During oogenesis in Drosophila, determinants that will dictate abdomen and germline formation are localized to the 'polar plasm' in the posterior of the oocyte. Assembly of the polar plasm involves the sequential localization of several messenger RNAs and proteins to the posterior of the oocyte, beginning with the localization of Oskar mRNA and Staufen protein during stages 8 and 9 of oogenesis. The mechanism by which these two early components accumulate at the posterior is not known. Directed transport along microtubules could be used to accomplish this localization. A fusion protein composed of the bacterial beta-galactosidase enzyme as a reporter was used, joined to kinesin, part of the plus-end-directed microtubule motor. The fusion protein transiently localizes to the posterior of the oocyte during stages 8 and 9 of oogenesis. Treatment with the microtubule-depolymerizing agent colchicine prevents both the localization of the fusion protein and the posterior transport of Oskar mRNA and Staufen protein. Furthermore, the fusion protein localizes normally in oocytes mutant for either oskar and staufen, but not in other mutants in which Oskar mRNA and Staufen protein are mislocalized. Thus, association with a plus-end-directed microtubule motor can promote posterior localization of a reporter protein during oogenesis. The genetic requirements for this localization and its sensitivity to colchicine, both of which are shared with the posterior transport of Oskar mRNA and Staufen protein, suggest that a similar mechanism may function in both processes (Clark, 1994).
The asymmetric localization of messenger RNA (mRNA) and protein determinants plays an important role in the establishment of complex body plans. In Drosophila oocytes, the anterior localization of Bicoid mRNA and the posterior localization of Oskar mRNA are key events in establishing the anterior-posterior axis. Although the mechanisms that drive Bicoid and Oskar localization have been elusive, oocyte microtubules are known to be essential. The plus end-directed microtubule motor kinesin I is required for the posterior localization of Oskar mRNA and an associated protein, Staufen, but not for the anterior-posterior localization of other asymmetric factors. Thus, a complex containing Oskar mRNA and Staufen may be transported along microtubules to the posterior pole by kinesin I (R. P. Brendza, 2000b).
To determine if kinesin I is involved in oocyte patterning, mitotic recombination was used to generate mosaic female flies containing clones of homozygous Khc null germ line stem cells. The production of eggs and embryos by the mosaic females suggests that germ line stem cells can proliferate and proceed through oogenesis without kinesin I. However, embryogenesis fails, despite fertilization by wild-type males. Most embryos arrest before blastoderm formation, but a few proceed into early gastrulation stages. This maternal lethal effect is completely rescued by a wild-type Khc transgene. Thus, germ line expression of KHC is required for normal embryogenesis (R. P. Brendza, 2000b).
Examination of embryos that reach the blastoderm stage reveals an absence of pole cells, the germ line precursors. To assay for earlier defects, the distributions of OSK and BCD mRNAs in Khc null oocytes were examined. The localization of BCD mRNA is normal, concentrated at the anterior during stages 8 to 10. In contrast, the localization of OSK mRNA is defective. It normally accumulates transiently at the anterior pole early in stage 8 and then moves to the posterior pole. In Khc null stage 8 to 10 egg chambers, OSK mRNA accumulates excessively at the anterior pole and is never concentrated at the posterior pole. This localization defect is completely rescued by a wild-type Khc transgene. Thus, although KHC is not required for anterior localization of either BCD or OSK mRNAs, it is required for the posterior localization of OSK. Perhaps kinesin I transports OSK mRNA along microtubules toward their plus ends and the posterior pole (R. P. Brendza, 2000b).
Given that microtubule-disrupting drugs prevent the posterior localization of OSK mRNA during oogenesis, the possibility was considered that the absence of KHC blocks OSK localization indirectly by disturbing oocyte microtubules. The shift of the oocyte nucleus from posterior to anterior poles during stage 6, which is microtubule-dependent, appears normal in Khc null oocytes. Furthermore, the anterior localization of the MTOC component Centrosomin is normal (R. P. Brendza, 2000b).
Microtubule organization was tested further by localizing a hybrid protein composed of the motor domain of KHC fused to a reporter enzyme, beta-galactosidase (beta-Gal). KHC::beta-Gal is thought to localize in regions of cells with high concentrations of microtubule plus ends. It is important to note that this chimeric protein does not rescue patterning defects in Khc null oocytes. In wild-type stage 9 to 10a oocytes, KHC::beta-Gal concentrates at the posterior pole. In Khc null oocytes, KHC::beta-Gal concentrates at the posterior pole in most instances. This suggests that in most of the stage 9 to 10a null oocytes with detectable amounts of KHC::beta-Gal, microtubule plus ends are concentrated at the posterior pole. This, and the indications that microtubules in the anterior end are normal, suggests that OSK mRNA mislocalization in Khc null oocytes is not due to a disruption of microtubule organization. In Khc null oocytes, KHC::beta-Gal staining is often not detected in oocytes, although it is visible in nurse cells. Perhaps efficient transport of the hybrid protein from nurse cells to oocyte requires the presence of native KHC (R. P. Brendza, 2000b).
Posterior transport of OSK mRNA is thought to depend on Staufen protein. Staufen is transiently localized at the anterior end of the oocyte during stage 8 where it may form a complex with OSK mRNA. If kinesin I transports such OSK-Staufen complexes along microtubules to the posterior pole, then Staufen protein should be mislocalized in Khc null oocytes. Immunostaining with anti-Staufen confirms this prediction. In wild-type stage 8 to 10 oocytes, Staufen concentrates at the anterior end early, appears in granules along the cortex, and then concentrates at the posterior end. Granular Staufen distribution was detected in most oocytes observed. This is consistent with the hypothesis that Staufen and OSK mRNA form complexes at the anterior cortex that are transported to the posterior pole. In Khc mutant oocytes, Staufen protein overaccumulates in the anterior end during stage 8, is not detected in granules, and does not concentrate at the posterior pole. Normal Staufen distribution patterns are restored in Khc null oocytes by the addition of a wild-type Khc transgene (R. P. Brendza, 2000b).
Thus, KHC, the force-generating component of the plus end-directed microtubule motor kinesin I, is required for the posterior localization of both OSK mRNA and Staufen protein. The participation of kinesin I in this mRNA motility process could be direct. It might attach specifically to osk-Staufen complexes at the anterior pole and transport them toward the posterior pole. However, initial tests for coimmunoprecipitation of KHC and Staufen from Drosophila ovary cytosol have not revealed any robust association, so perhaps the linkage is less direct. It is generally accepted that kinesin I transports membranous organelles toward microtubule plus ends. Thus, OSK and Staufen could localize to the posterior pole by virtue of association with mitochondria or other organelles carried by kinesin I. An alternative to these models is derived from the effect of a loss of KHC on the particulate staining pattern of Staufen. Before stages 7 to 8, while microtubules are still oriented with their plus ends toward the anterior, kinesin I might deliver, to the cortex, materials necessary for the assembly of transport-competent OSK-Staufen complexes. Thus, the lack of visible Staufen particles in Khc null oocytes may indicate that their assembly or persistence depends on kinesin I activity. New studies, using green fluorescent protein tags to follow the localization dynamics of OSK mRNA, Staufen, and organelles, may distinguish between these models and provide further insight into the mechanisms that drive the movements of maternal determinants for early developmental patterning (R. P. Brendza, 2000b).
A broadly conserved membrane-associated protein required for the functional interaction of kinesin-I with axonal cargo was identified. Mutations in sunday driver (syd) and the axonal transport motor kinesin-I cause similar phenotypes in Drosophila, including aberrant accumulations of axonal cargoes. GFP-tagged mammalian SYD localizes to tubulovesicular structures that costain for kinesin-I and a marker of the secretory pathway. Coimmunoprecipitation analysis indicates that mouse SYD forms a complex with kinesin-I in vivo. Yeast two-hybrid analysis and in vitro interaction studies reveal that SYD directly binds kinesin-I via the tetratricopeptide repeat (TPR) domain of kinesin light chain (KLC) with K(d) congruent with 200 nM. It is proposes that SYD mediates the axonal transport of at least one class of vesicles by interacting directly with KLC (Bowman, 2000).
Mutants in the actin nucleators Cappuccino and Spire disrupt the polarized microtubule network in the Drosophila oocyte that defines the anterior-posterior axis, suggesting that microtubule organization depends on actin. Cappuccino and Spire organize an isotropic mesh of actin filaments in the oocyte cytoplasm. capu and spire mutants lack this mesh, whereas overexpressed truncated Cappuccino stabilizes the mesh in the presence of Latrunculin A and partially rescues spire mutants. Spire overexpression cannot rescue capu mutants, but prevents actin mesh disassembly at stage 10B and blocks late cytoplasmic streaming. This study also shows that the actin mesh regulates microtubules indirectly, by inhibiting kinesin-dependent cytoplasmic flows. Thus, the Capu pathway controls alternative states of the oocyte cytoplasm: when active, it assembles an actin mesh that suppresses kinesin motility to maintain a polarized microtubule cytoskeleton. When inactive, unrestrained kinesin movement generates flows that wash microtubules to the cortex (Dahlgaard, 2007).
The main body axes of Drosophila are established during stages 7-9 of oogenesis when the oocyte microtubule (MT) cytoskeleton is reorganized to direct the asymmetric localization of bicoid (bcd), oskar (osk), and gurken mRNAs. At stage 7 of oogenesis, an unknown signal from the posterior follicle cells induces the disassembly of a microtubule-organizing center at the posterior of the oocyte, while new MTs nucleate from the anterior-lateral cortex with their plus ends extending toward the posterior pole. This results in the formation of an anterior-to-posterior gradient of MTs that directs the localization of bcd and osk mRNAs to the anterior and posterior poles of the oocyte, respectively, where they act to determine the anterior-posterior axis of the embryo. The polarized MT cytoskeleton is also required for the migration of the oocyte nucleus from the posterior of the oocyte to a point at the anterior margin, and this defines the dorsal-ventral axis by directing the localization of gurken mRNA to one side of the nucleus, where Gurken protein is secreted to induce dorsal follicle cell fates (Dahlgaard, 2007 and references therein).
The organization of the MTs changes during stage 10B, and they form parallel arrays around the cortex of the oocyte that drive a fast unidirectional movement of the oocyte cytoplasm, called ooplasmic streaming. Ooplasmic streaming requires the plus-end-directed MT motor, Kinesin, suggesting that the flows are generated by kinesin-dependent transport of organelles or vesicles. The cytoplasm is also in motion in oocytes from stages 8-10A, but these movements are slower and uncoordinated and have been named ooplasmic seething (Dahlgaard, 2007).
The polarized organization of the MTs at mid-oogenesis requires the function of par-1 and capu groups of genes. In mutants in the former group, which comprises par-1, lkb-1, and 14-3-3epsilon, the MTs appear to be nucleated all around the oocyte cortex, with their plus ends in the center. As a consequence, osk mRNA is mislocalized to a dot in the center of the oocyte, while bcd mRNA spreads from the anterior around most of the cortex. However, the localization of gurken mRNA is wild-type in these mutants. The polarity signal from the follicle cells induces the actin-dependent localization of PAR-1 to the posterior cortex of the oocyte, suggesting that asymmetric PAR-1 activity plays a key role in the polarization of the oocyte MT cytoskeleton (Dahlgaard, 2007).
Mutants in cappuccino (capu), chickadee (chic), and spire produce a distinct phenotype, in which the MTs form prominent arrays around the oocyte cortex during stages 8-10 and MT plus-end markers no longer localize to the posterior pole. These mutants also cause premature streaming of the oocyte cytoplasm, which resembles the cytoplasmic streaming seen in wild-type oocytes after stage 10B. As a result, both osk and gurken mRNAs are mislocalized, leading to abdominal defects in the embryo and ventralized eggs, although the localization of bcd mRNA is unaffected (Dahlgaard, 2007).
Actin-depolymerizing drugs produce identical MT and premature cytoplasmic streaming phenotypes to capu, chic, and spire mutants, indicating that actin is required for the correct organization of the MT cytoskeleton. Consistent with this, all three genes encode regulators of the actin cytoskeleton. Chickadee is Drosophila Profilin, which binds free G-actin protein to regulate actin dynamics; Spire is the founding member of a new family of actin nucleation factors that nucleate filaments from their pointed ends; Capu is a member of the Formin family of proteins, which also nucleate actin filaments, but in this case from their barbed ends (Dahlgaard, 2007 and references therein).
Although effects of actin depolymerization strongly suggest that actin plays a key role in the organization of the oocyte MT cytoskeleton, it is not clear which population of F-actin in the oocyte is responsible for this effect, or how Capu, Spire, and Profilin participate in the interaction between actin and MTs. One possibility is that Capu, Profilin, and Spire regulate MTs by directing the posterior recruitment of PAR-1, since they have been proposed to play a role in the organization of cortical actin, which is required for PAR-1 localization. This cannot account for all of the effects of the capu group mutants, however, since they produce a different phenotype from par-1 mutants. An alternative possibility is suggested by experiments showing that formin-related proteins can control the positioning or stability of MT plus ends. Bni1p is required for spindle positioning during early metaphase in budding yeast, through the recruitment of the plus ends of astral MTs to the bud tip. Bni1p localizes to the emerging bud tip and nucleates unbranched actin filaments. The myosin, Myo2p, then transports the MT plus ends along these actin cables to the bud tip, through its linkage to the plus-end-binding protein, Kar9p. In contrast, the mouse formin mDia1 acts independently of actin to stabilize MT plus ends at the leading edge of migrating NIH 3T3 cells, through a pathway that involves the inhibition of GSK3β and the plus-end-binding proteins, EB1 and APC. Thus, Capu may function in a similar way to either Bni1 or mDia to recruit or stabilize MT plus ends at the posterior of the oocyte (Dahlgaard, 2007).
