Emergent properties of proteostasis-COPII coupled systems in human health and disease

Abstract

In eukaryotic membrane trafficking, emergent protein folding pathways dictated by the proteostasis network (the ‘PN’) in each cell type are linked to the coat protein complex II (COPII) system that initiates transport through the exocytic pathway. These coupled pathways direct the transit of protein cargo from the endoplasmic reticulum (ER) to diverse subcellular and extracellular destinations. Understanding how the COPII system selectively manages the trafficking of distinct folded states of nascent cargo (comprising one-third of the proteins synthesized by the eukaryotic genome) in close cooperation with the PN remains a formidable challenge to the field. Whereas the PN may contain a thousand component, the minimal COPII coat components that drive all vesicle budding from the ER include Sar1 (a GTPase), Sec12 (a guanine nucleotide exchange factor), Sec23-Sec24 complexes (protein cargo selectors) and the Sec13-Sec31 complex (that functions as a protein cargo collector and as a polymeric lattice generator to promote vesicle budding). A wealth of data suggests a hierarchical role of the PN and COPII components in coupling protein folding with recruitment and assembly of vesicle coats on the ER. In this minireview, we focus on insights recently gained from the study of inherited human disease states of the COPII machinery. We explore the relevance of the COPII system to human biology in the context of its inherent link with the remarkably flexible folding capacity of the PN in each cell type and in response to the environment. The pharmacological manipulation of this coupled system has important therapeutic implications for restoration of function in human disease.

Keywords::

• COPII
• membrane trafficking
• chaperone folding
• proteostasis network
• inherited disease

Introduction

Exit of all proteins from the ER is managed by two universal systems; the conserved protein homeostasis or proteostasis network (PN) and the coat protein complex II (COPII) system (PN-COPII). The PN directs substrate protein folding by altering its folding energy landscape using a vast array of chaperones and degradation machineries (for recent reviews, see Balch et al. 2008, Morimoto 2008, Hutt et al. 2009, Powers et al. 2009, Prahlad and Morimoto 2009, Gidalevitz et al. 2010). The COPII system facilitates the selection and/or concentration of export permissive proteins and subsequent transport of cargo-laden vesicles from the ER (for recent reviews, see Gurkan et al. 2006, Fromme et al. 2008, Dancourt and Barlowe 2010, Russell and Stagg 2010). Many inherited and sporadic human diseases are linked to interruptions in protein cargo trafficking from the ER in response to the failure of the PN to prepare the protein fold for recognition by the COPII machinery (Powers et al. 2009).

The activity of the PN in protein folding is energetically linked to the construction of a multi-layered polymeric lattice that forms the COPII coat (Gurkan et al. 2006, Stagg et al. 2006, Stagg et al. 2008, Russell and Stagg 2010). COPII collects exportable cargo into lipid vesicle carriers that bud from the ER and transit to the Golgi for delivery to downstream compartments. Failure of protein cargo to be properly managed by the PN in response to amino acid substitutions/deletions within the polypeptide chain leading to misfolding (Powers et al. 2009), can result in decreased capture of the cargo by the COPII system and increased targeting to ER-associated degradation (ERAD). This can lead to numerous inherited diseases as a consequence of a loss of function and/or gain-of-toxic function phenotype. Surprisingly, although COPII genes are essential in yeast, defects in ER export and trafficking due to inherited mutations in the COPII machinery are not necessarily embryonically lethal. Rather, their deleterious effects are remarkably selective for specific types of cargo in particular cell and tissue types. Herein, we will highlight recent evidence of the effect of mutations in COPII components in protein trafficking through the exocytic pathway. We will discuss how the concept of protein substrate ‘selectivity’ by COPII impacts our understanding of the specificity of the operation of the PN and COPII trafficking pathways in generating a functional eukaryotic cell. We argue for a coupled role of the PN and COPII system in the differential management of cell, tissue and host protein substrate biologies that are responsible for both human health and disease.

