David Chan Research Group at Caltech

Molecular analysis of mitochondrial dynamics and membrane fusion

| | | |


The primary focus of our lab is to understand the role of mitochondrial dynamics in normal cellular function and human disease (for some general reading on this topic, please see the reviews listed in the publications section.  Mitochondria are remarkably dynamic organelles that undergo continual cycles of fusion and fission (Figure 1).  The equilibrium of these two opposing processes determines not only the overall morphology of mitochondria in cells, but also has important consequences for mitochondrial function.

Our research falls into several broad areas:

To address these issues, we use a wide range of approaches, including genetics, biochemistry, cell biology, and structural biology.

Figure 1: Mitochondria undergoing fusion. In these still frames from a time-lapse movie, two pairs of mitochondria (labeled blue) approach each other, contact briefly, and fuse. From Chen et al (2003) J Cell Biol 160, 189-200.

I. Cellular and physiological functions of mitochondrial fusion and fission

A typical mammalian cell can have hundreds of mitochondria. However, each mitochondrion is not an autonomous organelle, because fusion and fission events mix mitochondrial membranes and contents. As a result, such events have major implications for the function of the mitochondrial population. We are interested in understanding the cellular role of mitochondrial dynamics, and how changes in mitochondrial dynamics can affect the function of vertebrate tissues.

Much of our work focuses on proteins involved in mitochondrial fusion or fission. Three large GTPases are essential for mitochondrial fusion. The mitofusins (Mfn1 and Mfn2) are transmembrane GTPases embedded in the outer membrane of mitochondria, and OPA1 is a dynamin-related protein localized to the intermembrane space (Figure 2A). Mitochondrial fission requires the function of Drp1, a dynamin-related protein that recruited to the mitochondrial surface to promote fission (Figure 2B).

Figure 2: Molecules that mediate mitochondrial fusion and fission. (A) Mitofusins (Mfns) on the outer membrane and OPA1 in the intermembrane space mediate mitochondrial fusion. (B) Fis1 and Drp1 mediate mitochondrial fission. From Chen and Chan (2006) Curr Opin Cell Biol 18, 453-459.

To understand the role of mitochondrial fusion in vertebrates, we have constructed mice deficient in either Mfn1 or Mfn2. We find that mice deficient in either Mfn1 or Mfn2 die in mid-gestation, indicating an essential function for mitofusins during embryogenesis. In both cases, the embryonic lethality results from placental dysfunction. From mouse models, we have generated cellular systems to dissect mitochondrial dynamics. Embryonic fibroblasts lacking Mfn1 or Mfn2 display fragmented mitochondria (Figure 3), a phenotype due to a severe reduction in mitochondrial fusion. Cells lacking OPA1 or both Mfn1 and Mfn2 have completely fragmented mitochondria and show no detectable mitochondrial fusion activity. Our analysis indicates that mitochondrial fusion is important not only for maintenance of mitochondrial morphology, but also for cell growth, mitochondrial membrane potential, and respiration. In part, these defects arise from a loss of mtDNA nucleoids, suggesting that content mixing due to mitochondrial fusion plays an important protective role for the mitochondrial population within cells (Figure 4).

We have also generated mice with conditional alleles of Mfn1 and Mfn2 and are using these mouse lines to examine the role of mitochondrial fusion in adult tissues. For example, we have discovered that loss of Mfn2 results in a highly specific degeneration of Purkinje neurons in the cerebellum (Figure 5). These studies are highly relevant to our understanding of several human diseases in which defects in mitochondrial dynamics lead to neurodegeneration (see below). We are also developing mouse models to understand mitochondrial fission and to track mitochondrial dynamics in vivo.

Figure 3: Deletion of Mfn1 or Mfn2 results in fragmentation of mitochondria. Wildtype mouse embryonic fibroblasts (left panel) have tubular mitochondria that can be over 10 microns long (arrow). Cells lacking Mfn1 (middle panel) or Mfn2 (right panel) have fragmented mitochondria. This result illustrates the concept that mitochondrial morphology results from a balance between the opposing processes of fusion and fission--in the absence of fusion, fragmentation occurs due to unbalanced fission. From Chen et al (2003) J Cell Biol 160, 189-200.

Figure 4: The protective role of mitochondrial dynamics. In wildtype cells (a), mitochondria are highly dynamic. Occasional dysfunctional mitochondria (orange) can be complemented functionally through mitochondrial fusion, which results in content mixing. In cells lacking fusion (b), dysfunctional mitochondria have no pathway for complementation and thereby accumulate. From Detmer and Chan (2007) Nat Rev Mol Cell Biol 8, 870-879.