A different model has been proposed for the function of Capu and Spire, in which they act not as actin nucleators but as crosslinkers between MTs and cortical actin. The spire locus encodes multiple isoforms, including two short forms, Spire D and Spire C, that encompass the N-terminal and C-terminal halves of the longest isoform, respectively. Spire D contains the KIND domain and the 4 WH2 domains that have been shown to nucleate actin in vitro and when transiently expressed in mouse fibroblasts, whereas Spire C includes an mFYVE domain and a JNK-binding site. In binding studies with tubulin and actin in vitro, both Capu and Spire C induced the bundling of actin with MTs. In contrast, Spire D nucleated F-actin in vitro but did not interact with MTs and inhibited the actin/MT crosslinking activity of Capu and Spire C. This has led to the proposal that Capu and Spire C repress the cortical bundling of MTs and premature cytoplasmic streaming by crosslinking the MTs to the cortical actin, whereas Spire D is a negative regulator of this process (Dahlgaard, 2007).
To distinguish between the different models for the function of Capu and Spire, various domains of each protein were expressed in wild-type and mutant egg chambers in order to analyze their subcellular localizations and their effects on actin, MTs, and cytoplasmic streaming in vivo. The results indicate that neither the cortical localization nor the MT-binding activity of Capu and Spire is required for their function. Instead, Capu and Spire are shown to act to assemble a dynamic actin mesh in the oocyte cytoplasm (Dahlgaard, 2007).
Formin-related proteins play a key role in cell polarity in a number of systems and usually show a highly polarized distribution to one end of the cell. For example, Bni1p and For3p localize to the poles of budding and fission yeast, respectively, where they nucleate actin cables that are required for polarized growth, while mDia stabilizes MT plus ends at the leading edge of migrating fibroblasts. Although the Drosophila forming-related protein, Capu, is similarly required for the polarization of the oocyte MT cytoskeleton and for the formation of both the anterior-posterior and dorsal-ventral axes, the results reported in this study demonstrate that Capu regulates MTs by a very different mechanism from these other formins. Neither Capu nor its partner Spire shows a polarized distribution within the oocyte, nor do they play a direct role in MT organization in a particular region of the cell. Instead, they act together with Profilin to assemble an isotropic actin mesh in the oocyte cytoplasm, which maintains the polarized arrangement of MTs by suppressing kinesin-dependent cytoplasmic streaming (Dahlgaard, 2007).
This function for Capu and Spire contrasts with the recent proposal that they act at the oocyte cortex to regulate cortical polarity and to crosslink the actin and MT cytoskeletons. The results argue against this model for several reasons. First, cortical polarity appears to be unaffected in capu and spire mutant egg chambers. PAR-1 still localizes normally to the posterior cortex, and osk mRNA is specifically anchored at the posterior in spire mutant egg chambers, indicating that this region of the cortex is different from the anterior and lateral domains. Furthermore, the MTs show a normal association with the anterior and lateral cortex in capu and spire mutants, as is most clearly demonstrated by the wild-type MT arrangement in capu mutants in which kinesin function is impaired (Dahlgaard, 2007).
Second, although Capu and Spire interact with MTs in vitro, this activity does not appear to be required for their function in vivo. Spire D, which lacks the MT-binding domain, completely suppresses cytoplasmic streaming at all stages, whereas Spire C, which contains the domain, has no effect on the spire mutant phenotype. Thus, the MT-binding activity of Spire is not required for its in vivo activity. A similar argument can made for the MT-binding activity of Capu. Capu binds MT in vitro through its FH2 domain, and a P597T substitution in the capu2F allele blocks this activity. Despite this loss of MT binding, capu2F has the weakest phenotype of all capu alleles examined, indicating that the inability to interact with MT has little effect on Capu's in vivo activity. Furthermore, the weak phenotype of capu2F is more easily explained by an effect on actin nucleation, since a clear reduction in the actin mesh in was observed capu2F homozygous oocytes, although the P597T mutation was reported to have minimal effect on actin nucleation in vitro (Dahlgaard, 2007).
The localization of Capu and Spire also argues against a model in which they act exclusively to anchor MTs to the cortex. Neither GFP-tagged Capu nor Spire D is enriched at the oocyte cortex when visualized in living oocytes, even though these fusion proteins are functional, since they rescue the strongest alleles of capu and spire, respectively. This contrasts with a previous study in which both proteins were reported to localize to the oocyte cortex, and may reflect the fact that the latter examined their distribution in detergent-extracted and fixed samples. It therefore seems unlikely that the direct crosslinking of actin and MTs by Capu or Spire at the cortex plays a significant role in their function in the oocyte (Dahlgaard, 2007).
Instead, the results indicate that the Capu pathway functions to organize a dynamic network of actin filaments throughout the oocyte cytoplasm. This actin mesh is lost in capu, spire, and chic mutants, indicating that Capu, Spire, and Profilin are necessary for its formation. Furthermore, overexpression of Capu or Spire D induces an ectopic mesh in the nurse cells, while Spire D induces an ectopic mesh in late oocytes, strongly suggesting that both proteins play a direct role in its assembly. Indeed, the role of Capu in the formation of the cytoplasmic actin mesh may explain the seemingly paradoxical observation that capu mutants cause an increase in the amount of cortical actin in the oocyte. The failure to form the actin mesh in capu mutants should lead to a rise in the concentration of free G-actin in the oocyte, which may promote excess actin polymerization at the oocyte cortex by a Capu- and Spire-independent mechanism (Dahlgaard, 2007).
The presence of the ooplasmic actin mesh correlates perfectly with the polarized arrangement of the MTs in the oocyte. The mesh is present from stage 5 to stage 10A of oogenesis, which is the period during which the anterior-posterior gradient of MT persists, and the disappearance of the mesh at stage 10B coincides with the onset of fast cytoplasmic streaming and the rearrangement of the MT into parallel cortical arrays. Furthermore, the loss of the mesh in capu, spire, and chic mutants leads to premature streaming and the precocious formation of cortical MT arrays, whereas the overexpression of Spire D maintains the mesh during stage 11 and suppresses the normal rearrangement of the MT and streaming at this stage. Indeed, the density of the mesh correlates with the severity of the mutant phenotype, since the weakest alleles of capu and chic cause a reduction in the mesh without abolishing it entirely (Dahlgaard, 2007).
This revised view of the function of Capu and Spire is consistent with the known biochemical properties of the other formin-related proteins and Spire. In vitro studies have shown that formin-related proteins nucleate actin filaments through their FH2 domains and then remain associated with the barbed end, which they protect from actin-capping proteins, while increasing the rate of elongation through the interaction of the FH1 domain with Profilin/Actin complexes. Although Capu is not a typical formin, it contains well-conserved FH1 and FH2 domains, nucleates actin in vitro, and has been shown to interact with Profilin in yeast two-hybrid assays. Furthermore, the protection of the actin mesh from Latrunculin A-induced depolymerization by CapuΔN is consistent with a model in which the protein remains associated with the barbed ends and prevents their disassembly. Spire, on the other hand, nucleates actin filaments from their pointed ends and caps this end of the filament as it grows. Thus, both Capu and Spire have the capacity to nucleate and stabilize actin filaments, raising the possibility that each protein independently nucleates and stabilizes actin filaments in the mesh. This is consistent with the observation that overexpression of Capu can induce the formation of an actin mesh in the absence of Spire. The mesh induced by overexpressed Capu alone is significantly weaker than normal, however, and persists for a shorter time, while Spire D cannot form a mesh in the absence of Capu. Furthermore, the ability of GFP-CapuΔN to stabilize the actin mesh in the presence of Latrunculin depends on endogenous Spire activity. Thus, the two proteins must cooperate to form a normal mesh, and one possibility is that they assist each other by capping the opposite ends of filaments nucleated by the other. Since Spire D associates with Capu in vitro, it is even possible that they collaborate to nucleate the same filament and remain attached to opposite ends as it grows (Dahlgaard, 2007).
Although the mesh is essential for the polarized arrangement of the MTs in the oocyte, it appears to play a permissive rather than an instructive role, because the defects in MT organization and osk mRNA localization caused by its loss can be rescued by slowing the speed of kinesin. This suggests that the mesh normally serves to restrain kinesin-dependent motility and that the rearrangement of the MTs and premature streaming are a consequence of unrestricted kinesin activity. Kinesin is required both for the slow disorganized cytoplasmic movements during stage 9, called seething, and for the rapid directional streaming at stage 11, leading to the proposal that the motor generates ooplasmic flows by moving large organelles or vesicles along MTs. This suggests the following model for how loss of the actin mesh and unrestrained kinesin motility cause the rearrangement of the MT. In the absence of the actin mesh, there is an increase in the speed or frequency of kinesin-dependent organelle transport, resulting in a concomitant increase in the strength of the cytoplasmic flows that these movements generate. Since the MTs move with the cytoplasmic flows, the stronger flows will start to wash the MTs into alignment, thereby aligning the kinesin-dependent organelle movements, which will amplify the cytoplasmic flows still further. This positive-feedback loop then continues to coordinate and increase the flows until all MTs have been washed to the oocyte cortex, with the oocyte cytoplasm rapidly rotating inside (Dahlgaard, 2007).
This model raises the question of how the actin mesh restrains the kinesin-dependent cytoplasmic flows to prevent their amplification into cytoplasmic streaming. This could be an entirely passive process, in which the actin mesh increases the viscosity of the oocyte cytoplasm, thereby increasing the drag on kinesin-dependent transport. However, a model is favored in which the mesh plays a more direct role in the inhibition of kinesin-mediated movement of the cytoplasm, and one attractive possibility is that it tethers the cargoes of kinesin that generate the flows, thereby limiting their movement. One way that the organelles might be tethered to the actin mesh is by binding to either Capu or Spire, and it is interesting to note that Spire-D shows a punctate distribution that is consistent with an association with a population of organelles or vesicles. In addition, full-length Spire contains an mFYVE domain that is predicted to target it to endosomal membranes, and has been shown to colocalize with Rab11 on vesicular structures when expressed in tissue culture cells. This tethering mechanism is very similar to the function of mDia in the anchoring of endosomes to actin at the cell periphery, which inhibits their movement along MT, and also resembles the tethering of mitochondria in neuronal cells, where mDia nucleates actin filaments that anchor the mitochondria, without affecting the motility of lysosomes or peroxisomes (Dahlgaard, 2007).
A third possibility is that the mesh anchors the MTs within the cytoplasm and prevents them from being washed into alignment at the cortex by the cytoplasmic flows. It seems unlikely, however, that direct crosslinking of actin and MT by Capu and Spire is important in vivo, but some other protein may anchor the MT to actin. Alternatively, the actin and MTs could be crosslinked indirectly. For example, Capu or Spire could interact with a vesicle or organelle that is associated with MT, thereby linking the two cytoskeletons (Dahlgaard, 2007).
The formation of the actin mesh must be tightly regulated both spatially and temporally, since the mesh normally forms only in the oocyte and not the nurse cells and is disassembled during stage 10B to allow the onset of rapid streaming. Both Capu and Spire bind Rho-GTP, raising the possibility that one or both proteins are regulated by Rho). Indeed, the GFP-CapuΔN construct was generated to test if deletion of its Rho-binding domain would lead to a constitutively active form of the protein. However, overexpression of GFP-Capu or of full-length untagged Capu produces very similar effects to GFP-CapuΔN. The only obvious difference between the three constructs is the ability of CapuΔN to protect the actin mesh from Latrunculin A-induced depolymerization, but it is unclear whether this is due to constitutive activation of Capu or some other alteration to its activity. More importantly, our data suggest that the regulation of Capu activity is unlikely to determine where and when the mesh forms. Although overexpressed Capu can assemble a mesh in both the oocyte and the nurse cells in the absence of Spire until stage 10A, Spire D cannot induce the formation of an actin mesh in the absence of Capu. The ability of Spire to form an ectopic mesh in the nurse cells and in late oocytes therefore implies that endogenous Capu must be active in the nurse cells and during the late stages of oogenesis. This suggests that the regulation of Spire determines the temporal and spatial control of actin mesh formation and disassembly (Dahlgaard, 2007).