Role of the proteostasis network in trafficking biology

Protein biogenesis and function in biology is largely if not exclusively regulated and maintained by the protein homeostasis or proteostasis network (PN) (reviewed in Balch et al. 2008, Hutt et al. 2009, Powers et al. 2009, Hutt and Balch 2010). The PN is an integrated biological system that generates and protects the protein fold, and removes the protein in response to misfolding and/or loss of need (Ron and Walter 2007, Morimoto 2008, Hutt et al. 2009, Prahlad and Morimoto 2009, Gidalevitz et al. 2010, Hutt and Balch 2010, Korolchuk et al. 2010). It comprises over 1000 components including chaperones and folding enzymes such as the Hsc/p40/70 and Hsp90 systems (both general and specialized) (Ron and Walter 2007, Shamovsky and Nudler 2008), and a vast reservoir of degradation promoting components including cytosolic (proteasome) and membrane oriented (autophagy/lysosome) pathways (Konstantinova et al. 2008, Finley 2009). Importantly, multiple signaling pathways (Petersen et al. 2005, Morimoto 2008, Zhang and Kaufman 2008) regulate the composition and the concentration of PN components in response to developmental, environmental and misfolding protein challenges (Zhang and Kaufman 2008, Panowski and Dillin 2009, Prahlad and Morimoto 2009, Gidalevitz et al. 2010). Proteostasis networks not only differ vastly between cell types, but in response to genetic and/or environmental challenges to the cell/tissue/organism. Our appreciation of the emerging array of regulatory events responsible for the modulating the function of the PN in protein folding and function in human health and disease is in its infancy.

In considering the general function of the PN, an important point that is under-appreciated is that the composition of the local cellular PN sets an energetic baseline or ‘set-point’ for the ability of protein cargo to enter the exocytic pathway (Sekijima et al. 2005, Wiseman et al. 2007a, 2007b, Mu et al. 2008, Powers et al. 2009). This baseline is determined by the balance of folding and degrading PN components that can vary considerably between different cell types and can be altered by multiple signaling pathways (Powers et al. 2009, Hutt and Balch 2010). In other words, there is not a single quality control standard for ER export as the ER has no idea whether a protein has achieved a conformation that is functional. Rather, export is based on the general principle of folding energetics. Here, the folding of the polypeptide chain is variably managed by developmental and/or environmental factors that dictate the composition of the local PN. The set-point in each cell type is achieved through evolutionary constraints that reflect successful cell/tissue/organismal function (Wiseman et al. 2007b, Powers et al. 2009).

Thus, cargo folding plays a critical role in generating and maintaining membrane trafficking pathways. Indeed, in the absence of ‘recognizable’ folded protein cargo substrate, COPII budding does not occur (Aridor et al. 1999) (see below). Therefore, the PN is the initiating event in COPII biology. This is most evident during development. The ability of a plasma cell to generate high levels of immunoglobulin requires up-regulation of the PN (Brewer and Hendershot 2005, Mori 2009). At the protein substrate level, this is illustrated by the analysis of COPII interactions with the SREPB cleavage activating protein (SCAP), the sterol regulatory element binding protein-escort protein (Yellaturu et al. 2009, Jo and Debose-Boyd 2010, Sun et al. 2007). Here, alteration of the spacing between SCAP's ER exit motif and the ER membrane crucially interfered with COPII recognition and export, suggesting that PN management of the exit motif presentation affects COPII function. Finally, at the physiological level, complex kinase/phosphatase signalling pathways regulate the COPII system at the levels of both cargo selection and organization of ER budding (Farhan et al. 2010). The variable set-point in each cell type provided by the PN has a major impact on the ability of the COPII machinery to recognize protein cargo as a kinetically and/or thermodynamically acceptable substrate for transport (Gurkan et al. 2006, Hutt et al. 2009, Powers et al. 2009, Hutt and Balch 2010).

The basics of the COPII system

As a primer for understanding inherited COPII diseases, we will briefly discuss the essential features of COPII function in membrane traffic. More extensive discussions can be found in recent reviews (Gurkan et al. 2006, Fromme et al. 2008, Dancourt and Barlowe 2010, Russell and Stagg 2010).

