resistance (MDR) has hampered treatment and prognosis. Nanodelivery of anticancer agents is a new technology to be exploited in the treatment of patients, because it bypasses multispecific drug efflux transporters such as P-glycoprotein (ABCB1), multidrug resistance protein-1 (MRP1, ABCC1) and breast cancer resistance protein (BCRP, ABCG2). Drugs can be delivered to tumor tissue by passive and active tumor targeting strategies, which may reduce or reverse drug resistance. This review will mainly focus on MDR-associated proteins, as well as various nanoparticle formulations developed to overcome MDR in breast cancer.

Keywords:

• Breast cancer
• mechanisms
• multidrug resistance
• nanoparticles

Introduction

Breast cancer is one of the most common cancers in women worldwide, and the incidence has been increasing over the past 20 years in most countries (Cecchini et al., 2015; Scalia-Wilbur et al., 2016). Chemotherapy is a mainstay in the treatment of breast cancer. However, tumor chemotherapy not only acts on tumor cells, but also exerts toxicity on non-cancer cells, and often results in rapid development of acquired drug resistance in tumor cells, particularly by upregulation of multidrug resistance (MDR) transporters. MDR is a phenomenon of cross resistance to multiple drugs with different structures, targets and functional mechanisms (Perez-Tomas, 2006; Ullah, 2008). At present, it is believed that MDR is the most important self-protection mechanism of tumor cells and also the most important reason for the failure of chemotherapy. In recent years, nanotechnologies have made considerable contributions to control drug release and to overcome MDR. Nanodrug delivery systems refer to technologies that deliver drugs by loading them into nanoliposomes, polymer micelles, inorganic metal nanoparticles or other nanomaterials, with a particle size between 1 nm and 1 μm. This review will mainly focus on recent progress of nanoparticle-mediated anticancer drug delivery in breast cancer and highlight recent progress in the field.

MDR in breast cancer

MDR mediated by drug efflux

The most frequent cause of tumor cells resistance to multiple drugs is overexpression of active drug efflux transporters, which intercept drugs at the level of cellular membranes. These are either located at the plasma membrane or alternatively at the nuclear membrane. This results in inadequate drug concentrations at cellular target structures and reduction of cancer cell toxicity. The most extensively studied MDR transporters include P-glycoprotein (P-gp/ABCB1), multidrug resistance protein-1 (MRP1/ABCC1), breast cancer resistance protein (BCRP, ABCG2) and lung resistance protein (LRP). P-gp, MRP1 and ABCG2 are members of the human family of ATP-binding cassette (ABC) transporters.

P-glycoprotein (P-gp)

P-gp, encoded by the MDR1 gene, is a 170 kDa plasma membrane protein which is produced from a 140 kDa precursor protein in the Golgi apparatus (Gottesman et al., 2002; Saraswathy & Gong, 2013). P-gp is composed of two pseudo-symmetrical halves, each consisting of an N-terminal transmembrane spanning domain (TMD) and a carboxy-terminal nucleotide binding domain (NBD). In the full length protein, these halves are arranged in tandem. Each of the TMDs has six transmembrane domains. Two pseudosymmetrical ATP-binding sites are formed at the interface of the NBDs (Chin et al., 1989; Gottesman et al., 2002). P-gp acquires energy for active transport of anticancer drugs through ATP binding and hydrolysis (Chang, 2003). It is able to intercept cytostatic drugs before they enter the cytoplasm and transport them back to the cell exterior against a concentration gradient. This leads to the phenotype of multidrug resistance, because P-gp has a multispecific substrate profile (Saraswathy & Gong, 2013). A large number of drugs used in cancer chemotherapy are substrates of P-gp, including anthracyclines, topoisomerase inhibitors, vinca alkaloids, camptothecin, and taxanes (Sun et al., 2004; Zhou, 2008). Studies have indicated that even a moderate increase in MDR1 expression (as low as 5-fold) were sufficient to cause doxorubicin resistance (Pajic et al., 2009).