In summary, these results suggest that the Capu pathway controls the formation of an actin mesh, which acts as a switch between two alternative states of the oocyte cytoplasm, both of which are essential for the formation of a viable egg. During stages 5–10A, the mesh inhibits kinesin-dependent motility to allow the formation of the anterior-posterior MT array that directs the localization of oskar and gurken mRNAs, and this establishes the polarity of both body axes. Once oskar mRNA has been localized and anchored to the oocyte cortex and Gurken has signaled to polarize the dorsal-ventral axis, the actin mesh is disassembled. This relieves the inhibition of kinesin-dependent organelle movement and switches on fast ooplasmic streaming. As a result, the oocyte cytoplasm becomes thoroughly mixed with the cytoplasm that enters from the nurse cells during nurse-cell dumping, leading to a uniform distribution of maternal proteins and mRNAs throughout the egg. This is important for subsequent development, because most housekeeping functions in the embryo depend on maternal gene products, which must be evenly distributed in the egg, so that they are equally partitioned into all cells (Dahlgaard, 2007).
Preparations of kinesin, a microtubule-based force-producing protein, have been isolated from Drosophila melanogaster embryos by incubation of microtubules with a nonhydrolyzable ATP analogue and gel filtration of proteins released from the microtubules by ATP. These preparations induced MgATP-dependent microtubule gliding in vitro. Samples of Drosophila proteins that were active in motility assays possessed an average ATPase activity in solution of 17 nmol/min per mg that increased to an average of 106 nmol/min per mg in the presence of microtubules. The major polypeptides that copurified with these activities showed relative molecular masses of 115 kDa and 58 kDa. An antiserum raised against the 115-kDa polypeptide also recognized the 110-kDa component of squid kinesin preparations and the 130-kDa component of sea urchin kinesin preparations (Saxton, 1988).
A truncated motor domain of the alpha subunit of Drosophila kinesin was obtained by expression in Escherichia coli and purified to homogeneity in the presence of MgATP. This domain (designated DKH340) extends from the N terminus to amino acid 340. The isolated protein contains a stoichiometric level of tightly bound ADP and has a low basal rate of ATP hydrolysis in the absence of microtubules. The approximate equality of the ADP release rate and the steady state ATPase rate indicates that ADP release is the rate-limiting step in ATP hydrolysis in the absence of microtubules. The rate of ATP hydrolysis is stimulated 3000 fold-by addition of microtubules. One DKH340 binds tightly per tubulin heterodimer, but greater than one DKH340/tubulin heterodimer can be bound at higher ratios of DKH340/microtubules. These results are consistent with a model in which DKH340 cycles on and off the microtubule during hydrolysis of each ATP molecule (Huang, 1994a).
The DKH392 construct includes an additional 52 amino acids beyond the minimal motor domain of 340 amino acid residues, previously characterized as DKH340. DKH340 is a monomer in solution, but DKH392 is a dimer. In the presence of adenosine 5-(beta,gamma-imido)triphosphate, DKH392 binds to microtubules with a stoichiometry of two head domains (one DKH392 dimer) per tubulin heterodimer, in contrast to the tight binding of one DKH340 per tubulin heterodimer. Electron microscopy indicates that both DKH340 monomers and DKH392 dimers decorate microtubules with a periodicity of 8 nm (Huang, 1994b).
Kinesin is a heterotetramer composed of two 115-kD heavy chains and two 58-kD light chains. The microtubule motor activity of kinesin is performed by the heavy chains, but the functions of the light chains are poorly understood. Mutations were generated in the Drosophila gene Kinesin light chain (Klc), and the phenotypic consequences of loss of Klc function were analyzed at the behavioral and cellular levels. Loss of Klc function results in progressive lethargy, crawling defects, and paralysis followed by death at the end of the second larval instar. Klc mutant axons contain large aggregates of membranous organelles in segmental nerve axons. These aggregates, or organelle jams (Hurd, 1996b), contain synaptic vesicle precursors as well as organelles that may be transported by kinesin, kinesin-like protein 68D, and cytoplasmic dynein, thus providing evidence that the loss of Klc function blocks multiple pathways of axonal transport. The similarity of the Klc and Khc (Saxton, 1991; Hurd, 1996b) mutant phenotypes indicates that KLC is essential for kinesin function, perhaps by tethering KHC to intracellular cargos or by activating the kinesin motor (Gindhart, 1998).
The phenotypes of mutations in Drosophila Khc have been extensively studied. The Khc mutant phenotype is characterized by locomotion defects, progressive paralysis, and death during the larval stage of development. Paralysis is more severe at the distal (posterior) end of the larva, suggesting that the long motor axons innervating the posterior body wall muscles are more severely affected than the shorter axons of anterior segments. These phenotypes are caused by impairment of neuron function, as evidenced by electrophysiological defects including reduction of compound action potentials and the amplitude of excitatory junction currents, as well as reduction of the number of boutons at the neuromuscular junction. Light and electron microscopic analysis suggests that transport of a variety of cargos, both anterograde and retrograde, is blocked in Khc mutant larvae, leading to the formation of axonal swellings that accumulate diverse cargos (Gindhart, 1998).
The phenotype of null Klc mutations is quite similar to the Khc null mutant phenotype. KLC protein accumulation is observed at all stages of development; a single predominant species of 58 kD is observed. The presence of KLC at all stages of development suggests that KLC function may be required for developmental processes. However, embryonic development appears normal, and first instar larvae hatch apparently unhindered by the loss of Klc function. The lack of zygotic KLC accumulation during embryogenesis and the first larval instar is presumably compensated by a maternally supplied pool of KLC protein. However, the second larval instar is characterized by a progressive loss of vigor as the larvae exhibit increasing amounts of paralysis, eventually resulting in complete paralysis and death near the end of the second larval instar. Occasionally a Klc null individual can proceed to the third larval instar, but these escapers often lack the strength to shed their second instar cuticle. Similar to Khc mutants, progressive paralysis associated with loss of Klc function begins at the posterior end of the larva and proceeds anteriorly. It is thus likely that paralysis of Klc mutants, like Khc, reflects the differential requirement for kinesin-based transport in the longer axons of the posterior segments relative to the shorter axons innervating the anterior segmental muscles (Gindhart, 1998).
Combinations of less severe Klc mutations live to the late third larval instar or pupal stage of development. Like the null Klc mutants, the terminal phenotype of Klc hypomorphic alleles is larval paralysis, with the exception of weak mutants that pupate but fail to eclose. In addition, partial loss of Klc function often results in unusual locomotion defects. Wild-type larvae move along a surface by rhythmic contractile waves that originate at the posterior end and move in a concerted fashion toward the anterior. These contractile waves are facilitated by subcuticular body wall muscles controlled by motor axons whose cell bodies are located in the larval CNS. Normally, the dorsal and ventral muscles of each segment contract in unison, ensuring that the larva maintains contact with the surface. However, Klc mutations that allow survival to late third larval instar cause the posterior end of the larva to lift its tail off the medium. Severely affected individuals can be observed lifting the posterior 40% of their bodies. This tail-flipping phenotype is also observed in certain Khc mutant combinations (Hurd, 1996b). It has been proposed that the tail-flipping observed in Khc mutants is the result of a temporal gradient of paralysis such that the ventral body wall muscles lose muscle tone before dorsal body wall muscles. Contraction of the dorsal body wall muscles in the absence of counterbalancing ventral muscle contraction then causes the tail to flip upward. The tail-flipping and paralysis phenotypes observed in Klc mutant larvae are quantitatively rescued by the transgenic constructs GEN-KLC and MYC-KLC, providing strong evidence that the neuromuscular defects of Klc mutants result from the loss of Klc function (Gindhart, 1998).
Macromolecular structures required for synapse function, such as membranous vesicles, neurotransmitters, and the machinery controlling synaptic vesicle fusion and recycling, must be transported great distances from the point of synthesis in the cell body to the synapse. Similarly, for chemical signals received at the synapse to be acted upon by the neuron, the signals must be transported from the synapse to the cell body. Microtubule motors such as kinesin are necessary for these transport phenomena. Perhaps the paralytic and tail-flipping phenotypes associated with loss of Klc function are the result of a defect in axonal transport owing to mislocalization of kinesin cargoes. In fact, loss of Khc function is known to cause massive disruption of fast axonal transport, resulting in focal swellings packed with many different types of membrane-bound organelles, including mitochondria, prelysosomal mutivesicular bodies, and synaptic vesicle precursors. Loss of Khc function does not disrupt slow axonal transport, however, as molecules that undergo slow axonal transport, such as tubulin, are not observed in Khc organelle jams (Hurd, 1996b and Gindhart, 1998).
To determine if loss of Klc function results in formation of axonal organelle jams similar to those observed in Khc mutants, the cellular phenotype of Klc mutant axons of tail-flipping Klc mutant larvae was studied. Biochemical analyses in mammalian systems suggest that kinesin is associated with many different types of membranous vesicles, including synaptic vesicles and mitochondria. Therefore, the transport of proteins such as the synaptic vesicle components synaptotagmin (SYT) and cysteine string protein (CSP), which undergo high levels of microtubule-based axonal transport, was studied. Segmental nerves from control Klc/+ larvae exhibit punctate but relatively uniform CSP and SYT staining in the segmental nerves. However, large immunoreactive clusters of CSP and SYT are observed in the segmental nerves of Klc mutant larvae. These clusters appear to represent organelle jams similar to those observed in Khc mutant larvae (Hurd, 1996b). Small clusters of immunoreactivity are sometimes observed in control larvae, but their size and frequency are greatly reduced in comparison with Klc mutant segmental nerves. Immunostaining of Khc mutant larvae exhibiting the tail-flipping behavioral phenotype reveals similar staining patterns for SYT and CSP (Hurd, 1996b). These results suggest that, in addition to sharing behavioral phenotypes, Khc and Klc mutations appear to cause a similar disruption of fast axonal transport. Furthermore, the similar phenotype of Klc and Khc mutants suggests that both the light chains and heavy chains are necessary for kinesin function, and that the light chains may have a positive role in kinesin function (such as cargo binding) instead of being a negative regulator of kinesin activity (Gindhart, 1998).
The observation that organelle jams from Khc mutant larvae contain a variety of membranous organelles, including both anterograde and retrograde cargoes (Hurd, 1996b), led the authors to test directly the hypothesis that failure of kinesin transport in Klc mutants interferes with the ability of other molecular motors to transport their cargoes. To determine if Klc mutations interfere with transport pathways of other motors, immunolocalization was examined in two other motor proteins that are likely to play important roles in fast axonal transport. The first of these motors is KLP68D, which is a plus-end directed neuronal motor homologous to members of the murine KIF3 family and the sea urchin KRP85/95 family. The second motor examined, DHC, is the heavy chain of the retrograde fast axonal transport motor cytoplasmic dynein. KLP68D is obviously localized to large aggregates in Klc mutant larvae, whereas its localization in Klc/+ control segmental nerves is fairly uniform, suggesting that the presence of KLP68D aggregates is a direct consequence of loss of Klc function. Some of these aggregates colocalize with CSP, demonstrating that KLP68D protein is present in organelle jams. Identification of organelle jams that contain KLP68D but not CSP suggests that the composition of the membranous organelles in jams is heterogeneous. This heterogeneity may reflect the stochastic nature of organelle jam formation, or could be due to the presence of both sensory and motor axons in segmental nerves, each accumulating different ratios of intracellular cargos. While the degree of impairment of KLP68D-dependent transport in Klc mutant axons cannot be directly measured, the presence of high levels of KLP68D immunoreactivity in organelle jams suggests that loss of kinesin function causes a pleiotropic block of anterograde fast axonal transport (Gindhart, 1998).
To test whether the organelle jams found in Klc mutant axons may impede retrograde fast axonal transport, a comparison was made of immunolocalization of DHC in control and Klc mutant larvae. DHC is observed in large aggregates in Klc mutants, but not in Klc/+ control larvae, and the DHC aggregates sometimes colocalize with SYT. However, many DHC-labeled organelle jams do not colocalize with SYT-labeled organelle jams, suggesting again that the distribution of intracellular cargos within organelle jams is heterogeneous. This result demonstrates that cytoplasmic dynein is also present in organelle jams, and supports the hypothesis that loss of kinesin function inhibits multiple pathways of motor-driven fast axonal transport. However, the possiblity that Klc mutations disrupt fast retrograde axonal transport cannot be excluded. Cytoplasmic dynein is a fast retrograde motor in axons, but much cytoplasmic dynein in axons may be in an inactive state, transported by anterograde fast axonal transport to the synapse. Whether or not the DHC in organelle jams is in an active state cannot be examined; therefore, additional experiments are required to determine unequivocally the relationship between loss of Klc function and disruption of cytoplasmic dynein-dependent transport pathways. In conclusion, these results strongly suggest that behavioral defects observed in Khc and Klc mutant larvae are not solely due to a failure to localize specific kinesin cargos, but result from a general failure of several pathways of motor-mediated axonal transport (Hurd, 1996b; Gindhart, 1998).