Anterograde transport of protein substrates compatible with the operation of the PN is mediated by the formation of COPII-coated vesicles that bud from the ER. In the yeast Saccharomyces cerevisiae, COPII formation takes place randomly across the ER membrane (Rossanese et al. 1999), whereas in Pichia pastoris and mammalian cells budding is more site-specific. Coat formation occurs on ribosome free sub-domains of the ER referred to as transitional ER (t-ER), vesicular-tubular clusters (VTCs) or ER exit sites (ERES), identified morphologically by the presence of COPII proteins (Budnik and Stephens 2009). Typically, in P. pastoris the number of ERES is low (around 2–6) (Rossanese et al. 1999), whereas in mammalian cells this can vary up to a few hundred, half of which are clustered in a juxtanuclear array around the Golgi (Budnik and Stephens 2009). The size of ERES is normally in the range of several hundred nanometers, but is dependent on the kinetics of COPII activity and the local protein cargo composition and load level. ERES are relatively immobile structures when viewed in real-time, though their position is maintained by the combined activity of kinesin-1 and dynein-1 motor proteins (Gupta et al. 2008). Thus, differential construction of ERES as mentioned above may offer opportunities for the PN and COPII to cooperate in more effectively managing the export of specific types of cargo with special folding and trafficking needs.

Assembly of the COPII coat on ERES is a hierarchical process. It is initiated by the membrane localization of the GTPase Sar1, induced by the GDP/GTP exchange catalyzed by the guanine exchange factor, Sec12. The insertion of activated Sar1 into the lipid bilayer (possibly in the proximity of ERES) initiates membrane deformation that ultimately produces a curved vesicle (Bielli et al. 2005, Lee et al. 2005). Interestingly, the catalytic cycle of Sar1 appears to be the sole energy dependent event throughout the assembly process from recruiting protein cargo to the ultimate step triggering vesicle fission.

The activated ER-Golgi intermediate (ERGIC) Sar1 recruits the next layer of the COPII machinery to the ERES. Via direct interactions with Sec23, Sar1 engages the cytosolic Sec23–Sec24 (Sec23–24) adaptor complex to recruit cargo to be transported, forming pre-budding complexes (Gurkan et al. 2006, Stagg et al. 2006, Gurkan et al. 2007, Fromme et al. 2008). Sec23 can function as a GTPase activating protein (GAP) for Sar1, catalyzing hydrolysis by the insertion of an arginine finger into Sar1's active site. This drastically reduces the lifetime of GTP bound to Sar1 and is likely to play an important role in protein cargo selection as well as vesicle budding by functioning as a timer linking cargo recruitment with coat assembly and/or vesicle fission (Antonny et al. 2001, Bielli et al. 2005, Lee et al. 2005, Long et al. 2010). Though Sec23 and Sec24 are structurally very similar, Sec24 is the primary protein cargo adaptor in the COPII system. Three different binding pockets, all facing the ER bilayer, facilitate the interaction of Sec24 with specific ER-exit motifs that are cytoplasmically exposed on transmembrane (TM) cargo, or on TM cargo receptors in the case of soluble, luminal proteins (Mossessova et al. 2003, Fath et al. 2007, Mancias and Goldberg 2008). This also includes interactions with SNARES, a group of membrane-anchored proteins that are continuously recycled and conduct downstream fusion of the nascent vesicle with target pre-Golgi/Golgi membranes. Recognition of a wide variety of cargo export signals is further facilitated by four different isoforms of Sec24 (denoted as Sec24a–d). Interestingly, they play a major role in the unique phenotypes observed in human disease (see below).

Sec23–24 collects and concentrates nascent protein cargo (Tabata et al. 2009) in the inner shell of forming vesicles in response to the self-assembling properties of the outer shell Sec13–Sec31 (Sec13–31) complex. The heterotetrameric Sec13–31 subunits form a polymeric lattice around the Sec23–24 adaptor proteins (Stagg et al. 2007, 2008); the nucleation of COPII coat formation may be influenced in vivo by contacts between Sec31 and Sec23, although this remains to be directly demonstrated (Bi et al. 2007, Russell and Stagg 2010). Assembly of Sec13–31 into a polymeric lattice promotes vesicle budding via a process that utilizes Sar1 to catalyze fission from the ER membrane (Long et al. 2010). COPII generated vesicle carriers are thought to first deliver their protein cargo to the for further transit to the Golgi and latter post-Golgi destinations (Bi et al. 2007, Long et al. 2010).