Multidrug resistance protein

MRP, discovered in the process of studying the small cell lung cancer drug-resistant line H69AR, is another member of the human ABC family of plasma membrane associated transporters causing MDR (Cole et al., 1992). The MRP family consists of 13 proteins, most of which are associated with the transport of endogenous substances and xenobiotics. ABCC7/cystic fibrosis transmembrane conductance regulator (CFTR) is a chloride channel and ABCC8/sufonylurea receptor 1 (SUR1), and ABCC9/SUR2 are potassium channel regulators. Similar to P-gp MRP1 has been extensively studied over the past 25 years because of its prominent role in conferring multidrug resistance. MRP1 is a 190 kDa protein, consisting of 17 transmembrane domains with a P-gp-like core, but an additional N-terminal membrane spanning domain (TMD0) is present (Kruh & Belinsky, 2003; Leonessa & Clarke, 2003; Sodani et al., 2012). MRP1 is a basolateral transporter, the activity of which results in the movement of compounds into tissues that lie beneath the basement membrane. It is capable of transporting glutathione (GSH) and co-transporting drugs and-GSH and GSH-conjugated compounds (Evers et al., 1996). Efflux pumps involved in cellular export have been referred to as GSX pumps in the case of GSH conjugates (Ishikawa, 1992). Thence, the mechanism of MRP transport is different from that of P-gp (Kruh & Belinsky, 2003). Because MRP1 is ubiquitously expressed in human tissues, it is potentially present in most tumors and thus plays a role in drug resistance. MRP1 is able to confer resistance to anthracyclines, vinca alkaloids, epipodo-phyllotoxins, camptothecins and methotrexate, but not to taxanes, which are an important component of the P-gp profile. Besides, MRP1 is related to accelerated relapse in breast cancer, and a negative correlation between MRP1 expression and response to treatment has been found (Kruh & Belinsky, 2003; Sodani et al., 2012).

Breast cancer resistance protein

BCRP, encoded by the ABCG2 gene, was originally identified in an anticancer drug-resistant human breast cancer cell line under simultaneous treatment with mitoxantrone and inhibitors of P-gp. It was thus initially named mitoxantrone resistance protein (MXR) (Doyle et al., 1998; Ross et al., 1999; Chen et al., 2015). The protein, which belongs to the ABCG subfamily of ABC transporters, is an inverted half-transporter composed of an N-terminal nucleotide-binding site and a carboxyterminal transmembrane domain, which contains six membrane spanning helices (Gottesman et al., 2002; Gradhand & Kim, 2008). In its functional form, it is homodimeric and widely expressed in human tissues (Noguchi et al., 2009). Similar to ABCB1 and ABCC1 it plays an important role in intercellular drug absorption, distribution, metabolism, and excretion, as well as toxicity. The function as an active drug efflux transporter for anticancer drugs initially identified in breast cancer and later extended to other tumor types was eponymous for BCRP (Natarajan et al., 2012). In breast cancer cells, BCRP is not only expressed at the cell membrane, but also in intracellular vesicles that segregate and retain drugs, leading to drug compartmentalization. This is another reason for increased drug resistance of cells (Ifergan et al., 2005; Goler-Baron & Assaraf, 2011). Resistance of BCRP extends to the class of anthracyclines, but also to antifolatessuch as methotrexate (Doyle et al., 1998; Saraswathy & Gong, 2013).

