The concept of mitochondrial structural equilibrium is influenced by various metabolic and cellular signals that may significantly affect mitochondrial physiology, and substantially and ultimately impact cellular health (1,2). This concept in the recent research has been exploited to form new drugs for the efficient management of various diseases including diabetes, Alzheimers diseases, Parkinsons diseases and cancer (3,4). Compared to normal cells, mitochondria of cancer cells exhibit a stark difference in terms of structure and function (5). Cancer cells have adapted to a series of metabolic changes in order to achieve better survival regardless of existing conditions such as availability of nutrients. These are detailed in the well-regarded Warburg effect, named after Warburg who first reported this phenomenon in 1927 (6,7). The Warburg effect describes malignant metabolic alterations encompassing elevated glycolysis as well as defects in mitochondrial oxidative respiration (8). This extensive metabolic reprogramming also influences significantly the various catabolic processes of cancer cells including apoptosis, necrosis and autophagy (9).

Despite the numerous therapeutic advantages that may be realized by targeting mitochondria, there are very few examples of mito-active drugs (10). This is reflected by the challenges of delivering compounds to this organelle, which possesses a convoluted and complex membrane structure that prohibits access to its inner space. A well-established approach is to exploit the negative charge of the mitochondrial matrix in order to turn compounds toward mitochondria. This electrochemical gradient facilitates the accumulation of positively charged molecules (cations) in the mitochondrial matrix (11). However, translocation is energetically unfavorable for molecules that are hydrophilic in nature (12). This category of lipophilic and membrane-permeable cations is then able to pass through the hydrophobic interior of the phospholipid bilayer of biological membranes while retaining their overall positive charge. As a result, these lipophilic cations, for example the commonly used mitochondrial fluorescence dyes, such as JC-1 or MitoTracker^® dyes, are capable of accumulating at mitochondria (13).

Although a wide range of delocalized lipophilic cations (DLCs) may in principle be utilised for drug targeting to mitochondria, only a few such compounds have been extensively researched for this purpose. The most prominent example is triphenylphosphonium (TPP) cation, a positively charged phosphorus atom surrounded by three hydrophobic phenyl groups that impart an extended hydrophobic surface (14). The TPP moiety has been employed as a probe to explore the mitochondrial membrane potential for over 40 years and consequently its behaviour and interaction with mitochondria are well defined (11). The relative TPP concentration inside the negatively charged membrane compartments increases by one order of magnitude for every 60 mV of negative membrane potential. Due to the active transport of salt ions by membrane-bound pumps, the interior part of the plasma membrane is negatively charged relative to the exterior side. As a result, the plasma membrane potential ranges generally between −30 and −60 mV, which is sufficient to promote up to 10-fold accumulation of TPP inside the cell. Typical mitochondrial membrane potential is up to −180 mV, which facilitates a 1,000-fold accumulation of TPP inside mitochondria (10).

The most extensively investigated example showing enhanced therapeutic efficacy conferred by conjugation with TPP for mitochondrial targeting is MitoQ (15). Another example, gamitrinib, is the mitochondrial-targeted Hsp90 inhibitor 17-allylamino-17-demethoxygeldanamycin (17-AAG) via TPP or 1–4 tandem repeats of cyclic guanidinium (16). The construct exhibited profound improvement in efficacy compared to 17-AAG with excellent selectivity for cancer cells/tissues in vitro and in vivo. Accordingly, gamitrinib is currently being evaluated in clinical settings (17).

Another group of DLCs that have been examined for their application in directing compounds to mitochondria is the rhodamine group of fluorescent compounds. Rhodamine compounds have been used as histological stains and some of the members have been utilized in investigating the energy state of isolated mitochondria (18).