A COPII vesicle classically ranges in size from ∼60–100 nm (Matsuoka et al. 1998), although it may possibly accommodate much larger cargo such as pro-collagen and chylomicron particles given the flexibility inherent in the geometry of the outer Sec13–31 polymeric cage (Stagg et al. 2006, 2008). Both X-ray crystallization and cryo-electron microscopy (cryo-EM) have been powerful tools for resolving structural features of the COPII system; from partial fragments of the Sar1–Sec23–24 and Sar1–Sec23–Sec31 complexes (Bi et al. 2002, 2007), to the Sec13–31 heterotetrameric unit (Stagg et al. 2006, 2008, Fath et al. 2007, Lee and Goldberg 2010) to the octahedral cage (Sec13–13 alone) (Stagg et al. 2006, 2008) as well as the complete COPII coat (Sec23–24 plus Sec13–31) (Stagg et al. 2006, 2008, reviewed in Russell and Stagg 2010). Linking the data from both X-ray and cryo-EM approaches has been instrumental in refining our structural knowledge of the COPII coat (Gurkan et al. 2006, Lee and Goldberg 2010, Russell and Stagg 2010).

Each step of the COPII vesicle formation cycle as described above is controlled by the kinetic and thermodynamic parameters of the complete PN-COPII system (Wiseman et al. 2007b, Hutt et al. 2009, Powers et al. 2009, Hutt and Balch 2010) (Figure 1). These steps include: (1) The steady state activity of the Sar1–Sec12 activation step that maintains Sar1-GTP at the ER membrane; (2) the energetics protein cargo folding by the PN coupled to the differential rates of recruitment of protein cargo via their exit codes to the different Sec23–24 isoforms; (3) differential interactions of the isoform specific Sec23–24–cargo complexes with Sec13–31 cage components leading to Sec13–31 self-assembly to build the outer cage that collects and concentrates cargo; and (4) Sar1-dependent processes that are involved in the generation of membrane curvature and fission. Each of these processes appears to be strongly linked to folding dynamics and cargo size (Gurkan et al. 2006, Long et al. 2010). Thus, each of these kinetic and thermodynamic parameters is probably unique to each cargo, cell and tissue type, and COPII component composition. We propose that they are all likely to contribute to the phenotype of normal and diseased human (patho)physiology(s).