Lung resistance protein

LRP, another resistance-related protein was initially identified in the non-small cell lung cancer cell line SW-1573/2R120, and is encoded by the LRP gene (Scheper et al., 1993). The gene for LRP was confirmed to regionally localize to chromosomal segment 16p11.2, a location approximately 27 cM proximal to MRP (16p13.1) (Slovak et al., 1995). Although both the MRP and LRP genes map to the short arm of chromosome 16, they are rarely co-amplified and are thus not normally located within the same amplicon. Primary sequence analysis of LRPcDNA revealed that its amino acid sequence was 87.7% homologous with the brown rat vault protein (major vault protein, MVP). LRP is not a member of the ABC transporter superfamily. It is found in the cytoplasm and at the nuclear membrane (Scheffer et al., 1995; Slovak et al., 1995). Vaults are cellular organelles broadly distributed and highly conserved among diverse eukaryotic cells, suggesting that they play a role in fundamental cellular processes. Vaults localize to nuclear pore complexes and may be the central plug of the nuclear pore complexes. Vault structure and localization support a transport function for this particle which could involve a variety of substrates. Vaults accordingly may play a role in drug resistance by regulating the nucleo-cytoplasmic transport of drugs (Izquierdo et al., 1996a). Histopathology demonstrated an increased amount of MVP/LRP protein in cancer, which results in lower anticancer drug levels in the nucleus (Szaflarski et al., 2011). Studies by Wood et al. (Wood & Streckfus, 2015) showed that LRP concentration in saliva of breast cancer patients in stageIwas significantly higher than that of healthy women. In a panel of 61 cancer cell lines which had not previously been undergone laboratory drug selection, LRP was a superior predictor for in vitro resistance to MDR-related drugs as compared to P-gp and MRP, and LRP's predictive value extended to MDR unrelated drugs, such as platinum compounds (Izquierdo et al., 1996b). Similar to the ABC MDRs, LRP also causes broad resistance to a partially overlapping panel of compounds including anthracyclines, alkaloids, epipodophyllotoxin, but in addition leads to resistance to cisplatin, melphalan and several atypical MDR drugs (Izquierdo et al., 1996a).

Other mechanisms of MDR

In addition to enhanced drug efflux mediated by MDR transporters, decreased drug concentration in the tumor is influenced by the tumor microenvironment, which is characterized by hypoxia and low pH (acidic cellular environment). Acidic pH outside cells limits uptake of weak base drugs such as doxorubicin, adriamycin and mitoxantrone (Chen et al., 2016). HIF-1α is a hypoxia-activated transcription factor that upregulates target genes by binding to the Hypoxia Response Element (HRE). MDR1 is a HIF-1 target gene and by that HIF-1-meditated P-gp expression results in drug resistance (Huang et al., 2010; Videira et al., 2014; Kapse-Mistry et al., 2014). Besides, by increasing water solubility of cytotoxic drugs, Glutathione S-transferase isoenzymes (GSTs) reduce toxicity, promote excretion and decrease the effective drug concentration (Huang et al., 2010). Altered DNA damage repair can also cause tumor MDR. TopoisomeraseII (TopoII) is an ATP-dependent enzyme that catalyzes the topological passing of two double-stranded DNA segments, and creates an intermediate DNA-enzyme complex referred to as the “cleavable complex” (Miller et al., 1981; Tewey et al., 1984). Reduced levels of TopoII catalytic activity decreases drug-induced TopoII-mediated cleavage of DNA, which leads to reduction of DNA breaks and lower cytotoxicity in MDR cells (Deffie et al., 1989; Ogiso et al., 2002). Alterations in apoptotic or antiapoptotic pathway such as loss of p53 (a transcription factor) (Oshika et al., 1998; Wang et al., 2011; Saha et al., 2012; Wang et al., 2013), abnormal expression of Bcl-2 family (Youle & Strasser, 2008; Saraswathy & Gong, 2013), increased activity of phosphatidylinositol-3-kinase (PI3K)/protein kinase B (Akt) (PI3k/Akt) (Jin et al., 2003; Zhang & Liu, 2011) and aberrant action of nuclear factor kappa beta (NF-κβ) (Bentires-Alj et al., 2003) also play a role in the development of MDR in breast cancer cells.

