Of Yeast, Mice, and Men

A typical eukaryotic cell contains many types of membrane organelles with characteristic distributions within the cytoplasm. This distribution is facilitated in part by the organelle binding to a specific subset of motor proteins. Bound motors then transport the organelles to their proper destinations in the cell by moving them along microtubules or actin filaments. A key question of modern cell biology is how motor proteins recognize their target organelles. A number of recent articles address this question by showing that Rab proteins on the surface of organelles can function as a part of the recognition complex for motor proteins.

Rab proteins form the largest branch of the Ras superfamily of GTPases. They are found in organisms from yeast to human, and have been implicated in various functions within the cell including growth, protein trafficking, and the targeting and fusion of membrane bound organelles (Chavrier and Goud 1999; Pereira-Leal and Seabra 2000). The works of Bahadoran et al. 2001, Haddad et al. 2001, Hume et al. 2001, Raposo et al. 2001, and Stinchcombe et al. 2001, published in this issue, address the possibility that Rab proteins may also mediate motor/cargo interactions. They present compelling evidence to support this idea, while at the same time providing important information about some of the human disorders associated with organelle transport.

The most comprehensive evidence for the involvement of Rab proteins in intracellular transport is provided by recent studies demonstrating interactions between Rab proteins and the organelle-transporting motors of the myosin-V class in several experimental systems (Pruyne et al. 1998; Schott et al. 1999). Class V myosins are processive plus-end directed motors that have been implicated in intracellular transport in all eukaryotes from yeast to humans. They consist of two identical heavy chains and twelve light chains. The heavy chains contain an NH₂-terminal motor domain, a neck region with binding sites for light chains, a coiled-coil stalk, and a COOH-terminal globular domain that is involved in cargo binding (Reck-Peterson et al. 2000).

The first suggestion of a Rab/myosin-V interaction was provided by Pruyne et al. 1998 who showed that Myo2p, a yeast class V myosin, and Sec4p, a yeast Rab protein, colocalized to the growing bud in yeast cells (Pruyne et al. 1998). By using yeast with a conditional mutation in the MYO2 gene they were able to show that the localization of Sec4p was Myo2p dependent. Genetic evidence of an interaction between these two proteins in yeast was demonstrated by Schott et al. 1999 who showed that a mutation in MYO2 was synthetically lethal with a mutation in SEC4.

A direct physical interaction between Rab proteins and myosin-V was first demonstrated by Kumar et al. (Kumar, R., J. Navarre, and J.R. Goldenring. 1999. Mol. Biol. Cell. 10S:163a [abstr.]) who studied Rab11a, a Rab protein involved in the plasma membrane recycling system. By immunofluorescent staining of cultured cells, they showed that Rab11a is colocalized with myosin-Vb on vesicles containing the transferrin receptor. More importantly, by using a yeast two-hybrid system they demonstrated that Rab 11a binds the tail domain of myosin-Vb, suggesting that Rab11a may play a role in docking the motor to vesicles.

Rab–myosin-V interactions have been studied carefully in a series of recent papers that demonstrate a role for Rab proteins in myosin-Va–driven organelle transport in pigment cells of mice and humans (Menasche et al. 2000; Wilson et al. 2000; Bahadoran et al. 2001; Hume et al. 2001). In mammals, pigment melanin deposited in skin and hair is produced in melanocytes and transported to the cell periphery by the combined action of microtubule motors and myosin-Va (Wu et al. 1998). At the periphery of melanocytes, the pigment is secreted and then endocytosed by the adjacent epithelial cells. The genetics of pigmentation in mice has long been a subject of extensive study. One result attained by this analysis was the identification of myosin-Va as the protein encoded by dilute, a gene required for the transfer of pigment organelles from melanocytes to keratinocytes (Mercer et al. 1991). In addition to dilute, there are a number of other mutations that are characterized by similar phenotypes, and it is likely that the products of the genes affected by these mutations also work in the pigment organelle transport pathway. One of these genes, ashen, was found to encode Rab27a (Wilson et al. 2000), a Rab protein expressed in eye, lung, spleen, intestine, melanocytes, platelets, and pancreas (Seabra et al. 1995; Chen et al. 1997). Positional cloning has demonstrated that ashen mice have a mutation that encodes a nonfunctional version of Rab27a, and that the mutant phenotype can be rescued by addition of a BAC containing Rab27a (Wilson et al. 2000).

A similar picture of pigment transport mechanisms emerged from the analysis of Griscelli syndrome, a human autosomal recessive disorder that causes pigment abnormalities and neurological defects. Most cases also display immunological disorders that include haemophagocytic syndrome and defective exocytosis of lytic granules from cytotoxic T lymphocytes (Griscelli et al. 1978). Initially, the genetic defect in patients with Griscelli syndrome was mapped to the gene encoding myosin-Va. However, it has been shown recently that Griscelli patients with the immunological abnormalities have a mutation in the gene that encodes Rab27a rather than myosin-Va (Menasche et al. 2000).