Figure 1. The PN-COPII coupled system in health and disease. On an energy landscape, protein cargo folds to its native state via energy minima (folding intermediates) and maxima (the peak energy requirement for that step) illustrated by the blue (wild-type [WT]) and green (mutant) curves. Similarly during vesicle formation, COPII components energetically and kinetically interact with protein cargo and with one another (Gurkan et al. 2006). Again, each step can be described as a minima on the energy landscape (black (WT)) with a kinetic/thermodynamic barrier (maxima) separating progress through the COPII assembly pathway. The coupled PN-COPII ER export system can be described in detail as follows: (1) COPII assembly is initiated by the exchange of GDP for GTP on Sar1 by Sec12. Sar1-GTP inserts into the membrane and perhaps begins to initiate membrane curvature even at this early step. (2) Cargo is folded (from unfolded (U) or misfolded (M) states via intermediates (I) to energetically favorable functional states (N)) in the ER as indicated by the blue (WT) and green (mutant) energy profiles in the folding landscapes. Folding is in response to the local PN. This can be adjusted through the activity of multiple cytosolic (in the case of transmembrane (TM) protein cargo) and luminal (TM and soluble protein cargo) chaperones. Whereas the ‘WT’ protein engages the COPII system when reaching a certain energetic state (which will vary depending on the individual protein's circumstances) (blue circle), the mutant protein fails to energetically engage the COPII system if additional folding constraints (green circle) are not satisfied by the local PN. The arrival of the Sec23–24 complex to the ER membrane in response to Sar1-GTP engages exportable protein cargo (pending the activity of the local PN) and recruits this cargo through interactions with the highly dynamic Sar1–Sec23–24 complex that may abort in the absence of transport permissive protein cargo. The Sar1–Sec23–24 protein cargo complex (referred to as a ‘pre-budding complex’ (Aridor et al. 1998, Gurkan et al. 2006)) induces further membrane deformation that has an energetic/kinetic cost as indicated by the black energy landscape profile. (3) Pre-budding complexes recruit Sec13–31 subunits that self-assemble (Stagg et al. 2006) to form an outer structural cage that further deforms the membrane; again this is associated with a thermodynamic/kinetic cost. (4) Newly recruited Sar1-GTP molecules facilitate membrane curvature (Settles et al. 2010) to form a fission pore and upon GTP hydrolysis cause membrane fission releasing a productive COPII vesicle (Long et al. 2010). While the above steps 1-4 describe in a simplistic fashion the minimal operation of the wild-type COPII system, when diseases affect COPII machinery function, the energy barrier for the step associated with the specific function of the affected COPII component may be too high for successful interactions, stalling export at an intermediate state of the PN-COPII coupled assembly event leading to abortive budding (red). For example, mutations interfering with the binding sites of Sec24 may prevent cargo binding and progress would stall at an early step with cargo accumulating in the ER pending activity of the PN (step 2). This event could trigger the unfolded protein response (UPR) and place the cell/tissue in jeopardy. Specific mutations in Sec23 could affect many protein:protein interactions (e.g., with Sar1, Sec24, Sec31) (steps 2–4) given its apparent central role in COPII function. In the case of mutations that affect Sec13 or Sec31 function, it would be anticipated that the pre-budding complex forms and either accumulates in partial cage structures or is unstable and disassembles after failure to properly engage the Sec13–31 machinery in a way that correctly generates the polymeric lattice. In either case, the final stage of COPII cage formation does not occur in a productive way to generate Sar1 triggered membrane curvature or fission (step 4). Finally, Sar1 mutations, although required for initiating the COPII system function, could interfere at many steps of the process given its projected role also in fission (Lee et al. 2005, Long et al. 2010). For example, in CMRD, chylomicrons are visible within extended membranous compartments that have not budded, suggesting a failure to release the cargo-laden vesicles in an unknown fashion (step 4). Given that the mutation in each protein cargo or COPII component will affect functionality of the coupled PN-COPII export system in different ways, it is evident that there are many intermediate steps in the PN-COPII system that can be disrupted leading to disease, and each of these steps is tied to energetics of the local proteostasis environment. Treatment of diseases by altering the energetics of the PN-COPII system through pharmacological intervention (downward arrow on the COPII energy landscape) may lead to restoration of COPII-mediated export.

While Sar1, Sec23–24, and Sec13–31 are the minimal machinery required to reconstitute COPII-dependent budding in vitro, when considering the larger picture in vivo, a plethora of factors have been found which affect the ‘trinity’ of membrane traffic (e.g., protein cargo capture, vesicle coat assembly/fission and vesicle targeting/fusion) (Table I). These factors have been found to affect the above processes based on siRNA silencing studies, interaction data and/or over-expression phenotypes in vivo. Although the role for most of these factors is unclear at this juncture, interestingly, they mostly appear to target the function of Sec23, the GAP for Sar1, and the structural intermediates that are sensitive to Sec24 and Sec31 interactions driving coat assembly (Russell and Stagg 2010). Thus, Sec23 may be involved in unanticipated ways with coordinating the central task of handling a large diversity of protein cargo with folding environments unique to different cell, tissue and organismal physiologies.

Table I. Accessory factors involved in COPII function.

