P125, a COPII INTERACTING PROTEIN 4

108 CHAPTER Discussion This study demonstrated that p125A interacts with Sec31A and that this interaction is direct and independent of p125A’s interaction with Sec23A. p125A was also found to be associated with the Sec13/31 heterotetramer in the cytosol. Fluorescence microscopy and live-cell imaging studies showed that p125A is distributed to the ERES (ER exit sites). In Sec31A knockdown cells, the majority of p125A was redistributed to the cytosol, implying that p125A is targeted to the membrane via its interaction with Sec31A. Depletion of p125A by RNAi affects the ERES distribution of Sec31A, disrupts the Golgi morphology, and delays the export of a Golgi enzyme (GT-GFP) and VSVG out of the ER. p125A depletion also reduces the distribution of Sec31A and Sec13 to the membrane, suggesting that it might have a role in stabilizing the association of Sec13-Sec31 complex with the membrane. Previous study showed that the C-terminal 180 residue region of Sec31A expressed as GST fusion protein (GST-Sec31A), inhibited in vitro transport of VSVG protein from the ER to the Golgi (Tang et al., 2000), suggesting that it may compete with intact endogenous Sec31A for some limiting factor(s) critical for ER-Golgi transport. The results of the GST pull-down (Figure 3.1.3) and co-immunoprecipitation experiments (Figure 3.2.1) showed that p125A binds to the C-terminal 180 residues of Sec31A. Therefore, p125A may be one of the factors that were depleted by GST-Sec31A. The ability of p125A to interact with two COPII components also suggests that it is in close association with Sec23 and Sec31 on the membrane, and may play a role in the early secretory pathway at the ER exit sites. 109 Analysis of a series of different p125A fragments narrowed down the Sec31A binding domain to between residues 260-600 (Figure 3.2.2). Work by Mizoguchi and colleagues (Mizoguchi et al., 2002) had determined that the proline-rich region (residues 135–259) of p125A is responsible for the binding to Sec23A. The region in p125A for binding Sec31A does not overlap with the region that binds Sec23A, implying that binding of p125A to Sec31A is direct and may be independent from Sec23A (Figure 3.3). Sec13 and Sec31 exist as a heterotetramer in the cytosol (Salama et al., 1993; Salama et al., 1997; Shugrue et al., 1999; Tang et al., 2000; Lederkremer, et al, 2001; Kim et al., 2001). Figure 4.1 shows the schematic of the Sec13/Sec31 heterotetramer. They have a predicted combined molecular weight of about 370 kDa (Lederkremer, et al., 2001). However, this molecular weight cannot account for the detection of the heterotetramer complex at gel filtration fractions that corresponds a size of the 600700 kDa size (Tang et al., 2000). These results suggest that there are possibly other proteins within the complex. Gel filtration studies of HeLa cytosol showed that majority of p125A co-eluted in the same fractions as the Sec13/31 complex (Figure 3.4.1), suggesting that p125A may be associated with the Sec13/31 complex in the cytosol. In this study, both HA and myc tagged p125A was able to co-immunoprecipitate each other and also the Sec13/31 complex (Figure 3.4.3). This suggests that more than one molecule of p125A could be associated with the Sec13/31 heterotetramer. The coimmunoprecipitation of myc-p125AFL by HA-p125AFL may occur via direct interactions with each other, or through indirect interactions with the Sec13/31 110 heterotetramer complex. The Sec13/31 heterotetramer complex comprises of two molecules of Sec13 and two molecules of Sec31. The presence of two monomers of Sec31 suggests the existence of two putative p125A binding sites, and the p125ASec13/31 complex may therefore contain two copies of p125A. Immunoprecipitation of one molecule of p125A could co-precipitate the p125A-Sec13/31 complex that contains molecules of Sec13, Sec31A and another copy of p125A.   