These papers (Menasche et al. 2000; Wilson et al. 2000) suggest that Rab27a can work as a component of the pigment transport machinery that uses myosin-Va as a motor. In two recent works, one of which is published in this issue, it was shown that wild-type melanocytes transfected with a dominant negative form of Rab27a showed a perinuclear arrangement of melanosomes, similar to that observed in ashen and dilute melanocytes (Wilson et al. 2000; Hume et al. 2001; Wu et al. 2001). Conversely, it was demonstrated by Bahadoran et al. 2001 that transfection of wild-type Rab27a into melanocytes from a patient with Griscelli syndrome restored the wild-type distribution of melanosomes to the cell periphery. The same result was obtained in the mouse model using ashen melanocytes (Bahadoran et al. 2001; Wu et al. 2001).

In wild-type melanocytes, Rab27a is localized to the cytoplasmic face of melanosomes along with myosin-Va. However, in ashen melanocytes, myosin-Va failed to localize to melanosomes (Hume et al. 2001; Wu et al. 2001). These results indicate that not only does Rab27a colocalize with myosin-Va, but it is also required for motor binding to melanosomes. The simplest explanation for this observation is that Rab27a provides a binding site for myosin-Va on the surface of melanosomes. If this is the case, then the two proteins should interact with each other. Evidence for a physical interaction is provided by Hume et al. 2001 who show that antibodies against Rab27a are able to immunoprecipitate myosin-Va from melanocyte extracts. When taken together, the results of these experiments show that Rab27a is involved in targeting myosin-Va to melanosome membranes.

Studies of patients with Griscelli syndrome and ashen mice suggest that, in addition to regulating and anchoring myosin-Va, Rab 27a may have another function. It has been shown that patients with Griscelli syndrome caused by the mutation in Rab27a and ashen mice are deficient in the ability to secrete lytic granules from their cytotoxic T lymphocytes, resulting in a reduced ability to kill their targets (Menasche et al. 2000; Haddad et al. 2001; Stinchcombe et al. 2001). At the same time, Griscelli patients with the mutation in myosin-Va and dilute mice do not display these immunological defects (Menasche et al. 2000; Haddad et al. 2001). It could be that in T-lymphocytes, Rab27a is involved in the anchoring of some other motor required for lytic granule exocytosis, whereas in melanocytes, Rab27a is involved in the targeting of myosin-Va to pigment organelles. If this is the case, a mutation in the Rab27a gene would affect both types of organelles, whereas a mutation in myosin-Va would have no effect in T lymphocytes. Another possibility is that Rab27a is not involved in motor anchoring to lytic granules, but is required for the targeting and fusion of lytic granules with the plasma membrane. In this case, one may speculate that Rab27a plays a similar role in pigment cells by facilitating melanin exocytosis from melanocytes, a process that is still poorly understood.

Evidence is also emerging that suggests Rab proteins interact with microtubule motors. Echard et al. 1998 have demonstrated that a Golgi-associated Rab protein, Rab6, binds to a member of the kinesin family subsequently termed Rabkinesin-6, which appears to play a role in cytokinesis (Echard et al. 1998; Hill et al. 2000). Results obtained from studies with gunmetal mice (Stinchcombe et al. 2001) also suggest an interaction between Rab proteins and microtubule motors. Mice with the gunmetal mutation produce a partially functional Rab geranylgeranyl transferase (RGGT), an enzyme responsible for the prenylation and activation of several Rab family members (Detter et al. 2000). It is demonstrated in this issue by Stinchcombe et al. 2001 that lytic granules from gunmetal T lymphocytes do not polarize to the immunological synapse correctly, a process known to depend on microtubules (Kupfer et al. 1983). In contrast to this, lytic granule polarization in cytotoxic T lymphocytes from ashen mice appears to be normal, suggesting that Rab27a is not involved (Haddad et al. 2001; Stinchcombe et al. 2001). It could be the case that some other Rab protein not prenylated in gunmetal T lymphocytes is involved in the transport of lytic granules along microtubules.

The works in this issue (Bahadoran et al. 2001; Haddad et al. 2001; Hume et al. 2001; Raposo et al. 2001; Stinchcombe et al. 2001) indicate that one possible function of Rab proteins is to form a binding site for motor proteins on the surface of organelles. This would allow the motor proteins to bind the organelles, facilitating their transport and localization. The idea is an attractive one, in that Rab proteins are organelle-specific, which makes them well suited for such a function. It is interesting to note that while melanosomes share a number of membrane proteins with lysosomes, Raposo et al. 2001 suggests that there is a unique sorting pathway for their biogenesis. This pathway would allow melanosomes to possess a specific subset of Rab and motor proteins that would regulate their movement and exocytosis separately from lysosomes. Thus, Rab proteins may be one of the keys to unlocking the much bigger question of how motor proteins recognize specific organelles and transport them to their correct intracellular location.