Nanoparticle encapsulated anticancer drugs overcome MDR in breast cancer cells

Anticancer drugs encapsulated in nanoparticles can actively or passively target tumor cells and thus improve the therapeutic effect at the target site. This can reduce systemic toxicity of chemotherapy drugs and bypass certain forms of multidrug resistance (Palakurthi et al., 2012). Blood vessels in most solid tumors possess the following unique characteristics: excessive angiogenesis and high vascular density, leaky vascular architecture, strengthened vascular permeability and impaired lymphatic drainage (Shing et al., 1985; Maeda & Matsumura, 1989; Skinner et al., 1990). The leaky blood vessels contain gaps in the basement membrane. The abnormal lymphatic drainage system of tumors fails to remove macromolecules and lipids from the interstitial space, trapping and retaining these over a longtime period. This phenomenon has been characterized as the tumor-selective enhanced permeability and retention (EPR) effect (Maeda & Matsumura, 1989; Maeda et al., 2000; Maeda, 2001). In passive targeting, nanoparticles pass through gaps in the leaky blood vessels and get trapped by the abnormal draining lymphatic system, generating the EPR effect. Passive targeting can also occur when positively-charged nanoparticles electrostatically interact with the negatively-charged sialic acid and phospholipids present on the surface of tumor associated endothelial cells (Patra & Turner, 2014; Cerqueira et al., 2015). Nanoparticles modified by active biomolecules such as nucleic acids, peptides, sugars and antibodies can actively bind to cancer cells. Ideally nanoparticles with high affinity selectively combine with molecules to target such as sugars, proteins, folate, transferrin, aptamers or lipids. These are overexpressed on the surface of cancer cells and nanoparticle delivery thus minimizes damage to non-cancer cells (Bertrand et al., 2014). Active targeting is applied to specifically recognize cells through the interactions of active biomolecules, enhance drug endocytosis by the cell, decrease cytotoxicity on non-cancer cells, increase the concentration of drug, and exploit the EPR effect (Patra & Turner, 2014; Cerqueira et al., 2015).

Nanoparticles that are taken up by the cell via endocytosis often bypass and evade the ABC-transporters responsible for efflux of cytotoxic drugs once released into the cytoplasm (Song et al., 2010; Huang et al., 2011; Cerqueira et al., 2015). There are four main mechanisms of endocytosis: clathrin-mediated endocytosis, caveolae-mediated endocytosis, macropinocytosis, and other endocytosis. Clathrin-mediated endocytosis is the most widely studied mechanism involved in receptor-mediated uptake of nanoparticles, apart from receptor-independent endocytosis. The relevant receptors include the transferrin, low density lipoprotein (LDL) receptor and the epidermal growth factor receptors (EGFR) including the human epidermal growth factor receptor 2 (HER2). Nanoparticles entering into cells via clathrin-mediated endocytosis end up in lysosomes and are degraded to release drugs by the acidic environment of lysosomes. Caveolae-mediated endocytosis generates cytosolic caveolar vesicles following nanoparticles binding to the cell membrane. Ligands associated with caveolae-mediated endocytosis include folic acid, albumin, and cholesterol. Macropinocytosis, another nonselective endocytosis pathway, relies on actin-driven membrane protusions, which subsequently fuzes with and separates from the plasma membrane to generate macropinosomes (Hillaireau & Couvreur, 2009; Saraswathy & Gong, 2013). Besides, nanoparticles containing co-encapsulated P-gp inhibitors and anti-cancer drugs can be used to overcome P-gp meditated MDR (Livney & Assaraf, 2013; Kibria et al., 2014; Cerqueira et al., 2015). Wong et al. (Wong et al., 2006) combined doxorubicin and elacridar, a P-gp inhibitor, in polymer-lipid hybrid nanoparticles. The in vitro result showed that simultaneous delivery of those two drugs enhanced the treatment of multidrug-resistant breast cancer (Table 1).

Table 1. Main mechanisms of nanoparticles reversing tumor MDR.