Figure 4.1. A schematic diagram of the Sec13/Sec31 heterotetramer. The Sec13/Sec31 complex is arranged in Sec13/Sec31-Sec31/Sec13. Adapted from Stagg et al., 2006; Gurkan et al., 2006. Combined molecular weights of the Sec13/31 complex (370 kDa) with one copy of p125A (125 kDa) is only 505 kDa. However, if the molecular weight of two molecules of p125A was included (250 kDa), it will bring the total molecular weight of the complex to about 620 kDa, which falls within the range corresponding to the eluted size. The ability of p125A to form interactions with itself and Sec31A in the cytosol and the co-fractionation of p125A with Sec13/31 complex strongly supports the plausibility that p125A is in a ternary complex with the Sec13/31 heterotetramer in the cytosol (Figure 4.2), prior to its recruitment onto the membrane. Our results clearly suggest that p125A is a major protein in the cytosolic complex; however, the possibility of the presence of other cytosolic proteins that may be associated with the p125A-Sec13/31 ternary complex could not be excluded. The crystal structure of p125A has yet to be solved. Obtaining structural information of p125A in association 111 with the Sec13/31 complex and Sec23A would therefore be crucial in the understanding of the function of p125A and its role at the ERES. Figure 4.2. Schematic illustration of p125A-Sec13/Sec31 ternary complex in the cytosol Structural analysis of the Sar1-Sec23/Sec24 complex (Lederkremer et al., 2001, Bi et al., 2002) showed that Sec23/Sec24 exists as a bowtie-shaped complex, with a concave surface for binding membrane. The interface between Sec23 and Sar1 is stabilized by the bound GTP (Bi et al., 2002). Sec13/Sec31 exists as a heterotetramer which is arranged in the order of Sec13/Sec31-Sec31/Sec13. This Sec13/Sec31 architectural core is organized as a linear rod of 28 nm (Lederkremer et al., 2001; Matsuoka et al., 2001; Stagg et al., 2006; Fath et al., 2007). The Sar1-Sec23/Sec24 complex makes up the inner layer of the COPII coat while the Sec13/Sec31 heterotetramer makes up the outer layer of the COPII coat (Bi et al., 2007). Twentyfour copies of the Sec13/Sec31 linear rod can self-assemble to form the minimal COPII cage (Fath et al., 2007; Stagg et al., 2006). The molecular structure of the complete COPII cage containing both Sec23/Sec24 and Sec13/Sec31 complexes was recently solved using purified COPII proteins (Stagg et al., 2008). assembled COPII cage appears to have three layers. This self- The Sec13/Sec31 complex makes up the outer layer, while the middle layer comprises the Sec23/Sec24 complex. The innermost layer lacks regular structure and was suggested to be the unassembled 112 Sec13/Sec31 and/or the Sec23/Sec24 complexes (Stagg et al., 2008). The Sec13/Sec31 complex does not appear to contact the membrane directly but via interaction with the inner coat (Matsuoka et al., 2001; Lee and Miller, 2007). Sec31A interacts with Sec23A via its proline-rich region (Shaywitz et al., 1997, Shugrue et al., 1999). This proline-rich region contains the active fragment of Sec31A (residues 899 -847) which binds and stimulates the GAP activity of Sec23. This short 49 amino acid fragment of Sec31A (residues 899 to 947) is involved in the interaction with Sec23A and Sar1-GTP, and may be an anchorage point where the outer shell of the COPII coat can link to the inner layer (Bi et al., 2007). As p125A can bind to both Sec23A (Tani et al., 1999; Nakajima et al., 2002; Shimoi et al., 2005) and Sec31A, and have the ability to bind to phosphatidylinositol phosphate (Iinuma et al., 2007), it is likely to span across the inner and outer layer of the COPII cage, and may function to stabilise the interaction between Sec23A and Sec31A, and perhaps between these COPII components and the membrane. Immunofluoresence studies using p125A specific antibodies showed that p125A colocalizes very well with Sec31A (Figure 3.6.1), a marker for the ERES. At the ultrastructural level, immunoelectron microscopy showed that p125A was localized to regions between the ER and Golgi (but not within the Golgi stacks), including the cisGolgi area, similar to the immunolabeling pattern observed for Sec31A (Shimoi et al., 2005). All these indicate that p125A is indeed located at the ERES. Live cell imaging have revealed that ERES are very stable and long lived structures (Hammond & Glick, 2000; Stephens et al., 2000) that direct the recycling of COPII components during successive rounds of cargo selection and vesicle budding. Furthermore, ERES (marked by YFP-Sec23A) were also shown to undergo fusion 113 (Stephens et al., 2003). Consistent with these observations, p125A was observed to be stably colocalized to Sec31A at the ERES (Figure 3.7.2). ERES containing both Sec31A and p125A was also observed to undergo homotypic fusion (Figure 3.7.3). The intensity of the fused ERES seems brighter as compared to the intensities of the individual ERES prior to fusion. Fusion of the ERES site might be a mechanism to enlarge the ERES to allow the incorporation of larger cargo molecules such as collagen. The fact that p125A is in close association with Sec31A at the ERES supports the observation that both proteins are indeed interacting with each other in vivo. Figure 