Figure 5: Degeneration of cerebellar Purkinje cells in the absence of Mfn2. Wildtype Purkinje cells (left panel) in the cerebellum have elaborate dendritic arbors with many spines. Purkinje cells lacking Mfn2 (right panel) have greatly reduced dendritic arbors with few spines, and the cells eventually die. From Chen et al (2007) Cell 130, 548-562.

II. Molecular mechanism of membrane fusion and fission

The best understood membrane fusion proteins are viral envelope proteins and SNARE complexes. Viral envelope proteins, such as gp41 of HIV, reside on the lipid surface of viruses and mediate fusion between the viral and cellular membranes during virus entry (a topic we previously studied). SNARE complexes mediate a wide range of membrane fusion events between cellular membranes. In both cases, cellular and crystallographic studies have shown that the formation of helical bundles plays a critical role in bringing the merging membrane together. We would like to understand mitochondrial fusion at a similar level of resolution and to determine whether there are common features to these diverse forms of membrane fusion.

Mitofusins are the only conserved mitochondrial outer membrane proteins involved in fusion (Figure 6). Therefore, it is likely that they directly mediate membrane fusion. Consistent with this idea, mitofusins are required on adjacent mitochondria to mediate fusion (Figure 7). In addition, mitofusins form homotypic and heterotypic complexes that are capable of tethering mitochondria (Figure 8). We are trying to determine how tethered mitochondria, mediated by mitofusins, proceeds to full fusion. It should be noted that mitochondrial fusion is likely to be more complicated than most other intracellular membrane fusion events, because four lipid bilayers must be coordinately fused.

We are also exploring the roles of other proteins, such as OPA1, in mitochondrial fusion. In addition, we are using biochemical approaches to understand how mitochondrial fission complexes are assembled on the surface of mitochondria.

Figure 6: Schematic of mitofusin protein in the mitochondrial outer membrane. Note that both the N- and C-termini face the cytosol. HR1 and HR2 indicate two heptad repeat regions, motifs that generally form coiled-coil structures.


Figure 7: Mitofusins are required on adjacent mitochondria to mediate membrane fusion. In cell hybrids between wild-type and Mfn-null cells, mitochondria do not fuse. In this experiment, mitochondria from wild-type cells are labeled with GFP, and mitochondria from Mfn-null cells are labeled with dsRed. This lack of fusion suggests that mitofusin complexes act in trans. From Koshiba et al (2004) Science 305, 858-862.

Figure 8: The HR2 region of Mfn1 forms an antiparallel coiled coil that mediates tethering of mitochondria. Using X-ray crystallography, we found that the HR2 region (see Figure 6) forms a dimeric, antiparallel coiled coil. In a trans complex of mitofusin dimers, this structure would mediate tethering of mitochondria. In fact, certain mutants of Mfn1 can trap this tethered state. From Koshiba et al (2004) Science 305, 858-862.

III. Mitochondrial dynamics in human disease

           Two inherited human diseases are caused by defects in mitochondrial dynamics. Charcot-Marie-Tooth (CMT) disease is a neurological disorder that affects the peripheral nerves. Patients with CMT experience progressive weakness of the distal limbs and some loss of sensation. A specific type of CMT, termed CMT2A, is caused by mutations in Mfn2 that lead to degeneration of axons in peripheral nerves. We have used mouse and cellular models to understand how disease alleles of Mfn2 lead to neuronal dysfunction.

The most common inherited form of optic neuropathy (autosomal dominant optic atrophy) is caused by mutations in OPA1. This mitochondrial protein is localized to the inner membrane space and is essential for mitochondrial fusion. We have generated OPA1-null cells to understand the role of OPA1 in mediating membrane fusion and in maintaining mitochondrial function. In addition, we are developing biochemical assays to examine how OPA1 function is activated.

Finally, an understanding of mitochondrial dynamics will be essential for understanding a large collection of diseases termed mitochondrial encephalomyopathies. Many mitochondrial encephalomyopathies result from mutations in mitochondrial DNA (mtDNA). In mtDNA diseases, tissues maintain their mitochondrial function until pathogenic mtDNA levels exceed a critical threshold. Experiments with cell hybrids indicate that mitochondrial fusion, by enabling cooperation between mitochondria, can protect respiration even when >50% of mtDNAs are mutant. To understand the pathogenesis of mtDNA diseases, it is critical to explore how mitochondria can be functionally distinct and yet cooperate as a population within a cell. We anticipate that our studies with mice lacking mitochondrial fusion will help to shed light on this group of often devastating diseases. In addition, these studies may be relevant to understanding the link between mtDNA mutations, neurodegeneration, and age-related tissue degeneration.

IV. Additional research areas

In addition to mitochondrial dynamics, we have broad interests in other areas of mitochondrial biology. In particular, we have inititated research projects to understand the organization of mtDNA into nucleoids and to develop tools to analyze mtDNA function.