Designing an Organellar Postal Service: Delivery of Macromolecules to Mitochondria in Intact Cells

All cells in the body contain mitochondria: dynamic organelles that are essential for a myriad of functions. Defects of mitochondrial function can occur, however, resulting in significant disease. A method of carrying biomolecules to the mitochondrial matrix could therefore be of enormous potential benefit in therapeutics. It has proven to be notoriously difficult to target and import macromolecules to the matrix of these structures, particularly because of their double intracellular membranes. How can this problem be solved? Hatakeyama et al.¹ have now built on previous reports to generate a multilamellar nanoparticle that can be constructed around a cargo selected for mitochondrial targeting. This molecule can potentially target and fuse with endosomal and mitochondrial membranes, eventually releasing its cargo into the mitochondrial matrix. Although this approach has great potential, several questions remain regarding its current form.

Mitochondria house many critical proteins that are essential for the oxidative metabolism of foodstuffs, regulation of programmed cell death, and iron–sulfur cluster formation, among numerous other functions. Perhaps best characterized is their role in harnessing cellular respiration to the production of adenosine triphosphate in a process known as oxidative phosphorylation (OXPHOS). Of the more than 80 polypeptides that assemble into the five enzyme complexes that couple OXPHOS, 13 are encoded by the mitochondria's own genome, mitochondrial DNA (mtDNA). Defects of this multicopy genome are known to cause genetic disease. In addition, mutations have been associated with common neurodegenerative disorders and the aging process itself. There is therefore substantial interest in identifying methods to manipulate and repair this genome. Because there is no generally accepted method for transfecting mammalian mitochondria,² any method that could be designed to faithfully deliver biomolecules such as nucleic acids to the mitochondrial matrix would be invaluable.

The mitochondrial network is delineated by an outer mitochondrial membrane that contains a high density of voltage-dependent anion channels, allowing the free diffusion of molecules up to approximately 5–10 kDa. The inner membrane, which surrounds the mitochondrial matrix, is essentially an impermeable barrier, requiring a large family of substrate-specific transporters to import even low-molecular-weight charged molecules. Complex import mechanisms have evolved to transport the more than 1,000 different polypeptides necessary for maintaining mitochondrial structure and function. These mechanisms have been routinely exploited to import foreign proteins into mitochondria, but, although there have been sporadic reports of use of the protein import pathway for the uptake of other large macromolecules,^3,4,5 it has not won general support as a viable method. An alternative technique has been to utilize the potential of the mitochondrial membrane to target chemically designed transporter molecules. Murphy and Skulachev and colleagues have had great success in using caged lipophilic cation derivatives as vectors for targeting a variety of small reactive oxygen scavengers to the mitochondrial inner membrane.^6,7 The targeting of larger macromolecules to the mitochondrial matrix by hooking onto these targeting moieties, however, is proving more difficult. A possible method for targeting and importing DNA into the mitochondria of intact cells has circled the mitochondrial world for several years.⁸ Referred to as ProtoFection, this method focuses on coating the DNA to be imported with TFAM, a chimeric mitochondrial DNA-binding protein, downstream of a mitochondrial localization sequence and topped with an N-terminal 11-arginine protein transduction domain. It is unclear from current published data whether ProtoFection does indeed result in the delivery of exogenous DNA to the mitochondrion, although treatment with the DNA–protein complex or the chimeric TFAM protein alone does appear to result in extended alterations in mitochondrial physiology.⁹

In the absence of a generally accepted, robust mechanism for delivering foreign macromolecules to the mitochondrial matrix, Hatakeyama et al.¹ have been engineering a synthetic nanocarrier to resolve this problem. The novelty of their approach is to encapsulate the desired mitochondrial cargo in a series of liposomes with lipid compositions that optimize membrane fusion potential, with the addition of an external octa-arginine moiety to facilitate membrane transduction and localization. This group had previously synthesized a lipid nanocarrier, termed MITO-Porter, designed to target a specific cargo, propidium iodide, to the mitochondrial matrix.^10,11 This membrane-impermeant dye would intercalate with the matrix-localized mitochondrial DNA and fluoresce to give a specific signal, confirming matrix localization. Although the authors reported some tantalizing results, the data showed low-level efficiency of delivery. The authors have now extended this approach by encapsulating cargo in a series of multilayered fusogenic liposomes termed dual-function (DF)-MITO-Porter. All vesicles carry specific lipid compositions, starting with the outer layer, to fuse efficiently with the endosomal membrane, resulting in cytosolic import following pinocytosis. Removal of the outer vesicle then reveals two additional liposomes, which include the octa-arginine residues that help target the vesicle to the mitochondrion. Fusion occurs with both the outer and inner membranes, resulting in the final deposition of the core cargo into the matrix. In these latest experiments, the authors used DNase 1 as the cargo, with the intended proof of matrix import being the degradation of endogenous mtDNA and concomitant loss of mtDNA gene products, components of the mitochondrial OXPHOS machinery.

The production of such tailored multilamellar molecules is clearly an impressive display of nanotechnology. Many of the data compel one to believe that the vesicles have indeed delivered DNase 1 into the mitochondrial matrix. For example, the authors measured a depletion of mtDNA during the time course of the experiment. However, several observations remain unexplained. Why do the cells die while the steady-state levels of mtDNA-encoded OXPHOS components are unaffected? The toxicity of these current nanocarriers is a major concern. The authors attempted to measure the mitochondrial toxicity of these compounds directly using a well-known method based on the reduction of tetrazolium salts by dehydrogenases. Many life scientists would be familiar with this method, which uses the activity of mainly extramitochondrial dehydrogenases as an indicator of cell number.¹² It is therefore unclear why the authors claim that this assay specifically measures mitochondrial dehydrogenase or serves as a proxy for “mitochondrial function.” It is possible that the authors were measuring a more general cell dysfunction and perhaps cytotoxicity directly. One possible concern is that the DNase 1 itself is being delivered not only to the mitochondrion but also to the nucleus, leading to cell death. The simple way to refute this possibility would be to perform similar experiments in cells lacking mtDNA (rho0 cells), as it is difficult to see why delivering DNase 1 to rho0 mitochondria should be toxic to cells. A second possibility is that the fusogenic liposome inadvertently perturbs the mitochondrial infrastructure, leading to a form of cell death.

Even considering these major caveats, this is still a very promising methodology. There is great potential that such concerns will be resolved with time and that a robust method for the delivery of many sorts of macromolecules to the mitochondrial matrix will become available. It is exciting to speculate that iterations of this method may eventually produce the first technique for transfecting mammalian mitochondria in situ.