Peptide-based sequences form another prominent branch of mitochondrial-targeting moiety apart from DLCs. Peptide-based mitochondrial molecular transporters possess some advantages over DLCs, including better biocompatibility and relatively straightforward synthesis with amide coupling. These sequences are normally derived from cytosolic synthesized proteins destined for trafficking to mitochondria (19). In adition to the 13 proteins encoded by mtDNA, the remaining mitochondrial proteins are ciphered in the nuclear genome and synthesized in the cytosol. Sorting to mitochondria is facilitated through the N-terminal mitochondrial targeting sequence (MTS), in which Arg, Ala, or Ser are abundant, while negatively charged amino acids such as Asp or Glu are rare (20). Despite the weak consensus between different MTSs, they consist of alternating hydrophobic amino acids and positively charged amino acids that form amphiphilic α-helix, which is crucial for the import of nascent proteins into mitochondria (20). Nevertheless, these naturally occurring MTSs are lengthy and their application in targeted delivery is limited by challenges in generating conjugates that are soluble and cell permeable (4). In view of this, several synthetic peptides have been designed and reported to have mitochondrial targeting activity. Notably, some of them also have intrinsic pharmacological properties as well. For example, the Szeto-Schiller (SS) peptide antioxidants is a class of short peptide antioxidants that also possess the ability to penetrate into IMM (21).

A therapeutic compound may be selectively delivered to mitochondria via covalent conjugation with a mitochondrial-targeting moiety. Nevertheless, this strategy is inevitably accompanied by limitations. For example, DLCs are ineffective in directing large polar cargo and intrinsically toxic to mitochondria at high concentrations. In addition, the covalent conjugation of bioactive compounds with DLCs entails modification of the structure, which may lead to alteration of chemical properties and thus the resulting activity. As a result, an alternative DS that can overcome these limitations, i.e., developing a mitochondrial-targeting carrier to incorporate the drug without modification, is highly demanded.

In light of this, great attention has been recently devoted to the development of a DS that is able to: i) Cross several biological barriers; ii) protect bioactive agents from premature deactivation; and iii) elevate intracellular bioavailability of the drugs at their target site (22). One of the emerging DS is represented by nanocarriers, i.e., nano-sized materials (possessing at least one dimension in the range of 1–100 nm) that can incorporate/encapsulate therapeutic compounds and be functionalized to carry multiple moieties such as imaging agents or targeting ligands at their surface. Nanocarriers, encompassing natural and synthetic polymer conjugates, polymeric nanoparticles, lipid-based carriers (such as liposomes and micelles), dendrimers, carbon nanotubes (CNTs), gold, platinum and titanium oxide nanoparticles (including nano-shells and nano-cages), have been subjected to intensive research for potential therapeutic applications (23).

A favourable attribute of the nanocarrier as effective DS for cancer therapy is the ability to passively accumulate at tumour tissues due to an enhanced permeability and retention (EPR) effect. Accordingly, the newly formed tumour vessels for supplying nutrients to rapidly growing tumour mass are usually abnormal in architecture with poor alignment and wide fenestrations. In addition, it is well recognized that tumour tissues lack effective lymphatic drainage (24). This allows nanocarriers to preferentially accumulate at the tumour sites (24). Such a property has been harnessed for drug delivery to neoplastic tissues, i.e., passive targeting. Additionally, due to the high surface area-to-volume ratio of the nanocarrier, it is possible to achieve high ligand density of targeting moiety on the surface for selective targeting (25). Thus, nanocarriers potentially increase local drug concentration at the tumour site by carrying drug molecules to the tumour and releasing them only in response to altered physical conditions specific to the tumour microenvironment. Thus, in the context of targeting mitochondria for cancer therapy, the development of nanocarriers to carry mito-active agents for improved targeting and bioavailability at mitochondria is of great interest.

The engineering of mitochondrial-targeting polymeric nanoparticles demonstrated the flexible application of this nanocarrier for therapeutics against diseases such as cancer, diabetes, and Alzheimers disease (26). Notably, FDA-approved poly(lactic-co-glycolic acid) (PLGA) and polyethylene glycol (PEG)-based materials were used to synthesize the nanoparticle, providing a distinctive edge in terms of biocompatibility. This accentuates the potential of advancing the nanoparticle to clinical settings in the near future. The mitochondrial-targeted PLGA-b-PEG-TPP copolymer was synthesized by conjugating TPP at the terminal end of PEG. The polymeric nanoparticle was assembled by a polymer-blending technology via the nanoprecipitation method (27), which is capable of fine-tuning the resulting size and surface charges for optimal mitochondrial localization.