Nanoparticles	Property	Mechanisms of MDR reversal
Lipidosomes	Hydrophobic drug is dispersed in the lipid bilayer, while hydrophilic drug is wrapped in lipid bilayer	Membrane fluidization modulated by phospholipids (Lo, 2000)
		Inhibition of P-gp by anionic lipids (Kapse-Mistry et al., 2014) As the substrates for P-gp to reduce drug efflux (Van Helvoort et al., 1996; Bosch et al., 1997)
		PEGylated liposomes to prolong drug half-life (Yang et al., 2007)
		Endocytosis pathways to bypass active drug efflux (Kapse-Mistry et al., 2014)
Polymer Micelles	Core-shell structure spontaneously formed by amphiphilic copolymers in aqueous solution among which the shell is hydrophilic and the kernel is a hydrophobic drug reservoir composed by hydrophobic block	Improving drug solubility (Kataoka et al., 2001) Caveolae-mediated endocytosis to bypass drug efflux (Alakhova et al., 2010; Sahay et al., 2010) ATP depletion (Batrakova et al., 2001)
		Direct inhibition of P-gp (Kabanov et al., 2003)
Dendrimers	Polymer with spherical shape, dendritic skeleton and a number of functional groups on its surface	Large amounts of drug delivered due to high pore volume and surface area (Qin et al., 2014)
		Surface modification easily accomplished (Hong et al., 2007)
		Use endocytic pathways to bypass active drug efflux (Lu et al., 2011)
MSNs	High drug-loading capacity due to large pore volume and surface area	Uptake mainly by macropinocytosis clathrin mediated by folic acid receptors and caveolae-mediated mechanism ( Slowing et al., 2006; Huang et al., 2011; Shen et al., 2011)
		Inhibition of P-gp expression (Shen et al., 2011)

Liposomes

Liposomes have an aequeous core surrounded by a phospholipid bilayer. Therefore, they are able to carry both lipophilic and hydrophilic drugs (Qin et al., 2014). Liposomes, which are mainly composed of natural phospholipids, are thought to be pharmacologically inactive with minimal toxicity and therefore have good biocompatibility (Sercombe et al., 2015). It was shown that anionic membrane lipids (cardiolipin and phosphatidylserine) inhibited P-gp by direct interaction. This enhanced cellular absorption and cellular toxicity compared to free drugs (Kapse-Mistry et al., 2014). Phospholipids have been shown to be substrates of P-gp in many cell lines and P-gp was considered to be a phospholipid translocase with broad specificity (van Helvoort et al., 1996; Bosch et al., 1997). Lo (2000) proposed that modulation of P-gp by phospholipids could be brought about by substrate substitution or membrane fluidization, thereby increasing epirubicin cytotoxicity. Encapsulation of drugs in liposomes prevents rapid inactivation, degradation and dilution in the circulation (Hua & Wu, 2013). Kang et al. (2009) compared retention of Rhodamine 123, a fluorescent probe, often used as a paradigmatic P-gp and BCRP substrate, loaded into liposomes with different compositions in two kinds of breast cancer cells: wild type MCF-7/WT cells and P-gp overexpressing MCF-7/P-gp cells. The results showed that liposome encapsulation increased the retention of Rhodamine 123 in MCF-7/P-gp cell, but not in MCF-7/WT cells. Liposomes tend to accumulate preferentially in tumors because of the EPR effect. However, the delivery system is prone to rapid elimination by the reticuloendothelial system (RES) (Kang et al., 2009). To enhance liposome stability and improve the circulation times in the blood, PEGylated liposomes (long-circulating liposomes) were introduced. PEGylated liposome formulations of paclitaxel established by Yang et al. (2007) showed increased biological half-life of paclitaxel (17.8 (±2.35)h versus 5.05 (±1.52)h, reduced uptake in the RES and increased uptake in tumor tissues in rats compared to conventional liposomal formulations. Liposomes, that are not specifically targeted, can become targeted liposomes by linking selective ligands to their surface, which bears great potential for site-specific delivery of drugs. Examples of such targeted liposomes are: immunoglobuline modified liposomes (immune liposomes (ILs)), folate- or sugar-modified liposomes. Modified with selective ligands, liposome encapsulated drugs undergo targeted endocytosis and thus are not subject to active drug efflux by MDR transporters (Kapse-Mistry et al., 2014). Alternatively, higher temperature and lower pH in tumor tissue can be exploited by manufacturing temperature and pH-sensitive liposomes, which release drugs in target tissue in a temperature and pH-dependent manner (Qin et al., 2014; Vaidya et al., 2016).