An important question is whether the kinesin light chains are necessary for the binding of kinesin to its intracellular cargoes. Whereas some results suggest that the KHC tail domain is sufficient for kinesin-cargo interaction (Skoufias, 1994; Bi, 1997), other experiments demonstrate that KLC function may be necessary for a subset of kinesin binding to intracellular cargoes (Stenoien, 1997). Attempts were made to test directly the role of KLC in cargo binding by studying immunolocalization of KHC in Klc mutant axons. The absence of KHC from Klc mutant organelle jams would support a model in which KLC is necessary for cargo attachment. In contrast, the observation of KHC immunoreactivity in the absence of KLC would strongly suggest that some kinesin-cargo interactions are KLC-independent. Immunostaining of Klc1/ Df(3L)8ex94 mutant larvae with antisera to KHC and CSP demonstrates that KHC is present in Klc larval organelle jams. This result suggests that KLC is dispensable for kinesin-cargo interactions; however, it is possible that KHC is tethered to cargoes by residual light chains present in Klc mutant larvae. The allelic combination used for KHC immunolocalization, Klc1/Df(3L)8ex94, makes low but detectable levels of KLC protein. KLC protein, like KHC, is also observed in Klc mutant larvae, and its immunoreactivity partially overlaps CSP. Biochemical fractionation experiments also demonstrate that nearly all the residual KLC protein in light chain mutants is membrane-bound, suggesting attachment to organelle cargoes. Although these results support a role for KLC in kinesin-cargo interactions, additional experiments such as immunoelectron microscopy will be required to define the relative contributions of KLC and KHC to cargo attachment at high resolution, and to demonstrate the direct association of KHC with membranous organelles in Klc-dependent organelle jams (Gindhart, 1998).
The hypothesis was tested that amyloid precursor protein (APP) and its relatives function as vesicular receptor proteins for kinesin-I. Deletion of the Drosophila APP-like gene (Appl) or overexpression of human APP695 (an alternatively spliced version of APP) or APPL constructs causes axonal transport phenotypes similar to kinesin and dynein mutants. Genetic reduction of kinesin-I expression enhances while genetic reduction of dynein expression suppresses these phenotypes. Deletion of the C terminus of APP695 or APPL, including the kinesin binding region, disrupts axonal transport of APP695 and APPL and abolishes the organelle accumulation phenotype. Neuronal apoptosis was induced only by overexpression of constructs containing both the C-terminal and Ab regions of APP695. The possibility is discussed that axonal transport disruption may play a role in the neurodegenerative pathology of Alzheimer's disease (Gunawardena, 2001).
Although reducing the amount of kinesin-I to 50% of normal by deleting one of two copies of either the klc or khc gene ordinarily has no significant phenotype, such a reduction in an animal overexpressing APP proteins that contain the cytoplasmic C terminus is predicted to significantly enhance the axonal blockage phenotype. This behavior is expected, because if kinesin-I becomes limiting by virtue of binding excess APP C termini, then further reduction of kinesin-I by deleting one gene copy should dramatically enhance the axonal transport phenotype. To test this prediction, larvae were generated that overexpressed APPL or APP695 and that were also heterozygous for a null mutation, khc8, so that kinesin-I was reduced to 50% of normal. Although larvae overexpressing APPL or APP by themselves or reduced in KHC dosage alone have no striking organismal phenotype, larvae combining these two features exhibit a dramatic new neuromuscular phenotype. These larvae flip their tail and head upwards during crawling, rocking back and forth as they struggled to crawl. Their neurons also contained an enhanced number of organelle accumulations. The extent of accumulations in larvae expressing wild-type APP695 and APPL in the context of reduced KHC dosage was comparable to homozygotes for kinesin-I or dynein mutants and was similarly lethal. Quantitative analysis has revealed a statistically significant difference between siblings with a normal and reduced dose of KHC (Gunawardena, 2001).
To confirm the specificity of the genetic interactions observed between reduction in KHC and overexpression of APP695 or APPL, larvae heterozygous for a null mutation of klc [Df(3L)8ex94, which removes the entire kinesin light chain gene] were generated in combination with constructs expressing APPL and APP695. Neurons from these larvae contain an increased number of organelle jams relative to larvae with a normal dose of KLC. Quantitative analysis reveals a statistically significant difference between siblings containing a normal and reduced dose of KLC, although the extent of enhancement is not as dramatic as when the dosage of KHC is reduced. Although these larvae do not show a larval neuromuscular phenotype as dramatic as was observed when the dosage of KHC was reduced, these larvae show a clear posterior paralysis phenotype. This result is again consistent with a direct functional interaction of the C-terminal region of APP695 and APPL with kinesin-I (Gunawardena, 2001).
Although reducing the amount of cytoplasmic dynein to 50% of normal by deleting one of two copies of either dynein heavy chain (dhc) or dynein light chain (dlc) genes ordinarily has no significant phenotype, such a reduction in an animal overexpressing APP family members that contain the cytoplasmic C terminus is predicted to suppress significantly the severity of the axonal accumulation phenotype. The basis for this prediction is that dynein drives retrograde axonal transport, which is antagonistic to kinesin-I-mediated anterograde axonal transport. In addition, many vesicles or organelles that exhibit net anterograde movement experience periods of retrograde movement owing to the simultaneous presence of kinesins and dyneins on the same vesicle or organelle. Thus, vesicle stalling and axonal accumulations induced by APP are predicted to be ameliorated by dynein reduction by (1) reducing the rate at which vesicles and organelles moved by dynein are transported into regions that have stalled or accumulated vesicles caused by APP expression; or (2) reducing the contribution of dynein-driven movement to a vesicle experiencing stalling because of reduced kinesin-driven activity; this reduction should attenuate vesicle stalling by restoring the balance of movements toward normal (Gunawardena, 2001).
To test this prediction, larvae overexpressing APP695 or APPL were generated that were also heterozygous for either a deficiency of either dhc [Df(3L)GN24] or dlc (roblk). Reduction of dynein suppresses the extent of organelle accumulations in APP695 or APPL transgenic lines. In addition, no significant larval crawling phenotype was observed in these animals. Surprisingly, reduction in dynein dosage also rescues the inviability of males overexpressing APP695. No effect is seen in the lines expressing a C-terminal deletion of APP695 or APPL. Thus, reduction in dosage of a retrograde motor protein appears to be sufficient to decrease organelle accumulations induced by APP or APPL expression (Gunawardena, 2001).
Whether larvae bearing heterozygous deletions of dynein components and the Appl gene in combination would exhibit abnormal axonal transport was investigated. In contrast to the situation with kinesin and APPL, female larvae with one copy (50% dose) of Appl and one copy of dhc or dlc do not have typical axonal accumulations. Intriguingly, male larvae bearing a deletion of the Appl gene, and hence lacking all APPL function, in combination with one copy of dhc or dlc also lack axonal accumulations. Thus, reduction in dynein levels appears to suppress axonal accumulations induced by loss of APPL (Gunawardena, 2001).
Genetic analysis in Drosophila strongly supports the hypothesis that mammalian APP and its homolog, APPL, have kinesin-I receptor functions in vivo. The genetic data and tests complement the in vitro biochemical evidence for a kinesin-I receptor function for APP. In these experiments, APP has been shown to form a complex with conventional kinesin by directly binding to KLC. Transport of APP depends upon kinesin-I and KLC in particular. In addition, the finding that, in Drosophila, APP695 can enter and be transported down axons to neuromuscular junctions and that this transport depends upon the cytoplasmic C terminus containing the proposed KLC binding region supports this view. In toto, these data strongly support the hypothesis that APP functions as a kinesin-I receptor. Perhaps APP bound to kinesin-I may be required for the axonal transport of a subset of cargoes, such as vesicles containing signaling or other molecules used at the synapse. Identifying these vesicles and their cargoes is an important next step (Gunawardena, 2001).
A surprising finding is the suppression of APP and APPL-induced organelle accumulations by genetic reduction of cytoplasmic dynein. Although further work is needed to define the mechanism of this suppression, one simple explanation comes from previous observations about the functionally antagonistic relationship of dynein and kinesin. A general observation is that kinesin and dynein are both present on many of the same axonal vesicles and organelles. Such vesicles and organelles often exhibit alternating anterograde (kinesin) and retrograde (dynein) movements, with net anterograde or retrograde movement resulting from a regulated bias in the balance of opposing movements along the microtubule. Thus, reduction of kinesin-I on non-APP vesicles caused by binding of kinesin-I to excess APP might cause vesicle stalling and organelle accumulations. Stalling of these vesicles and subsequent phenotypes might be rescued by reducing the antagonistic component of movement produced by dynein (Gunawardena, 2001).
An important related issue is whether APP and APPL have functions in addition to their likely roles as kinesin-I receptors on vesicles. Thus, one extreme possibility is that all phenotypic effects caused by genetic manipulation of APPL or APP result from either titration of available kinesin-I or failure to deliver other components of vesicles whose movement depends upon APP or APPL. However, a number of observations support the view that APP and APPL may have additional functions, perhaps mediated by the extracellular domain of the protein. For example, the secreted form of APP has been implicated in the regulation of hemostasis and neuroprotection, while the intact molecule may be involved in cell-extracellular matrix adhesion and in the sequestration of potentially toxic transition metals. There is also evidence that APPL may have a role in synapse differentiation. Thus, APPL and APP may have additional roles at the nerve terminal following their role in the axon as receptors for kinesin-I-dependent transport (Gunawardena, 2001).
It is also striking that several other proteins thought to play a role in Alzheimer's disease have recently been linked to the axonal transport machinery. For example, axonal transport defects are observed in transgenic mice expressing human ApoE4, a gene whose allelic state is associated with increased risk of AD. In the axonal blockages found in these animals, accumulations of synaptophysin, neurofilaments, mitochrondria, and vesicles are seen. Similarly, overexpression of tau protein, a major component of neurofibillary tangles, has been proposed to inhibit kinesin-I-dependent trafficking of vesicles, mitochondria, and endoplasmic reticulum in cultured cells and the PNS of transgenic mice. It is also striking that JIP-1/2 proteins, which are scaffolds for components of JNK signaling pathways, may link kinesin-I to ApoER2 and LRP, which are receptor proteins for ApoE. The allelic state of LRP itself has also been reported to be a predisposing factor for AD. Thus, it is conceivable that not only motor proteins but their cargoes and receptors, such as APP, ApoE, and perhaps secretases and other signaling molecules, coaccumulate when axonal transport is impaired in any way. Accumulating these proteins together at the same site, at the same time, may by itself be neurotoxic, may cause induction of cellular suicide signals, may block neurotrophic and other signaling needed for neuronal viability, or may lead to biochemical changes causing excess production of Aß, any or all of which may lead to cell death. Whether these correlations are unrelated or truly indicative of a causative link remains to be tested (Gunawardena, 2001).
Pathogenic polyQ proteins cause axonal transport defects and neuronal apoptosis:
To address whether axonal transport defects are selective to the pathogenic Huntingtin protein, or whether they are a feature of polyQ proteins in general, the effects were examined on axonal transport of various proteins containing polyQ tracts of different lengths and in different contexts. 179Y-GAL4 and APPL-GAL4 were crossed to lines encoding proteins with either a 'normal' length, nondisease-causing polyQ repeat region or proteins with an expanded, disease-causing polyQ repeat region. These proteins consisted either entirely of polyQ repeats (20Q, 22Q, 108Q, 127Q) or polyQ repeats embedded in the C-terminal region of the polyQ disease protein Machado-Joseph disease (MJD) protein (MJD-27Q, MJD-78Q). Proteins with normal length polyQ regions (22Q, MJD-27Q) were found to be present within axons, based upon the smooth staining seen with either an antibody against polyQ or an antibody against the HA tag, and these proteins were found to accumulate at neuromuscular junctions, suggesting that they are normally transported within larval axons. In contrast, proteins with expanded polyQ repeats (MJD-78Q, 127Q) exhibited prominent polyQ staining within organelle blockages, while reduced staining was observed at the neuromuscular junctions, suggesting impaired transport of pathogenic polyQ proteins (Gunawardena, 2003).
The extent of axonal accumulations induced by polyQ repeats was length dependent, since a correlation was observed between the number of polyQ repeats and the amount of axonal accumulations. Larvae expressing 20Q, 22Q, or MJD-27Q were similar to wild-type in that they exhibited no axonal accumulations. Larvae expressing the pathogenic proteins MJD-78Q, 108Q, or 127Q exhibited a severe sluggish larval movement phenotype, with prominent axonal accumulations observed in all instances. The expression of MJD-78Q and 127Q at 29°C was also observed to be very toxic such that larvae expressing these proteins never survived to adulthood and died at second or third instar larval stage. Western blot analysis ruled out a general expression difference between proteins with normal length polyQ repeats and proteins with expanded polyQ repeats since, if anything, more MJD-27Q expression was observed compared to MJD-78Q. To reduce the level of toxicity, the amount of MJD-78Q and 127Q made was reduced by growing animals at 25°C (GAL4 activity is temperature dependent). Organelle accumulations were still reduced although these larvae now survived much longer, dying at early pupal stages (Gunawardena, 2003).
To confirm the results seen by immunofluorescent staining, EM analysis was conducted on larvae expressing 127Q and on severe genotypes in which expression of MJD-78Q or 127Q was combined with a 50% reduction in KHC gene dose. Prominent axonal blockages were observed characteristic of those observed in homozygous mutations of motor protein genes. Mutant larval nerves also contained enlarged axons, some almost four or five times the diameter of those observed in wild-type. Sometimes 'holes' were observed lacking organelles within the nerve, perhaps indicative of degeneration (Gunawardena, 2003).