4.3. A schematic illustration of p125A’s interactions with Sec13/Sec31complex and the Sec23/Sec24 complex at the ERES membrane. When Sec31A was depleted from HeLa cells, the punctate distribution of p125A was noticeably reduced. The residual punctate staining may result from its association with Sec23A, suggesting that the ERES distribution of p125A is at least partially determined by its interaction with Sec31A. On the other hand, when p125A was depleted, the punctate staining of Sec31A appears to be reduced. The distribution of 114 Sec31A became less concentrated at the perinuclear region, and more diffuse around the cell periphery (Figure 3.23). Golgi morphology was also altered when p125A was depleted from HeLa cells. The same was observed for Sec13 depleted but not Sec31A depleted cells. Depletion of Sec31A did not cause any major alteration to the Golgi morphology probably because of the presence of another Sec31 homologue, Sec31B. Sec31B shares 47.3% identity to Sec31A and may be able to functionally compensate for the loss of Sec31A, though Sec31B may play a lesser role in p125A membrane association (Shugrue et al., 1999; Tang et al., 2000; Stankewich et al., 2006). Similar to Sec13 depleted cells, the Golgi in p125A depleted cells also appeared fragmented and was dispersed around the perinuclear region. When p125A was expressed at moderate levels, it induces the coalescence of ERES and ER-Golgi interface at the perinuclear region (Shimoi et al., 2005, Iinuma et al., 2007). These observations suggest that p125A not only plays a part in the organisation of the ERES, but also required for the maintenance of the Golgi integrity. How does p125A function in maintaining the Golgi integrity is not known. The Golgi disruption phenotype may be a direct or indirect effect of the function of p125A. A fragmented Golgi phenotype was also observed in cells depleted of tethering factors or SNAREs such as GM130, p115, Syntaxin 5, Syntaxin 18, etc., which are involved in ER-to-Golgi transport (Suga et al., 2005; Marra et al., 2007; Satoh et al., 2008; Iinuma et al., 2009). The lack of these proteins prevented the homotypic fusion of ER-Golgi intermediate compartments, and/or the fusion of ER-Golgi intermediate compartments with the cis-Golgi. Evidence suggested that the continual flux of proteins and membrane from the ER to the Golgi is essential for the maintenance of 115 the Golgi ribbon (Shorter et al., 1999; Zolov et al., 2005). However, it is still unclear if these membrane fusion processes require distinct factors or are simply vesiclefusion that occurs as part of the secretory traffic (Puthenveedu et al., 2006). Why the depletion of p125A should affect the Golgi structure is not clear. As p125A contains a possible dimerization domain (SAM domain), it may function as a factor that is required for the homotypic fusion of vesicles of the ER intermediate compartments. Another possibility is that p125A, like Sec13, is required for the anterograde transport from the ER to the Golgi. Consequently, depleting Sec13 or p125A may impair this pathway and reduce the number of vesicles budding from the ERES. As large amounts of ER membrane is required to maintain Golgi, the reduction in membrane input from the ER also means that the amount of specific proteins that may be necessary for Golgi maintenance is also reduced. Therefore, a lack of anterograde transport from the ER to the Golgi may lead to the altered Golgi integrity. Several studies have revealed that even though Golgi integrity is affected, secretory cargo could still be delivered to the Golgi complex, albeit with a lower efficiency (Kondylis and Rabouille, 2003; Marra et al., 2007; Diao et al., 2008). For example, in GM130 depleted cells, the Golgi is abberated. However, the transport of the VSVG protein into the Golgi was only delayed but not inhibited. Glycosylation of the VSVG protein in these cells were also not affected (Marra et al., 2007). Similar results were observed in p125A depleted cells. The export of VSVG protein from the ER to the medial-Golgi was delayed, as monitored by the acquisition of EndoH resistance (Figure 3.34). The rate of Golgi reassembly after BFA treatment was also delayed in p125A depleted cells (Figure 3.30). These results imply that p125A may be involved in protein export from the ERES. 116 As overexpression and depletion of p125A affected the normal organisation of the ERES (Tani et al., 1999; Shimoi et al., 2005), it may play a role in the organisation of the ERES (Shimoi et al., 2005; Iinuma et al., 2007). A recent study demonstrated that