Polymer micelle nanoparticles

Polymeric micelle nanoparticles with a core-shell structure are spontaneously formed by amphiphilic copolymer. The shell is hydrophilic and the kernel is a hydrophobic drug reservoir composed by a hydrophobic block copolymer. The hydrophilic shell inhibits protein adsorption, which decreases the polymeric micelles involved in foreign body reaction while improving drug solubility (Kataoka et al., 2001). The pluronic block copolymers are one of the polymeric micelle nanoparticles widely used in various drug delivery systems. Pluronics are triblock copolymers of poly(ethylene oxide) (PEO) and poly(propylene oxide) (PPO), arranged in PEO-PPO-PEO structure. They have a core containing PPO blocks, which serves to incorporate hydrophobic drugs (Alakhova & Kabanov, 2014). Pluronic block copolymers are thought to associate with energy metabolism in MDR cancer cells. It is rapidly taken up by the cells via a caveolae-mediated endocytosis pathway and co-localizes with mitochondria after 15 min. Moreover, it suppresses complex I and complex IV of the mitochondrial respiratory chain, reduces oxygen consumption and leads to ATP depletion in MDR cells, while non-MDR cells require significantly higher doses of Pluronic to achieve similar depletion (Alakhova et al., 2010; Sahay et al., 2010). Studies showed that the Pluronic P85 (P85) resulted in a substantial decrease in ATP levels selectively in MDR cells, thereby inhibiting P-gp and sensitizing MDR cells (Batrakova et al., 2001). Using isolated Pgp-, MRP1-, and MRP2-overexpressing membranes, Batrakova et al. (2004) evaluated the effects of P85 on the kinetic parameters V_max, K_m and first-order rate constant (V_max/K_m) of ATP hydrolysis by these ATPases. The effect of P85 on P-gp, MRP1 and MRP2 ATPase activity was as follows: MRP1 < MRP2 ≪ P-gp. Likewise, studies by Kabanov et al. (2003) indicated that the effect of P85 on cells with respect to ATP depletion correlated with the levels of expression of the multidrug transporter, primarily that of P-gp.

Micelles carrying drugs can selectively accumulate and release their cargo in tumors through the EPR effect (passive targeting), or when modified by specific ligands (active targeting). Thus, the micelles are ideal carriers for active and passive targeting approaches with the aim to enhance efficacy and reduce toxicity. More recent researches (Lee et al., 2005; Mohajer et al., 2007) showed that micelles with pH sensitivity or surface modifications, have improved performance in overcoming MDR. Studies on folic acid modified pH-sensitive doxorubicin micelles (mixture of poly-histidine-polyethylene glycol-folic acid and poly-l-lactic acid-polyethylene glycol-folic acid) revealed that drug was released from the endosome into the cytoplasm and nucleus due to the formation of nanomicelles from early endosomes. The latter was triggered by lower pH in this compartment (about 6). The micelle drug delivery system overcame P-gp efflux due to folate-receptor mediated endocytosis. This increased intracellular drug accumulation, prolonged residence time of intracellular drug, and enhanced cytotoxicity in resistant MCF-7/DOX cells (Lee et al., 2005; Mohajer et al., 2007; Kim et al., 2008). Studies by Wang et al. (2007) on four different paclitaxel (PTX) containing micelles prepared with Pluronic-P105, a mixture of P105 and L101, P105 modified by folic acid (FOL-P105) and a mixture of P105 and L101 modified by folic acid (FOL-PL) revealed that all micellar formulations had an improved anticancer effect on human breast cancer resistant MCF-7/DOX cells as compared to PTX in solution. FOL-P105 and FOL-PL micelles were more toxic than micelles without folic acid, indicating combined active targeting of ligands with the direct effect of Pluronic on multidrug resistant cells was superior in enhancing cytotoxicity (Kabanov et al., 2002; Batrakova et al., 2003,2004; Sharma et al., 2008). Further research (Werle & Hoffer, 2006) revealed that Chitosan-(4-thiol Butanimidamide)/GSH-mixed micelles not only increased the absorption of Rhodamine123 and Saquinavir, but also inhibited P-gp-dependent ATPase, leading to decreased drug efflux.