To test directly whether pathogenic polyQ proteins block transport by inducing nonmoving blockages in axonal processes, live analysis was performed of vesicular movement within whole-mount larval axons. YFP-tagged human amyloid precursor protein (APP-YFP) was expressed either in the presence or absence of MJD-78Q, using the GAL4 driver pGAL4-62B SG26-1, which is expressed in only a small population of motor neurons. Neurons expressing only APP-YFP contained many actively motile vesicles moving at velocities of approximately 1 microm/s; large bright nonmotile accumulations such as those seen in the presence of MJD-78Q were never observed. Neurons coexpressing MJD-78Q and APP-YFP revealed nonmoving large, bright aggregates of APP-YFP. Thus, polyQ expression can interfere with transport of APP-YFP vesicles (Gunawardena, 2003).
TUNEL analysis was used to test whether proteins with expanded polyQ repeats can induce neuronal apoptosis. A large increase in neuronal apoptosis was observed in lines expressing MJD-78Q and 127Q, but not in lines expressing the nonpathogenic control proteins 22Q or MJD-27Q. Anti-polyQ staining revealed obvious nuclear inclusions within larval brain cells. Many stained nuclei were obviously enlarged and may be undergoing apoptosis. Expression during embryonic cycle 14 revealed smooth cytoplasmic staining for MJD-78Q protein, while staining for 127Q revealed obvious punctate cytoplasmic aggregates with some nuclear aggregates. At later cycles, both MJD-78Q and 127Q were observed as punctate cytoplasmic and nuclear aggregates. In addition, embryos expressing both MJD-78Q and 127Q died soon after they hatched into larvae, indicating substantial polyQ toxicity on normal development. Taken together, these results confirm that proteins with expanded polyQ repeats cause axonal transport defects, perhaps by blocking axonal processes by polyQ accumulations, neuronal cell death, and neurodegeneration (Gunawardena, 2003).
It is possible that polyQ proteins bind and deplete critical components of the molecular motor machinery, perhaps via a Drosophila version of HAP1. This hypothesis makes two predictions: (1) genetic reduction of motor protein dosage should worsen the phenotypes caused by proteins with polyQ expansions by further depleting the motor protein supply, and (2) motor protein depletion should be observable with biochemical methods. To test this hypothesis, proteins with expanded polyQ repeats were expressed and levels of dynein and kinesin were reduced. While a 50% reduction in the dose of KHC has no significant phenotype on its own, when combined with pathogenic polyQ repeats, it dramatically enhances the axonal organelle accumulation phenotype. Similarly, while a 50% reduction in the dose of DLC or components of the dynactin complex (p150Glued, Arp1, and dynamitin) also normally have no significant phenotypes on their own, when combined with pathogenic polyQ proteins, these reductions substantially enhance the organismal phenotype leading to early larval lethality. This finding is consistent with the observation that the neuronal APPL-GAL4 driver turns on during embryonic stage 15 as observed by the expression pattern of UAS-GFP. The enhanced lethality precluded analysis of axonal transport in these genotypes. Interestingly, organelle accumulations now appeared in transgenic lines expressing normally nonpathogenic poly Q repeats (22Q and MJD-27Q) with a 50% reduced dose of dynein, consistent with the hypothesis that all of these proteins may titrate motor proteins, but to varying extents. To test directly for motor protein depletion mediated by expression of proteins with expanded polyQ regions, early embryos expressing httex1-20Q (Huntingtin with non-expanded polyQ repeats), MJD-27Q, MJD-78Q, httex1-93Q, and 127Q (Huntingtin proteins with expanded polyQ repeats), were examined using the early embryonic GAL4 driver da-GAL4, which turns on at the blastoderm stage based on its UAS-GFP expression pattern (Gunawardena, 2003).
Two considerations led to an evaluation of the effects of polyQ proteins on available motor protein pools by assessing soluble levels of motor proteins: (1) if motor proteins are titrated from normal cargoes by binding to large aggregates, it may be difficult to distinguish motor proteins bound to sedimentable cargoes from sedimentable aggregates; (2) it is not possible to measure the amount of each motor protein associated with normal cargoes owing to the lack of information about such cargoes and how such cargoes fractionate relative to polyQ aggregates. Thus, soluble levels of motor proteins were assessed under the hypothesis that aggregated polyQ proteins may bind motor proteins and deplete both soluble and cargo bound pools in parallel (Gunawardena, 2003).
At 6 hr of development, no significant change in the amount of total motor protein present in these embryos was observed. In contrast to the normal amounts of total motor proteins, an obvious reduction was observed in the amount of soluble motor proteins in embryos expressing MJD-27Q, MJD-78Q, and 127Q compared to wild-type embryos (i.e., da-GAL4 alone and yw). The amount of soluble DHC, DIC, p150Glued, KHC, and KLC were reduced, with no change observed in tubulin, actin, HDAC3, and Rab8. However, for reasons that are not clear, syntaxin was upregulated. Similar observations were evident from 12 and 16 hr embryo collections. The effect of httex1-93Q expression on levels of soluble motor proteins was not obvious at 6 hr of development. However, at 18 hr of development, there is an obvious reduction in soluble p150Glued and KLC in embryos expressing httex1-93Q but not httex1-20Q or wild-type. Expression of polyQ proteins in embryos was obvious as detected by anti-HA antibody and confirmed by anti-polyQ antibody. The level of 127Q was difficult to evaluate, perhaps due to the formation of aggregates, and was convincingly observed only after immunoprecipitation with anti-HA antibody. These observations indicate that expanded polyQ proteins can deplete or sequester available soluble motor proteins, perhaps into polyQ aggregates. The high expression level of MJD-27Q in embryos and the observed depletion of soluble motor proteins in these embryos are consistent with the finding that axonal blockages can be observed in larvae expressing MJD-27Q when motor protein gene dose is reduced by 50%. This finding is also consistent with the proposal that motor titration and aggregation may act in concert to poison axonal transport (Gunawardena, 2003).
Enhanced expression of chaperones restores neuronal transport and suppresses cell death caused by pathogenic polyQ proteins
The neurodegenerative adult eye phenotype caused by polyQ expansion proteins in Drosophila is suppressed by excess chaperone proteins (Warrick, 1999). This suppression has been proposed to occur by modulating soluble properties of pathogenic polyQ proteins, by preventing abnormal interactions with other proteins, or by rescuing chaperone depletion (Bonini, 2002). Whether expression of excess HSC70 protein would suppress axonal blockages and neuronal cell death was tested. Expression of UAS-HSC70 using APPL-GAL4 in the presence of MJD-78Q and 127Q restores axonal transport within larval nerves and suppresses neuronal death. PolyQ accumulations were absent within larval nerves, while cytoplasmic and punctate polyQ staining was present within larval brains. However, while these larvae were now able to pupate (expression of MJD-78Q or 127Q alone causes death at second or third instar larval stages), they still failed to eclose, suggesting that polyQ toxicity was still sufficient to cause lethality. Expression of HSC70 by itself did not cause axonal blockages or neuronal cell death. These results suggest that chaperones could 'clear' larval axons of blockages caused by polyQ proteins and suppress cell death within the larval brain, although organismal toxicity was not completely suppressed (Gunawardena, 2003).
Do axonal defects instigate neuronal dysfunction?
The pathogenic polyQ proteins accumulate in both axonal and nuclear inclusions. To dissect the relative contributions of nuclear and axoplasmic inclusions to the phenotype, transgenes were used that expressed proteins that had different subcellular localizations.
One transgene encoded a protein with an expanded polyQ repeat with a nuclear localization sequence (MJD-65QNLS). While expression of MJD-65QNLS within the larval brain caused neuronal apoptosis as observed by TUNEL staining, organelle accumulations within larval axons were absent. These larvae pupated but failed to eclose. While reduction in dynein dose by 50% with excess MJD-65QNLS had no effect, reduction in kinesin dose by 50% with excess MJD-65QNLS caused a small number of accumulations, perhaps due to continued motor titration by these proteins even when targeted to nuclei (Gunawardena, 2003).
To test further if dying neuronal cells induce axonal transport defects, the axonal transport phenotype was analyzed of the cell death gene reaper, which also induces neuronal apoptosis. Transport following reaper expression in these genotypes appeared to be normal based on immunostaining with synaptic vesicle markers even though high levels of neuronal apoptosis were induced. In addition, a 50% reduction in KHC combined with excess reaper expression had no effect on axonal transport. These findings emphasize that not all neuronal death is associated with axonal accumulations, and that the axonal transport defects induced by pathogenic polyQ proteins may be specific to cytoplasmic aggregations of expanded polyQ proteins (Gunawardena, 2003).
To test for cytoplasmic or axoplasmic toxicity, a protein with an expanded polyQ repeat with a nuclear export sequence (MJD-77QNES) was studied. Expression of MJD-77QNES within larval neurons caused large numbers of synaptotagmin-containing organelle accumulations within larval nerves. Consistent with primarily cytoplasmic localization of this protein, polyQ/HA staining was absent from cell nuclei within the larval brain, with bright anterior staining (just distal to the brain) present within larval nerves. Neuronal apoptosis as determined by TUNEL staining was completely absent and these larvae died at second or third instar, similar to MJD-78Q. Quantitative analysis indicates that the extent of organelle accumulations within MJD-77QNES is comparable to accumulations observed in mutations of motor proteins, suggesting that perhaps the extent of accumulations causes lethality. Similar to MJD78Q, a 50% reduction in the dose of KHC with MJD-77QNES enhanced organelle blockages, while 50% reduction in DLC combined with MJD-77QNES caused early larval lethality. Additionally, expression of MJD-77QNES in the adult eye using GMR-GAL4 caused a severe degenerative eye phenotype, indicating that cytoplasmic polyQ protein can cause degeneration of adult neurons (Gunawardena, 2003).
To directly test if excess MJD-77QNES sequestered motor proteins, embryos expressing MJD-27Q, MJD-78Q, and MJD-77QNES were compared with wild-type (da-GAL4). While embryos expressing both MJD-78Q and MJD-77QNES showed normal levels of total motor proteins, they exhibited a striking reduction in the amount of soluble motor proteins. MJD-77QNES also exhibited high molecular weight aggregates, which could be immunoprecipitated with the HA antibody. These high molecular weight aggregates also contained sequestered DHC. Although phenotypically normal when expressed in larvae, MJD-27Q appears to titrate more DHC into high molecular weight aggregates than do MJD-78Q, MJD-77QNES, or 127Q. It is conceivable that pathogenic MJD-78Q, MJD-77QNES, and 127Q form high molecular weight aggregates that were not possible trap or to solubilize using current protocols. Indeed, dramatic phenotypes are only observed in MJD-78Q, MJD-77QNES, and 127Q. Additionally, similar to embryos expressing MJD-78Q or 127Q, embryos expressing MJD-77QNES died soon after they hatch into larvae, indicating significant polyQ toxicity on normal development. It is possible that the 'soluble' polyQ aggregates observed in embryos expressing MJD-77QNES represent a subclass of misfolded proteins, while the class of insoluble aggregates, which were not possible to observe directly on SDS-PAGE, may be responsible for polyQ toxicity (Gunawardena, 2003).
To distinguish if MJD-78Q blockages and MJD-77QNES blockages are comparable, whether blockages caused by MJD-77QNES expression can be suppressed by excess HSC70 was tested. Expression of MJD-77QNES with excess HSC70 completely suppressed axonal accumulations, polyQ aggregates, and rescued larval lethality to pupae, suggesting that axonal blockages caused by either MJD-78Q or MJD-77QNES were comparable. PolyQ-containing accumulations were also absent in larval nerves. However, excess HSC70 was not sufficient to suppress organismal lethality (Gunawardena, 2003).
In a genetic screen for Kinesin heavy chain (Khc)-interacting proteins, APLIP1, a neuronally expressed Drosophila homolog of JIP-1, a JNK scaffolding protein (Taru, 2002), was discovered. JIP-1 and its homologs have been proposed to act as physical linkers between kinesin-1, which is a plus-end-directed microtubule motor, and certain anterograde vesicles in the axons of cultured neurons (Verhey, 2001). Mutation of Aplip1 causes larval paralysis, axonal swellings, and reduced levels of both anterograde and retrograde vesicle transport, similar to the effects of kinesin-1 inhibition. In contrast, Aplip1 mutation causes a decrease only in retrograde transport of mitochondria, suggesting inhibition of the minus-end microtubule motor cytoplasmic dynein (Pilling, 2005). Consistent with dynein defects, combining heterozygous mutations in Aplip1 and Dynein heavy chain (Dhc64C) generate synthetic axonal transport phenotypes. Thus, APLIP1 may be an important part of motor-cargo linkage complexes for both kinesin-1 and dynein. However, it is also worth considering that APLIP1 and its associated JNK signaling proteins could serve as an important signaling module for regulating transport by the two opposing motors (Horiuchi, 2005).