phosphatidylinositol 4-phosphate can induce recruitment of Sec23/Sec24 complex to the membrane by Sar1 in an ATP independent manner, suggesting that phospholipids have an active role in the nucleation of ERES (Blumental-Perry et al., 2006). p125A also contains a phospholipase A1-like domain (Tani et al., 1999), and was found to bind phosphatidylinositol phosphate (Iinuma et al., 2007). p125A binds to Sec23A via residues 135-259 (Mizoguchi et al., 2002) and Sec31A via residues 260-600. Sec31A binds p125A via its C-terminal 180 amino acids (residues 1040 -1220) and to Sec23A via residues 879-1114 (Bi et al., 2007). These binding sites are adjacent to each other, suggesting that the interactions of all three proteins may occur in close proximity to one another. The ability of p125A to interact with COPII proteins and its possible interaction with membrane lipids suggests that p125A may have a role in anchoring COPII components to the membrane, stabilising the interactions between the inner and outer layer of the COPII coat. In p125A depleted cells, the levels of Sec13 and Sec31A associated with the membrane was substantially reduced, while the levels of Sec23A was slightly but not abnomally affected (Figure 3.36). Membrane association of Sec13 and Sec31A were not completely abolished in the p125A depleted cells, probably because both Sec13 and Sec31A have interactions with other membrane associated proteins. Sec31A binds to Sec23A and Sar1 on the membrane (Bi et al., 2007) and Sec13 was found to be strongly associated with KIAA0310p, a mammalian homologue of yeast Sec16 which is recruited to the ERES by Sar1 (Iinuma et al., 2007). These results suggest that p125A may function downstream of Sec23A recruitment, and that the association 117 of Sec13 and Sec31A, but not Sec23A to the membrane, is partially dependent on p125A. Differential centrifugation of membrane and cytosolic fractions showed that the majority of p125A was distributed to the membrane. p125A specific antibodies also stained punctate structures that colocalizes very extensively to ERES (Shimoi et al., 2005, current work). It was previously proposed that the phopholipase homology domain of p125A is a primary determinant for membrane attachment (Tani et al., 1999, Shimoi et al., 2005), and the N-terminal region that interacts with Sec23A coordinates membrane specificity (Mizoguchi et al., 2000). However, from the current study, there is evidence to suggest that p125A associates to the membrane via its interaction with Sec31A. Upon depletion of Sec31A, the association of p125A with the membrane was greatly reduced (Figure 3.23 and Figure 3.35), indicating that Sec31A is essential for the attachment of p125A to the membrane, suggesting that rather than anchoring the Sec13/31 complex to the ERES, p125A may serve to stabilise the association of the heterotetramer to the membrane. The mammalian homologue of Sec16p, KIAA0310p (Watson et al., 2006, Iinuma et al., 2007, was proposed to function in cooperation with p125A to nucleate ERES (Iinuma et al., 2007). KIAA3010p was identified together with p125A in a GSTSec23A pulldown experiment (Tani et al., 1999). Sec16p is a large ER membrane peripheral protein that is required for transport vesicle budding from the ER (Supek et al., 2002, Watson et al., 2006). It was found to interact with Sec23 (Espenshade et al, 1995; Gimeno et al., 1996; Shaywitz et al., 1997; Iinuma et al., 2007), Sec24 (Gimeno et al., 1996; Shaywitz et al., 1997; Iinuma et al., 2007), Sec31 (Gimeno et al., 1996; Shaywitz et al., 1997) and Sec13 (Iinuma et al., 2007). 118 Overexpression of either KIAA0310p or p125A results in different effects on marker proteins of the early secretory pathway. Overexpression of KIAA0310p caused a marked decrease in the perinuclear and peripheral staining for Sec23A and Sec31A (Iinuma et al., 2007), which are ERES markers (Paccaud et al., 1996; Tang et al., 2000). In KIAA0310p overexpressed cells, the ERGIC marker, ERGIC53 (Klumperman et al., 1998; Appenzeller et al., 1999; Hauri et al., 2000) was redistributed to the ER, while GM130, a cis-Golgi protein (Nakamura et al., 1995) was diffused (Iinuma et al., 2007). On the other hand, overexpression of p125A induced aggregation of Sec31A to p125A positive aggregates at the perinuclear region (Tani et al., 1999; Iinuma et al., 2007). Furthermore, both ERGIC53 and GM130 were also observed in p125A positive aggregates (Iinuma et al., 2007). The different effects of overexpression of p125A and KIAA0310p on ERES, ERGIC and the Golgi may indicate the different roles of p125A