Dendrimers

Dendrimers are polymers with spherical shape, dendritic skeleton and a number of functional groups on its surface. In this drug targeting delivery system drugs may either be encapsulated within the internal cavity of the polymer or coupled with the surface functional groups, which resulted in improved drug-stability and decreased cytotoxicity (Qin et al., 2014). It was considered that dendrimer phthalocyanine was internalized into cells via endocytosis. Anionic dendrimer phthalocyanine (DPc) electrostatically interacted with poly(ethylene glycol)-b-poly(l-lysine) block copolymer (PEG-PLL) to form polyion complex(PIC) micelles. A doxorubicin containing DPc-encapsulated PIC micelle (DPc/m) was shown to internalize into cells via endocytosis, and accumulate in the endosome/lysosome compartments after irradiation. Finally, the drug would accumulate in the nucleus in drug-resistant MCF-7 cells (Lu et al., 2011). A G3 PAMAM-NH(2) dendrimer-chlorambucil conjugate exerted a more potent antiproliferative effect and stronger inhibition of [(3)H]thymidine incorporation into DNA in both MDA-MB-231 and MCF-7 breast cancer cells than chlorambucil alone (Bielawski et al., 2011). Targeted dendrimer nanoparticles are often prepared with surface-modified ligands and the targeting moieties such as sugar, folate, antibody, peptide or epidermal growth factor, which are conjugated to the dendrimers cause preferential accumulation of drug in the target tissue or cells than untargeted controls or free drug, thereby producing a stronger reduction of the tumor size (Hong et al., 2007). It has been shown that in analogous manner biotin (a cofactor for carboxylation reactions taken up in rapidly proliferating tissues such as cancer cells) was conjugated to the dendrimer to improve the uptake of anticancer drugs in tumor cells (Yang et al., 2009).