To identify proteins that influence kinesin-1-based axonal transport, genetic interaction tests were done to search for mutations that act as dominant enhancers of Kinesin heavy chain. A number of such E(Khc) mutations were found that caused synthetic axonal transport phenotypes (i.e., larval paralytic 'tail flipping' and organelle-filled 'axon swellings') when combined with a Khc null (Khc27/+; E(Khc)/+). Tail flipping was not seen and swellings were rare in Khc27/+ or E(Khc)/+ single heterozygotes. A subset of E(Khc) loci cause tail flipping and swellings when homozygous mutant in a wild-type Khc background, suggesting that the products of those loci have direct roles in axonal transport. That subset includes Kinesin light chain (Klc), Dynein heavy chain 64C (Dhc64C), Glued and an unknown locus on chromosome 3 initially designated E(Khc)ek4 (abbreviated as ek4) (Horiuchi, 2005).
To gain more insight into the functions of ek4 products, a number of phenotypic tests were done. Homozygous ek4 mutant larvae showed classic posterior paralysis and axonal swelling phenotypes with severities similar to those caused by strong hypomorphic Khc genotypes. However, in contrast to such Khc mutants, which die during larval and pupal stages of development, ek4 mutants survive to become active, fertile adults. Severity comparisons with a null [Df(3L)Fpa2] indicate that the ek4 mutation is a strong hypomorphic allele, causing nearly a complete loss of function. These observations suggest that wild-type products of the ek4 locus have important axonal transport functions in larvae and that they have a positive functional relationship with kinesin-1. However, ek4 is not itself essential, suggesting that its products contribute to only a subset of kinesin-1 functions (Horiuchi, 2005).
To test the effects of ek4 mutations on kinesin-1-dependent fast axonal transport, time-lapse confocal microscopy was used. GFP-neuronal synaptobrevin (GFP-nSyb) was used to image transport vesicles, while cytochrome c oxidase-GFP (mito-GFP) (Pilling, 2005) was used to image mitochondria. These constructs were expressed in motoneurons of larvae by virtue of Gal4-UAS promoters that were activated by P[GawB]D42-Gal4 (abbreviated D42), a motor neuron Gal4 driver. With this system, it has been shown that hypomorphic Khc mutations cause anterograde and retrograde flux reductions for GFP-nSyb (60%–70%) and for mito-GFP (75% and 90%), supporting the hypothesis that normal dynein function in some processes depends on kinesin-1 (Pilling, 2005). Both anterograde and retrograde GFP-nSyb flux were reduced ~35% in ek4 mutant axons, supporting the idea that wild-type ek4 products facilitate some kinesin-1 functions. Surprisingly, ek4 mutant axons showed no change in anterograde mito-GFP flux and a 60% reduction in retrograde flux. Currently, the only mutations known to cause a similar unidirectional inhibition of retrograde mitochondrial flux are in Dhc64C (~80%) (Pilling, 2005), which encodes the motor subunit of cytoplasmic dynein (Horiuchi, 2005).
To further test the possibility that ek4 influences dynein, additional genetic interaction tests were done. Consistent with the original genetic screen for dominant enhancers of Khc, ek4 acts as a dominant enhancer of Kinesin light chain (Klc), causing synthetic tail flipping and axonal swelling phenotypes. No such interaction was seen when ek4 was combined with a mutant allele of Klp64D, which encodes an anterograde axonal motor of the kinesin-2 family. However, when ek4 was combined with a mutant allele of Dhc64C, synthetic tail flipping and axonal swelling phenotypes were seen. In summary, these results support the hypothesis that wild-type ek4 gene products facilitate vesicle transport by kinesin-1 and mitochondrial transport by cytoplasmic dynein (Horiuchi, 2005).
To identify the ek4 locus, meiotic recombination and deletion mapping approaches were initially used. The results indicated a position near the tip of the left arm of chromosome 3 within the 61F3-4 cytological region. That interval included APP-like interacting protein 1 (Aplip1), a gene that encodes a neuronally expressed Drosophila homolog of c-Jun N-terminal kinase (JNK)-interacting protein 1 (JIP-1), a scaffolding protein that has been shown to bind Kinesin light chain (KLC), a reelin receptor (ApoER2), and Alzheimer's amyloid precursor protein (APP), as well as JNK pathway kinases (Taru, 2000; Verhey; 2001; Yasuda, 1999). It has been proposed that JIP-1 and its close relative JIP-2 link kinesin-1 with axon vesicles to facilitate anterograde vesicle transport. Similar kinesin-1 linker functions have been proposed for an unrelated JNK scaffolding protein, sunday driver (syd, JSAP, JIP-3), and for APP, although the APP-kinesin relationship may be mediated by APLIP1/JIP-1. A P element transgene that included Aplip1 and flanking sequences fully rescued the tail flipping and partially rescued the axonal swelling phenotypes of larvae that were doubly heterozygous for Khc27 and ek4. Finally, sequencing of the Aplip1 locus from ek4 mutant animals revealed a single base change that converts a conserved proline at position 483 to leucine. This proline is within a conserved 11 amino acid C-terminal region (KBD) that has been shown to be important for binding of mammalian JIP-1 to KLC (Verhey, 2001). The transgenic rescue and sequencing results confirm that ek4 is a mutant allele of the Aplip1 gene, and hence it will be referred to as Aplip1ek4 (Horiuchi, 2005).
To determine whether the P483L mutation affects KLC-APLIP1 binding, epitope-tagged versions of KLC and APLIP1 were used for immunoprecipitation studies. After coexpression of Myc-KLC and wild-type Flag-APLIP1 in S2 cultured cells, anti-Myc antibody precipitated both proteins. Removal of the 11 amino acid KBD from Flag-APLIP1 eliminated detectable binding to Myc-KLC. Furthermore, changing proline 483 to either leucine or alanine substantially reduced KLC binding. This shows that P483 is indeed important for KLC binding, which suggests that at least some of the Aplip1ek4 mutant phenotypes are due to poor association of APLIP1 and kinesin-1 (Horiuchi, 2005).
If APLIP1 links kinesin-1 to anterograde transport vesicles in Drosophila axons, as has been proposed for JIP-1 in vertebrates (Verhey, 2001), APLIP1 should localize in axons and such localization should depend on its ability to bind KLC. To test those predictions, flies were transformed with P elements that carried either full-length UAS-Flag-Aplip1 or UAS-Flag-Aplip1ΔKBD. When driven by D42-Gal4, the two constructs produced equivalent levels of mRNA, which were many times in excess relative to the endogenous gene in larvae. Western blots of larvae with anti-Flag were not successful, but both the full-length and the ΔKBD Flag-tagged proteins were seen at equivalent levels in Westerns of transfected S2 cells, suggesting that both were stable. Interestingly, D42-Gal4-driven expression of one copy of full-length UAS-Flag-Aplip1 in motoneurons causes dramatic tail flipping and nearly 100% lethality during late larval and pupal stages. In larval nerves, it causes axon swellings that stain intensely for vesicles (anti-Syt) and APLIP1 (anti-Flag) . D42-Gal4-driven expression of the deletion construct caused no tail flipping or lethality. It did cause some axon swellings in larval nerves, and Flag-APLIP1ΔKBD staining was visible in those swellings. However, the overall amount of staining in nerves was substantially reduced relative to the amount seen after expression of the full-length protein (Horiuchi, 2005).
The presence of residual Flag-APLIP1ΔKBD in larval nerves indicates that some is transported into axons despite the fact that its binding to kinesin-1 is compromised. JIP-1 as well as APLIP1 is known to form multimers (Taru, 2002; Yasuda, 1999). Indeed, immunoprecipitation tests indicate that tagged APLIP1 and APLIP1ΔKBD can form stable multimers with one another. Thus, it is possible that in larval neurons, endogenous wild-type APLIP1 mediates linkage of some transgenic Flag-APLIP1ΔKBD to kinesin-1. Overall, these results suggest that binding between APLIP1 and KLC is an important factor in the presence of APLIP1 in axons, providing in vivo support for the hypothesis that APLIP1 is transported anterograde by kinesin-1 (Horiuchi, 2005).
To test the possibility that APLIP1 is associated with dynein-driven retrograde transport as well as with kinesin-1-driven anterograde transport, transgenic flies were developed carrying a UAS-GFP-Aplip1 transgene that expressed a stable fusion protein. When combined with the D42-Gal4 driver, some transformant lines showed paralysis and GFP-filled swellings, similar to the Flag-APLIP1 lines. Time-lapse imaging did not reveal obvious transport, suggesting that the GFP-APLIP1 was transported in a form too dispersed for imaging of discrete punctate signals. Turning to a classic axonal transport approach, a method was developed for nerve ligation in Drosophila larvae. A homozygous UAS-GFP-Aplip1 D42-Gal4 transformant line was used in which there were few axonal swellings and little visible axonal GFP fluorescence, presumably because of low expression. Intact live larvae were constricted with a fine synthetic fiber midway between head and tail to compress their segmental nerves. After 4 hr, they were partially dissected, the ligation threads were cut, dissection was completed, and the nerves were imaged. Distinct compressed regions were flanked by bright accumulations of GFP-APLIP1 on both the proximal and distal sides. This provides a strong indication that APLIP1 is carried not only by anterograde, but also by retrograde axonal transport (Horiuchi, 2005).
By using an in vivo genetic approach to identify proteins that contribute to the mechanism of kinesin-1-driven anterograde axonal transport, this study has identified APLIP1, a Drosophila homolog of the JNK-interacting protein JIP-1. In vivo axonal transport analysis with intact nervous systems suggests roles for APLIP1 in anterograde and retrograde transport of nSyb-tagged vesicles and in retrograde transport of mitochondria. Similar neuronal phenotypes were seen with either Aplip1 inhibition or overexpression, suggesting that correct stoichiometry of APLIP1 and its interacting proteins is critical for normal organelle transport. The influence of APLIP1 on nSyb vesicle transport in both directions could be explained simply by its importance for kinesin-1 function. Khc is required for normal retrograde dynein activity as well as for anterograde kinesin-1 activity, probably because of a physical or regulatory relationship between the two motors. Alternatively, APLIP1 might make separate contributions to kinesin-1-driven anterograde and dynein-driven retrograde vesicle transport (Horiuchi, 2005).
The selective influence of APLIP1 on retrograde, but not anterograde, transport of mitochondria, as well as Aplip1-Dhc64C genetic interactions, suggests that APLIP1 does have distinct, kinesin-independent functions in dynein-driven transport, at least for mitochondria. Considering how APLIP1 and other JIP-1-related proteins contribute to axonal transport mechanisms, binding studies suggest they may be structural components of kinesin-1-cargo linkage complexes. However, the APLIP1 influence on retrograde mitochondria, the well-known scaffolding role of APLIP1/JIP-1 in the JNK signaling pathway, and indications that JNK may influence motor linkage must also be kept in mind. Mitochondrial transport and distribution in axons responds dramatically to extracellular signaling and may also respond to intracellular signaling stimulated by changes in mitochondrial membrane potential. APLIP1 might be important in these or in other pathways that regulate dynein-cargo linkage and/or mechanochemistry. Future tests for a physical APLIP1-dynein association and for influences of JNK signaling on axonal transport may provide important insights into the microtubule-based transport mechanisms required to sustain neurons and other large asymmetric cells (Horiuchi, 2005).
Long-distance organelle transport toward axon terminals, critical for neuron development and function, is driven along microtubules by kinesins. The biophysics of force production by various kinesins is known in detail. However, the mechanisms of in vivo transport processes are poorly understood because little is known about how motor-cargo linkages are controlled. A c-Jun N-terminal kinase (JNK)-interacting protein (JIP1) has been identified previously as a linker between kinesin-1 and certain vesicle membrane proteins, such as Alzheimer's APP protein and a reelin receptor ApoER2. JIPs are also known to be scaffolding proteins for JNK pathway kinases. Evidence is presented that a Drosophila ubiquitin-specific hydrolase (Fat facets) and a JNK signaling pathway that it modulates can regulate a JIP1-kinesin linkage. The JNK pathway includes a MAPKKK (Wallenda/DLK), a MAPKK (Hemipterous/MKK7), and the Drosophila JNK homolog Basket. Genetic tests indicate that those kinases are required for normal axonal transport. Biochemical tests show that activation of Wallenda (DLK) and Hemipterous (MKK7) disrupts binding between kinesin-1 and APLIP1, which is the Drosophila JIP1 homolog. This suggests a control mechanism in which an activated JNK pathway influences axonal transport by functioning as a kinesin-cargo dissociation factor (Horiuchi, 2007).
Maintaining proper distributions of protein complexes, RNAs, vesicles, and other organelles in axons is critical for the development, function, and survival of neurons. The primary distribution mechanism relies on long-distance transport driven by microtubule motor proteins. Components newly synthesized in the cell body, but needed in the axon, bind kinesin motors that carry them toward microtubule plus ends and the axon terminal (anterograde transport). Neurotrophic signals and endosomes, examples of axonal components that require transport to the cell body, bind dynein motors that carry them toward minus ends (retrograde transport). The importance of these processes is highlighted by the observation that mutation of motors and other transport machinery components can cause neurodegenerative diseases in humans and analogous phenotypes in model organisms (Horiuchi, 2007).