and KIAA0310p in the organisation and function of ERES. Depletion of either p125A or KIAA0310p also did not induce a similar effect on the ERES. Depletion of KIAA0310p abolished almost all the punctate staining of ERES markers (Watson et al., 2006; Bhattacharyya et al., 2007; Iinuma et al., 2007). In p125A depleted cells, all ERES showed reduced labelling of COPII. These results suggest that KIAA0310p may function to nucleate the ERES (Watson et al., 2006; Bhattacharyya et al., 2007; Iinuma et al., 2007), while p125A may be responsible for stabilising the COPII coat assembly. The Golgi was fragmented in both p125A and KIAA3010p knockdown cells (Watson et al., 2006; Iinuma et al., 2007; current work). In KIAA0310p depleted cells, Golgi reassembly after BFA treatment was markedly inhibited (Bhattacharyya et al., 2007), 119 and VSVG transport to the plasma membrane was impaired (Watson et al., 2006; Iinuma et al., 2007). A less severe phenotype was observed in p125A knockdown cells. The rates for Golgi reassembly and VSVG transport were delayed but not completely inhibited. This could be due to the abolished and/or reduced rate of COPII vesicle formation in p125A depleted cells. In view of KIAA0310p’s role in nucleating the ERES (Watson et al., 2006; Bhattacharyya et al., 2007; Iinuma et al., 2007), and that p125A stabilises the COPII coat, it would be interesting to study the combinatorial knockdown effect of both p125A and KIAA0310p on the ERES, ERGIC, Golgi and protein trafficking. [...]... Bhattacharyya et al., 2007; Iinuma et al., 2007) In p12 5A depleted cells, all ERES showed reduced labelling of COPII These results suggest that KIAA0310p may function to nucleate the ERES (Watson et al., 2006; Bhattacharyya et al., 2007; Iinuma et al., 2007), while p12 5A may be responsible for stabilising the COPII coat assembly The Golgi was fragmented in both p12 5A and KIAA3010p knockdown cells (Watson... either KIAA0310p or p12 5A results in different effects on marker proteins of the early secretory pathway Overexpression of KIAA0310p caused a marked decrease in the perinuclear and peripheral staining for Sec2 3A and Sec3 1A (Iinuma et al., 2007), which are ERES markers (Paccaud et al., 1996; Tang et al., 2000) In KIAA0310p overexpressed cells, the ERGIC marker, ERGIC53 (Klumperman et al., 1998; Appenzeller... (Watson et al., 2006; Iinuma et al., 2007; current work) In KIAA0310p depleted cells, Golgi reassembly after BFA treatment was markedly inhibited (Bhattacharyya et al., 2007), 119 and VSVG transport to the plasma membrane was impaired (Watson et al., 2006; Iinuma et al., 2007) A less severe phenotype was observed in p12 5A knockdown cells The rates for Golgi reassembly and VSVG transport were delayed but... (Iinuma et al., 2007) The different effects of overexpression of p12 5A and KIAA0310p on ERES, ERGIC and the Golgi may indicate the different roles of p12 5A and KIAA0310p in the organisation and function of ERES Depletion of either p12 5A or KIAA0310p also did not induce a similar effect on the ERES Depletion of KIAA0310p abolished almost all the punctate staining of ERES markers (Watson et al., 2006; Bhattacharyya... due to the abolished and/or reduced rate of COPII vesicle formation in p12 5A depleted cells In view of KIAA0310p’s role in nucleating the ERES (Watson et al., 2006; Bhattacharyya et al., 2007; Iinuma et al., 2007), and that p12 5A stabilises the COPII coat, it would be interesting to study the combinatorial knockdown effect of both p12 5A and KIAA0310p on the ERES, ERGIC, Golgi and protein trafficking... al., 1999; Hauri et al., 2000) was redistributed to the ER, while GM130, a cis-Golgi protein (Nakamura et al., 1995) was diffused (Iinuma et al., 2007) On the other hand, overexpression of p12 5A induced aggregation of Sec3 1A to p12 5A positive aggregates at the perinuclear region (Tani et al., 1999; Iinuma et al., 2007) Furthermore, both ERGIC53 and GM130 were also observed in p12 5A positive aggregates . p12 5A or KIAA0310p also did not induce a similar effect on the ERES. Depletion of KIAA0310p abolished almost all the punctate staining of ERES markers (Watson et al., 2006; Bhattacharyya et al.,. effects on marker proteins of the early secretory pathway. Overexpression of KIAA0310p caused a marked decrease in the perinuclear and peripheral staining for Sec2 3A and Sec3 1A (Iinuma et al., 2007),. the inner layer (Bi et al., 2007). As p12 5A can bind to both Sec2 3A (Tani et al., 1999; Nakajima et al., 2002; Shimoi et al., 2005) and Sec3 1A, and have the ability to bind to phosphatidylinositol