Mesoporous silica nanoparticles

Mesoporous silica nanoparticles (MSNs) can hold large amounts of drugs due to high pore volume and surface area. Drugs are loaded into the mesoporous pore to improve drug-stability and have targeted and sustained action. MSNs contribute to improving the solubility of hydrophobic drugs due to the water-soluble silanol groups on their surface, and can be modified for targeted delivery on the basis of different functionalities. Conjugation of targeting moieties with specific recognition on tumor cells include folic acid, monoclonal antibodies, glycoproteins, peptides, nucleic acids, which provide a means for effectively and accurately delivering drugs to target tissues (Trewyn et al., 2007; Mamaeva et al., 2013). It was reported that the main pathway of cellular uptake of MSNs was macropinocytosis, which is a form of endocytosis that accompanies cell surface ruffling and that this uptake was distinct from micropinocytosis including clathrin-coated vesicle endocytosis. Doxorubicin-loaded MSNs (DMNs) induced higher accumulation of doxorubicin in drug-resistant tumors than free doxorubicin. MSNs themselves caused down-regulation of P-gp expression. This could be the main reason for the enhanced anti-cancer activity of DMNs against P-gp over-expressing MCF-7/ADR cells. When DMNs were endocytosed into cells, doxorubicin would not be available for efflux by P-gp (Shen et al., 2011). Huang et al. (2011) used the endosomal pH-sensitive MSN (MSN-Hydrazone-Dox) for controlled release of doxorubicin in cells while evading its P-gp-mediated efflux. Both in vitro and in vivo studies showed that MSNs could serve as an efficient nanodelivery system entering cell via endocytosis and thus bypassing P-gp mediated efflux. Further studies (Slowing et al., 2006) showed that the MSN drug delivery system in combination with modifications by folic acid enhanced cellular uptake of drugs via receptor-mediated endocytosis. Moreover, Slowing et al. (2006) investigated the possible mechanism of endocytosis of the MSNs by introducing different surface functionalities. The authors prepared a fluorescein-functionalized MSN (FITC-MSN), and the functional groups 3-amino-propyl (AP), N-(2-aminoethyl)-3-aminopropyl (AEAP), N-folate-3-aminopropyl (FAP), guanidinopropyl (GP), and 3-[N-(2-guanidinoethyl)guanidino]propyl (GEGP) were grafted onto the external surface of FITC-MSN. The results indicated that the FITC- and FAP-MSNs were endocytosed via a clathrin-pitted mechanism mediated by folic acid receptors, while the endocytosis of AP- and GP-MSNs were endocytosed via acaveolae-mediated mechanism. The uptake mechanism for GEGP remained unclear. These studies indicated surface functionalities to play a role in the uptake of MSNs. The cytotoxicity of the MSN drug delivery system with different pore size on MCF-7/ADR cells showed that the drug delivery system with a pore size of 12.6 nm was more powerful than that of 3.2 nm. This confirmed that the anti-tumor activity in breast cancer cells is dependent on pore size, whereby apoptosis increased with increasing nanoparticle pore size (Gao et al., 2011; Jia et al., 2013).

In addition to MSNs, inorganic nanoparticles used in tumor treatment include quantum dots, carbon nanotubes, silica nanoparticles, gold nanoparticles, iron oxide magnetic nanoparticles and ceramic nanoparticles, which provide additional opportunities for the treatment and diagnosis of tumors.

Conclusions and outlook

MDR in tumor cells is still one of the major impediments in the clinical treatment of breast cancers. Overall thorough research is needed for further understanding the mechanism of MDR as a prerequisite for reversing various MDR forms in breast cancer, avoid MDR-related pathways, and enhance the efficacy of anti-tumor drugs. Nanodrugs hold potential in mediating sustained drug release, selectively targeting tumor cells, bypassing drug efflux, improving drug therapeutic indices, and reversing an MDR phenotype in tumor cells. Compared with free drugs, drugs encapsulated by nanomaterials can enter the target cells through different mechanisms to avoid the drug transporters P-gp, MRP and BCRP and to increase the intracellular accumulation of drugs and by that reverse MDR. The mechanisms leading to MDR are complex, and application of nanocarriers combined with targeting technologies to overcome or reverse MDR has been recognized as an important and promising field of research in recent years. It also holds great potential for tumor diagnostic procedures.

Although nanomaterials, with excellent physiological properties have been applied more and more widely in cancer treatment and diagnosis, studying on the potential toxicity of nanomaterials on cells and organism has not yet caught up with the rapid development and wide use of this technology. Using high-throughput methods such as metabolomics to evaluate the potential toxicity of nanoparticles holds excellent prospect for research and clinical applications in the future.

Declaration of interest

The authors report no conflicts of interest in this work.

We acknowledge the financial support from the National Natural Science Foundation of China (81273707, 81173215), the Ministry of Education in the New Century Excellent Talents (NECT-12-0677), the Natural Science Foundation of Guangdong (S2013010012880), the Science and Technology Program of Guangzhou (2014J4500005), the Science Program of the Department of Education of Guangdong (2013KJCX0021, 2015KGJHZ012), and the Science and Technology Program of Guangdong (2015A050502027).