Two key questions are (1) how do cargoes link to particular motors, and (2) how are such linkages regulated to ensure appropriate pickup and dropoff dynamics? For kinesin-vesicle linkages, scaffolding proteins have emerged as key connectors. For example, the cargo-binding kinesin light chain (Klc) subunit of kinesin-1 binds not only the kinesin-1 heavy chain (Khc) but also JNK-interacting proteins (JIPs). Vertebrate JIPs can bind multiple components of the JNK signaling pathway, e.g., JNK itself, upstream activating kinases (MAPKKs), and regulatory kinases (MAPKKKs). JIPs can also bind vesicle-associated membrane proteins, such as ApoER2, which is a reelin receptor, and APP, a key factor in Alzheimer's disease. Therefore, JIP scaffolding proteins are likely to link JNK pathway kinases and kinesin-1 to vesicles carrying these membrane proteins. This raises an interesting question: Are the JNK pathway kinases simply passive hitchhikers on the kinesin-1/JIP/vesicle complex, or can they actively regulate its transport (Horiuchi, 2007)?
A genetic screen was conducted for factors that control kinesin-JIP linkage during axonal transport. The screen was based on the previous observation that neuron-specific overexpression of Aplip1, which encodes the Drosophila JIP1, causes synaptic protein accumulation in axons, larval paralysis, and larval-pupal lethality, the classic axonal-transport-disruption phenotypes caused by Khc and Klc mutations. Why might overexpression of the JIP1 cargo linker for kinesin-1 disrupt axonal transport? The disruptive effect requires APLIP1 (JIP1)-Klc binding. It may be that excess APLIP1 (JIP1) competes with other Klc-binding proteins, for example, different linkers that may attach kinesin-1 to other cargoes. In search of factors that can disrupt or antagonize APLIP1 (JIP1)-Klc binding, a screen was performed for genes that can suppress the axonal-transport phenotypes when co-overexpressed with Aplip1. An 'EP' collection of fly strains capable of the targeted overexpression of endogenous Drosophila genes was screened and P{EP}fafEP381, a line that overexpresses fat facets (faf), was identified as a strong suppressor of the APLIP1 (JIP1)-induced lethality and other neuronal overexpression phenotypes (Horiuchi, 2007).
Faf protein antagonizes ubiquitination and proteasome-mediated degradation of its target proteins. Interestingly, Faf was recently reported to stimulate a Drosophila neuronal JNK signaling pathway that is regulated by the MAPKKK Wallenda (Wnd), a homolog of dual leucine zipper-bearing kinase (DLK) that is known to bind JIP1. Overexpression of faf leads to increased levels of Wnd (MAPKKK) protein and thereby causes excessive synaptic sprouting through a pathway that requires the Drosophila JNK homolog Basket. It was found that mutating just one copy of wnd blocked the suppression of Aplip1 overexpression by P{EP}fafEP381. This suggests that faf overexpression suppresses APLIP1 (JIP1)-Klc interaction by elevating the level of Wnd (MAPKKK). Consistent with this, direct overexpression of wnd in neurons with a wild-type transgene (UAS-wnd) was as effective as P{EP}fafEP381 in suppressing UAS-Aplip1-induced axonal accumulation of synaptic proteins. Equivalent expression of a 'kinase-dead' mutant transgene (UAS-wndKD) did not suppress the defects. Thus, Wnd (MAPKKK) and its downstream phosphorylation targets may actively regulate APLIP1 (JIP1)-Klc binding in neurons (Horiuchi, 2007).
If Wnd (MAPKKK) signaling plays a role in normal axonal transport, disrupting its function should cause axonal-transport phenotypes. Consistent with this, wnd loss-of-function mutations (wnd1/wnd2) in an otherwise wild-type background caused accumulation of synaptic proteins in axons. The accumulation phenotype was rescued by motoneuron expression of the wild-type wnd transgene but not by equivalent expression of the kinase-dead mutant transgene. The likely target of Wnd (MAPKKK) kinase activity is the Drosophila homolog of MKK7, Hemipterous (Hep), a MAPKK that activates Bsk (JNK). Mutation of hep also causes axonal accumulations, as does neuronal expression of a dominant-negative mutant bsk transgene. The results of these genetic-inhibition tests combined with those of the Aplip1-overexpression-suppression tests suggest that a Wnd (MAPKKK)-activated JNK pathway influences fast axonal transport by regulating APLIP1 (JIP1)-Klc binding (Horiuchi, 2007).
Is a Wnd (MAPKKK)-Hep (MAPKK)-Bsk (JNK) signaling module bound by APLIP1 (JIP1)? Although all three components of the homologous vertebrate module (DLK-MKK7-JNK) bind JIP1, APLIP1 (JIP1) lacks a conserved JNK-binding domain, and it does not bind directly to Bsk (JNK). However, APLIP1 (JIP1) does bind Hep (MAPKK), Klc, and the Drosophila APP homolog APPL. To determine whether Wnd (MAPKKK) associates with Hep (MAPKK) and APLIP1 (JIP1), coexpression and immunoprecipitation tests were performed in Drosophila S2 cultured cells. Wnd (MAPKKK) did not coprecipitate with APLIP1 (JIP1). However, Hep (MAPKK) did coprecipitate with APLIP1 (JIP1), and Wnd (MAPKKK) coprecipitated with Hep (MAPKK). Thus, Wnd (MAPKKK) may bind and influence the APLIP1 (JIP1)-kinesin complex via Hep (MAPKK) (Horiuchi, 2007).
Can Wnd (MAPKKK) and Hep (MAPKK) control the binding of APLIP1 (JIP1) to Klc? When expressed in S2 cells, APLIP1 (JIP1) and Klc exhibit strong binding, as assessed by coimmunoprecipitation. Coexpression of wild-type Wnd (MAPKKK) partially inhibited APLIP1 (JIP1) binding to Klc, but coexpression of a kinase-dead mutant Wnd (MAPKKK) did not. Wild-type Hep (MAPKK) also caused a partial inhibition of APLIP1 (JIP1)-Klc binding, and a constitutively active mutant Hep (MAPKK) caused nearly complete inhibition. Finally, coexpression of wild-type Wnd (MAPKKK) and Hep (MAPKK) together caused an almost complete inhibition of APLIP1 (JIP1)-Klc binding. In addition to inhibiting APLIP1 (JIP1)-Klc binding, Wnd-Hep activation in S2 cells increased the level of Bsk (JNK) activation. Hence, there is a correlation between decreased levels of APLIP1 (JIP1)-Klc binding and elevated levels of Bsk (JNK) activation. This suggests that, despite the lack of a known JNK-binding site on APLIP1 (JIP1), Bsk (JNK) may be the kinase that disrupts the APLIP1 (JIP1)-Klc complex. These results suggest that Wnd (MAPKKK) activation of Hep (MAPKK), and perhaps also Hep (MAPKK) activation of Bsk (JNK), can regulate the linkage between kinesin-1 and a cargo complex via the JIP1-like scaffolding protein, APLIP1 (Horiuchi, 2007).
Hep (MAPKK) may regulate the APLIP1 (JIP1) complex either by activating JNK or by a mechanism independent of JNK. Observations that motoneuron-specific inhibition of Bsk (JNK) caused transport defects similar to those caused by mutations in wnd and hep and that decreased APLIP1 (JIP1)-Klc binding in S2-cell lysates coincided with increased phosphorylated Bsk (JNK) support pathway 1, i.e., Hep (MAPKK) activation of Bsk (JNK), which then directly or indirectly inhibits APLIP1 (JIP1)-Klc binding. A second pathway, Pathway 2, employs an alternative mechanism in which activated Hep (MAPKK) does not need Bsk (JNK) to inhibit APLIP1 (JIP1)-Klc binding. There is little current evidence that Hep or its vertebrate MAPKK homolog MKK7 have phosphorylation targets other than Bsk (JNK). However, that does not exclude the possibility that activated Hep induces in APLIP1 (JIP1) a direct conformational change that causes Klc dissociation. Regardless of how Hep (MAPKK) disrupts binding, when kinesin-1 is not attached to cargo via JIP1, it can fold into a compact form that does not interact with microtubules. Hence, the activated Wnd (MAPKKK) pathway could both inhibit APLIP1 (JIP1)-Klc binding and cause dissociation of kinesin-1 from microtubules. Consistent with this, recent studies report that stimulation of JNK pathways in cultured cells or axoplasm can disrupt the association of kinesin-1 with microtubules (Horiuchi, 2007 and references therein).
From a broader perspective on axonal-transport regulation, it is interesting to consider that there are multiple types of kinesin-1 cargoes, that there are various JIPs that could be specific for different cargoes, and that different MAPKKKs can associate with different JIPs. By sitting at the top of a classic signaling cascade, MAPKKKs such as Wnd are in a good position to differentially control the transport of specific subsets of anterograde kinesin-1 cargoes in response to specific cellular signals. It is known in mammals that other MAPKKKs such as MLK, ASK1, and MEKK1 can bind JIP scaffolding proteins. It will be interesting to determine whether they too influence kinesin-cargo interactions (Horiuchi, 2007).
This work provides the first demonstration that a kinesin and its transport functions can be influenced by a MAPKKK. More specifically, the MAPKKK Wnd and its downstream MAPKK Hep can regulate attachment of a JIP1 cargo linker to kinesin-1. These results also provide the first indication that ubiquitination pathways, by way of MAPKKKs, could be important for proper regulation of axonal transport. Finally, our results suggest that JNK pathway kinases are not just hitchhikers on the axonal kinesin-1/JIP/cargo complex; rather, they can actively regulate its transport dynamics (Horiuchi, 2007).
Axonal transport is required for the elaboration and maintenance of synaptic morphology and function; this study demonstrates that Liprin-α is required for trafficking of synaptic vesicles. Liprin-αs are scaffolding proteins important for synapse structure and electrophysiology. A reported interaction with Kinesin-3 (Kif1a) suggested Liprin-α may also be involved in axonal transport. Aberrant accumulations have been discovered, at the light and ultrastructural levels, of synaptic vesicle markers (Synaptotagmin and Synaptobrevin-GFP) and clear-core vesicles along Drosophila Liprin-α mutant axons. Analysis of presynaptic markers reveals reduced levels at Liprin-α synapses. Direct visualization of Synaptobrevin-GFP transport in living animals demonstrates a decrease in anterograde processivity in Liprin-α mutants but also an increase in retrograde transport initiation. Pull-down assays reveal that Liprin-α interacts with Drosophila Kinesin-1 (Khc) but not dynein. Together, these findings suggest that Liprin-α promotes the delivery of synaptic material by a direct increase in kinesin processivity and an indirect suppression of dynein activation. This work is the first to use live observation in Drosophila mutants to demonstrate the role of a scaffolding protein in the regulation of bidirectional transport. It suggests the synaptic strength and morphology defects linked to Liprin-α may in part be due to a failure in the delivery of synaptic-vesicle precursors (Miller, 2005; full text of article).
Three prominent models have been proposed to explain the regulation of bidirectional transport: (1) a substitution model in which only one set of motors is present on the cargo at a given time, (2) a tug-of-war model in which both anterograde and retrograde motors are bound and always active but differ in their number on the cargo, and (3) a coordinate-regulation model in which both sets of motors are bound but one group is inactive. The observations that dynein is associated with anterograde transported cargos and kinesin is associated with retrogradely transported vesicles containing synaptic components argue against the substitution model of transport for SVPs. If the tug-of-war model were correct, then the shift in flux that was observed in Liprin-α mutants would correspond to a change in the number of bound active motors and a skewing of the velocity profile. However, because no such shift is observed, the current results are most consistent with a model in which coordinate regulation mediated through Liprin-α modulates transport. Because Liprin-α interacts with kinesins but not dynein, the data suggest that Liprin-α directly promotes kinesin activity or cargo-association, which then leads to dynein inhibition through some additional component(s) (Miller, 2005).
In light of the observations that disruption of kinesin alters the morphology and electrophysiological properties of synapses, these observations suggest that the synaptic defects seen in mutants of LAR and the Anaphase Promoting Complex may be mediated in part by Liprin-α’s role in axonal transport. As a scaffolding protein with multiple known partners and motors, Liprin-α is in an ideal position for integrating and transducing information to regulate the delivery of cargoes to and from the synapse (Miller, 2005).
To address questions about mechanisms of filament-based organelle transport, a system was developed to image and track mitochondria in an intact Drosophila nervous system. Mutant analyses suggest that the primary motors for mitochondrial movement in larval motor axons are kinesin-1 (anterograde) and cytoplasmic dynein (retrograde), and interestingly that kinesin-1 is critical for retrograde transport by dynein. During transport, there was little evidence that force production by the two opposing motors was competitive, suggesting a mechanism for alternate coordination. Tests of the possible coordination factor P150Glued suggested that it indeed influenced both motors on axonal mitochondria, but there was no evidence that its function was critical for the motor coordination mechanism. Observation of organelle-filled axonal swellings ('organelle jams' or 'clogs') caused by kinesin and dynein mutations showed that mitochondria could move vigorously within and pass through them, indicating that they were not the simple steric transport blockades suggested previously. It is speculated that axonal swellings may instead reflect sites of autophagocytosis of senescent mitochondria that are stranded in axons by retrograde transport failure; a protective process aimed at suppressing cell death signals and neurodegeneration (Pilling, 2006).