One of the most extensively reviewed mitochondrial-targeted nanocarriers is dequalinium-based liposome-like vesicles (DQAsomes) (28). Dequalinium is a molecule with two symmetrical cationic charge centres separated by a hydrophobic carbon chain. DQAsome is made from the self-assembly of dequalinium chloride into vesicle-like aggregates in an aqueous suspension. This cationic colloidal DS formed is converted to DQAsome/DNA complexes by interaction with DNA. DQAplexes were shown to enter cells via endocytosis (29). The advantage of the DQAsome is the ability to evade endosomal compartments.

Given the versatility and effectiveness of CNTs in delivering various drugs, the role of CNTs in mitochondrial medicine has been investigated. The application of CNTs in organellar targeting is currently at its infancy stage. In an exploratory study, fluorescein isothiocyanate (FITC)-labelled PL-PEG (PL-PEG-FITC)-functionalized SWCNTs (SWCNT-PL-PEG-FITC) were serendipitously found to be localized at mitochondria in tumour and normal cells (HeLa, MCF-7, ASTC-a-1, COS-7 and EVC304) in a mitochondrial transmembrane potential-dependent manner (30). By contrast, in RAW264.7 murine macrophage, SWCNT-PL-PEG-FITC was observed to be mainly distributed in lysosomes instead of mitochondria, possibly due to lysosomal trafficking following phagocytosis. Notably, when SWCNTs were functionalized with molecules that are normally internalized by receptor-mediated endocytosis (FA and bovine serum albumin), the entire constructs were found to distribute to lysosomes (31). However, the SWCNTs would travel to mitochondria ensuing direct membrane penetration when the functionalized moiety was a small molecule chemical compound (FITC and PI). Since SWCNTs are a potent NIR-absorbing nanomaterial for the thermal destruction of tumour cells (31), this specific mitochondrial localization property was capitalized. It was observed that SWCNT-PL-PEG induced apoptosis in HeLa cells via mitochondrial damages after irradiation with NIR laser.

The design and synthesis of CNTs for active mitochondrial targeting has been reported recently (19). In that study, MWCNTs were functionalized covalently with an established MTS derived from the N-terminal region of subunit VIII of human cytochrome c oxidase. The unique mitochondrial localizing property of the MTS-MWCNTs, which was fluorescently labelled with a sulforhodamine B probe, was validated in RAW264.7 and HeLa cells using wide-field epifluorescence microscopy, confocal microscopy, and TEM (19). The colocalization of the conjugates with mitochondria was further confirmed by the presence of MWCNTs in isolated mitochondria observed under TEM. Different from the findings of Zhou et al (30), whereby the active design of CNTs for mitochondrial targeting was unnecessary, in the present study, targeting peptide was essential to direct CNTs to mitochondria.

Interaction of CNTs with biological species has not been comprehensively elucidated, as it has been observed to be a threat to the environment as well as human health. The interest of developing CNTs for biomedical applications means it is imperative to investigate and scrutinize the impact of CNTs especially on human. Over the past decades, numerous studies have been published on the interactions between CNTs with cells in vitro in terms of uptake, intracellular distribution, potential expulsion and even destruction/metabolism (32–35). However, the large number of dissimilarities existing between CNT batches employed in different research groups impose challenges in comparisons and the drawing of conclusions of the findings.

The nature of the functional molecules on CNTs surface also plays a determining role in the resulting cytotoxicity of CNTs construct. Several other independent groups also reported that CNTs coated by biological molecules, such as DNA, amphiphilic helical peptides, and serum proteins, were non-toxic to cells (36). Nevertheless, widespread concerns persist on the issue that CNTs are particularly strong in structure. In addition, it is generally recognized that large and pristine CNTs, which are poorly suspendable in aqueous solution, may form bundles and aggregates that induce inflammation and granuloma formation (37). On the other hand, such toxicities are not commonly observed with smaller and individualized CNTs (38). Furthermore, functionalization of CNTs with well-established biocompatible moieties, such as PEG, has conferred reduced in vivo toxicity after being i.v. injected into animals compared to the pristine counterparts (39). Nevertheless, issues of the effects of CNTs on reproductive functions and immune responses have only been partially addressed (40,41). Thus, it is essential to systematically investigate the impact of CNTs to living systems in all aspects before advancing CNTs into the clinical stage as a drug carrier.

It can therefore be determined that use of mitochondria targeting against cancer is promising. However, many improvements are required for proper and more efficient clinical use.