The presence of different types of motor proteins on one organelle raises interesting questions about how they are coordinated to accomplish proper cargo transport and distribution. Coordination of cytoplasmic dynein and an unidentified plus-end motor on lipid droplets in Drosophila embryos is sensitive to the dominant-negative Gl1allele, implicating the P150Glued component of dynactin in the coordination mechanism. The current results are consistent with cohabitation of dynein and kinesin-1 on axonal mitochondria and with their activities being coordinated to avoid antagonism, but the coordination mechanism is insensitive to either the Gl1 allele or neuron-targeted Gl RNAi. Both those inhibition approaches enhance, rather than reduced, plus- and minus-end mitochondrial runs in motor axons. When a more complete disruption of P150Glued in larval neurons becomes possible, perhaps a role in the coordination mechanism will become evident. However, the current data suggest that axonal P150Glued slows mitochondrial runs in both directions. It is an elongated protein with a microtubule-binding site at one end that is thought to facilitate dynein processivity by acting as a tether to hold the dynein and its attached cargo near the microtubule. Because mitochondria are sufficiently large to accommodate many motor complexes and they are confined near microtubules by the small diameter of axons, processivity in axons may not need assistance from P150Glued. In the confines of the axon, perhaps the microtubule-binding action of P150Glued actually generates drag that retards rather than facilitates mitochondrial runs (Pilling, 2006).
The results are consistent with a model in which directional programming is controlled by signals that partition mitochondria into distinct anterograde, stationary, and retrograde classes. It is suggested that mitochondria with fresh, cell body-derived components generate a local signal that represses cytoplasmic dynein, activates kinesin-1, and thus dictates anterograde transport toward the terminal. In the axon, areas requiring ATP generate local signals that convert anterograde mitochondria to stationary by activating static cross-links with the cytoskeleton. Over time, damage to mitochondrial components from reactive oxygen species causes a decline in energy-producing capacity that triggers release from cytoskeletal anchoring, activation of dynein, and repression of kinesin-1. After retrograde transport to the cell body, senescent mitochondria are enclosed in membranes by autophagocytosis and then are degraded by fusion with lysosomes. Regulatory mechanisms that control transitions between the three states may be triggered by extracellular signals, local ATP and ion concentrations, mitochondrial inner membrane potential, or other internal cues from mitochondria. Definition of those control mechanisms should be an exciting endeavor in the coming years (Pilling, 2006).
The genesis and nature of axonal swellings is of great interest because they are a prominent pathological feature of Alzheimer's and some motor neuron diseases. The discovery that kinesin-1 and dynein mutations in Drosophila cause axonal swellings led to the proposal that they are a consequence of traffic jams, i.e., halted organelles impede passage of others causing stochastic pile-ups that locally block transport and force axon swelling. However, it is clear from the direct observations presented in this study that organelles in swellings can be quite mobile and that swellings do not cause a general blockade of mitochondrial transport (Pilling, 2006).
If axonal swellings are not caused by localized transport blockades, what causes them? One answer is suggested by the prominence of large autophagosomes and lysosomal organelles within swellings. Mitochondria with damaged proteins and weak membrane potential undergo a membrane permeability transition that allows release of signals for apoptotic and necrotic cell death. That permeability transition also triggers rapid autophagocytosis and lysosomal degradation of mitochondria that presumably help suppress the cell death signaling. It is suggested that mitochondria stranded by failed retrograde transport pass the permeability transition in the axon, release cell death signals and stimulate local axonal autophagocytosis. Structural and physiological changes that accompany the autophagocytosis cause local enlargement of axon diameter. If this hypothesis is true, then the swellings associated with some neurodegenerative diseases may be triggered by failed retrograde transport and the lingering presence of spent mitochondria in axons (Pilling, 2006).
Support for this stranded mitochondria hypothesis comes from classic studies of physical blockades of axonal transport. Anterograde organelles accumulate on the proximal (cell body) side of a blockade, and retrograde organelles accumulate on the distal side. Proximal organelles can return to the cell body by switching to retrograde transport, but distal organelles are trapped in the axon. Among the normal-looking organelles on the distal side, many autophagosomes and lysosomal organelles accumulate. The source of those organelles has been puzzling, because they are not common in normal axons. As proposed for autophagocytic/lysosomal organelles in the axonal swellings of Drosophila kinesin-1 and dynein mutants, those distal blockade-induced organelles could reflect local autophagy induced by trapped mitochondria that have passed the membrane permeability transition. Further study of the relationship between mitochondrial transport behavior, physiology, and autophagy may provide new insight into the genesis of axonal swellings and contribute to understanding of the pathology of neurodegenerative diseases (Pilling, 2006).
Mitochondria are distributed within cells to match local energy demands. The microtubule-dependent transport of mitochondria depends on the ability of Milton (Milt) to act as an adaptor protein that can recruit the heavy chain of conventional kinesin-1 (kinesin heavy chain [KHC]) to mitochondria. Biochemical and genetic evidence demonstrate that kinesin recruitment and mitochondrial transport are independent of kinesin light chain (KLC); KLC antagonizes Milton's association with KHC and is absent from Milton-KHC complexes, and mitochondria are present in klc-/- photoreceptor axons. The recruitment of KHC to mitochondria is, in part, determined by the NH2 terminus-splicing variant of Milton. A direct interaction occurs between Milton and Miro, a mitochondrial Rho-like GTPase, and this interaction can influence the recruitment of Milton to mitochondria. Thus, Milton and Miro are likely to form an essential protein complex that links KHC to mitochondria for light chain-independent, anterograde transport of mitochondria (Glater, 2006).
Milton has been shown to be required for mitochondrial transport within Drosophila photoreceptors (Stowers, 2002). Mitochondria are absent from Milt photoreceptor axons, but are normally distributed and appeared to be functional in their cell bodies. Although devoid of mitochondria, their axons and synapses are otherwise surprisingly normal in their general architecture, possessing microtubules, synaptic vesicles, and active zone specializations. Thus, the transport defect is selective for mitochondria (Stowers, 2002; Gorska-Andrzejak, 2003). The mechanism of Milton's action has remained unknown, but Milton is associated with mitochondria and coimmunoprecipitates with kinesin heavy chain (KHC) in extracts of fly heads (Stowers, 2002). The mammalian homologues milton 1 and 2, which are also called O-linked N-acetylglucosamine–interacting protein 106 (OIP106) and gamma-aminobutyric acid A receptor–interacting factor-1 (GRIF-1), also colocalize with mitochondria and coimmunoprecipitate with KIF5B, which is a mammalian homologue of Drosophila KHC. Therefore, it has been suggested that Milton acts as an adaptor or regulator of the mitochondrial anterograde motor (Glater, 2006 and references therein)
This study examined the involvement of Milton in kinesin-mediated mitochondrial motility and, thus, in the essential process of distributing mitochondria within the cell. From these studies a model was derived of a protein complex that includes kinesin and adaptor proteins that link kinesin to the mitochondrion. These proteins are also likely to serve as a focal point for regulating mitochondrial motility (Glater, 2006).
In vivo, milton is required for the axonal transport of mitochondria throughout the nervous system (Stowers, 2002). Milton associates with kinesin-1 via a highly conserved domain located between residues 138 and 450. This association can recruit kinesin to mitochondria in COS7 cells and appears to activate plus end–directed transport of mitochondria, as judged by their redistribution to aggregates in the periphery of many cells transfected with both milton and KHC. These findings provide a mechanistic explanation for the absence of mitochondria from milton axons and terminals. Consistent with this model, the motors that endogenously associate with Milton, KHC in Drosophila, and KIF5 members in mammals have previously been implicated in the axonal transport of mitochondria (Glater, 2006).
The association of Milton with mitochondria appears to be mediated, in part, by its interactions with Miro, and this probably accounts for the failure of mitochondrial transport in the axons of miro mutants (Guo, 2005). This proposal is supported (a) by the ability of a truncated cytosolic form of miro to act as a dominant negative and displace milton from mitochondria, and (b) by the ability of overexpressed full-length miro to recruit to mitochondria a truncated milton (residues 1–750) that could not independently localize there. However, additional interactions for tethering milton to mitochondria are likely, as a COOH-terminal portion of milton (residues 847–1,116) also localizes to the organelle. The difficulty of purifying mitochondria from limited numbers of homozygous miro larvae prevents a direct determination of the amount of milton on mitochondria that lack miro (Glater, 2006).
The role of miro in kinesin-mediated transport does not preclude additional roles for miro. Indeed, such functions are likely because a Miro homologue, GEM1p, is found in yeast, where mitochondrial motility is chiefly actin-based, and GEM1 mutants have abnormal mitochondrial distributions. In addition, it will be of interest to determine the relationship of Milton and Miro to Syntabulin (Cai, 2005), which is another protein that has recently been proposed to link kinesin to mitochondria (Glater, 2006).
Unexpectedly, it was found that axonal transport of mitochondria did not require the light chains of the kinesin-1 motors and that light chains were, indeed, absent from the milton–kinesin complex. When expressed in COS7 and HEK293T cells, the association between Milton and KHC was inhibited by KLC. In fly homogenates, KLC was not detected in immunoprecipitates of the Milton–KHC complex. Mitochondria were abundant in the axons of klc–/– photoreceptors. Thus, this mitochondrial motor provides an exception to the conventional tetrameric structure of kinesin-1. Precedent for KHC-based transport that is KLC independent has been reported in Neurospora crassa, sea urchins, neuronal dendrites, and the transport of RNA particles (Glater, 2006).
The interaction of Milton with KHC was not only KLC independent, but was inhibited by KLC overexpression in transfected COS7 and HEK293T cells. Therefore, a pool of KHC without KLC is required for Milton to associate with KHC in vivo. Previous studies have found evidence for such a pool in bovine brain. In light of the current findings, it may be appropriate to consider the light chains as one of several cargo adaptors for kinesin-1, of which Milton is another (Glater, 2006).
Mitochondria are not static. In dividing cells they go through orchestrated movements to distribute themselves between the daughter cells. Within axons they typically alternate between stationary and moving states and can reverse their direction. They arrest in the presence of elevated Ca2+, including Ca2+ that is derived from the activation of synaptic receptors, and respond to the activation of neurotrophin receptors and various intracellular signals. It is noteworthy that, in addition to linking kinesin to the mitochondria, the Milton–Miro complex provides several possible mechanisms for the regulation of transport. These include the alternative splicing of Milton, the posttranslational modification of Milton, and the modulation of the state of miro (Glater, 2006).
The choice of NH2 terminus splicing variant can influence KHC's association with the adjacent region of Milton. In particular, KHC did not associate with Milton-C, although it contains the KHC-association domain that is common to all the isoforms. The NH2 terminus of Milton-C presumably inhibits the interaction with KHC and might, thereby, reserve a pool of mitochondria for retention in the cell body. Alternatively, the inhibition may not be constitutive in vivo, but, instead, might undergo regulation by additional factors and thereby control the recruitment of kinesin. In this context, it may be noteworthy that multiple bands of Milton are detected on immunoblots from fly heads. Most of the milton isoforms in these homogenates are in an association with KHC, as determined by immunodepletion with anti-KHC. However, there is one major band, representing nearly half of the endogenous Milton, which does not appear to be associated with KHC (Stowers, 2002). Thus, additional motors may associate with Milton, and particularly with Milton-C. Milton may also be involved in such processes as mitochondrial fission and elongation, and such a role might explain the clustering of mitochondria when Milton and Miro are overexpressed (Glater, 2006).
The alternative splicing of Milton may also represent an adaptation of the complex to the needs of particular cell types. Antiserum P1–152, which binds only to Milton-A, labels a subset of the structures in the Drosophila brain that are recognized by antibodies to the common regions (Stowers, 2002). Thus, there is tissue specificity in the choice of splicing variant. To date, ESTs for Milton-D have only been found in a testes library; therefore, Milton-D may correspond to the male-specific milton transcripts on Northern blots (Stowers, 2002) and be necessary for the elongation of mitochondria along the axoneme of sperm (Glater, 2006).
Posttranslational modifications are also likely to regulate mitochondrial motility. In particular, the COOH-terminal portions of the mammalian miltons bind to, and are substrates for, the cytosolic glycosylating enzyme O-GlcNAc transferase (OGT). Drosophila OGT has been identified by mass spectroscopy in immunoprecipitates of Milton from fly homogenates. In addition, GlcNAc-modified Milton is associated with kinesin in vivo in Drosophila, although the physiological consequences of this conserved modification are not known (Glater, 2006).
Mitochondrial motility is a feature of most, perhaps all, eukaryotic cells. In neurons, much of this motility is microtubule based, with kinesin as the plus end–directed motor. This motility, and its regulation by a variety of signals, permits the mitochondria to be distributed in accordance with local energy use. Inadequate mitochondrial function in axons or dendrites can result in decreased synapse formation, a failure to maintain synaptic transmission, or axonal degeneration. The identification of milton and miro as key components of the mechanism for mitochondrial transport by KHC should lead to a greater mechanistic understanding of the regulation of mitochondrial movement (Glater, 2006).
Home page: The Interactive Fly © 1997 Thomas B. Brody, Ph.D.
The Interactive Fly resides on the
Society for Developmental Biology's Web server.