Protein	Function	Reference
ALG-2	ALG-2 localizes to ERES via Ca^2+-dependent interactions with Sec31 and stabilizes the localization of Sec31 at ERES.	(Bentley et al. 2010, Yamasaki et al. 2006)
Lysophospholipid acyltransferase (LPAT)	A small molecule inhibitor of an ER-associated LPAT (CL-976) blocks a very late fission step of COPII trafficking.	(Brown et al. 2008)
Nucleoside diphosphate kinase (Nm23H2)	Promotes COPII vesicle formation independent of its catalytic function. siRNA silencing of Nm23H2 leads to defective COPII assembly at steady state and reduced kinetics of ER export.	(Kapetanovich et al. 2005)
p125	Localizes to ERES and interacts with Sec23 and Sec31. SiRNA silencing disrupts perinuclear ERES and cis-Golgi. Displays homology to Phospholipase A1. Participates in the organization of ERES. Found in complex with Sec13–31 in the cytosol. Function unknown.	(Ong et al. 2010, Shimoi et al. 2005, Tani et al. 1999)
p150glued	p150glued, a subunit of the microtubule-dependent dynactin complex, directly interacts with Sec23. This interaction couples ER export to the microtubule cytoskeleton.	(Watson et al. 2005)
p38 MAP kinase	Negatively regulates COPII recruitment in nocodazole-arrested “mitotic” cytosol.	(Wang and Lucocq 2007)
PCTAIRE	Physically interacts with Sec23. siRNA depletion or over-expression of a kinase-inactive form inhibits COPII function at an early step early in ER export.	(Palmer et al. 2005)
p97	AAA NSF family member. p97 has been shown to support ERES biogenesis in a cell-free in vitro system.	(Roy et al. 2000)
Phospholipase D (PLD)	Upon activation by Sar1, PLD supports Sar1-dependent membrane tubulation, the subsequent recruitment of Sec23–24 and Sec13–31 COPII complexes to the ERES and ER export.	(Pathre et al. 2003)
PI4-kinase,(PtdIns4P)	PtdIns4P localize to ERES and supports Sar1-induced proliferation and constriction of ERES membranes. PI4K-IIIa siRNA silencing reduces the number of ERES.	(Blumental-Perry et al. 2006, Farhan et al. 2008)
Sec16	Acts as a platform and/or regulator for COPII assembly at ERES. Promotes the recruitment of Sec23–24 and Sec13–31 and stabilizes the pre-budding complex on synthetic liposomes. Over-expression as well depletion causes disorganization of ERES and the inhibition of ER to Golgi transport. Physically interacts with COPII components Sec23 and Sec13.	(Connerly et al. 2005, Espenshade et al. 1995, Hughes et al. 2009, Watson et al. 2006)
Sed4p*	Sed4p inhibits the GAP action of Sec23p on Sar1p.	(Saito-Nakano and Nakano 2000)
Smy2p*	Assists with COPII vesicle formation by interacting with Sec23p-Sec24p complex.	(Higashio et al. 2008)
STAM	Physically interacts with Sec13–31. Localizes mainly to ERES (depending on Sar1 activity). Over-expression causes fragmentation whereas siRNA depletion induces condensation of the Golgi. Both over-expression and depletion inhibit VSVG trafficking from the ER.	(Rismanchi et al. 2009)
TANGO-1	Promotes cargo loading at ERES and physically interacts with Sec23–24.	(Saito et al. 2009)
Yip1A	Enriched in ERES. Yip1A efficiently incorporates into COPII vesicles, colocalizes with Sec13-31 and physically interacts with Sec23–24. Over-expression of cytoplasmic domain inhibits VSVG trafficking from the ER to Golgi.	(Heidtman et al. 2003) (Tang et al. 2001)

*Only the yeast homologue has been identified.

COPII and the accessory components described above are integral to the budding process. However, it is important to emphasize that protein cargo plays an essential role in vesicle formation in vivo. Interestingly, COPII vesicle formation on synthetic liposomes in vitro was not hindered by the absence of cargo, leading to early views that that cargo is not required for vesicle formation (Matsuoka et al. 1998). In contrast, COPII vesicle formation in vivo requires the presence of transiting cargo (Aridor et al. 1999) with more recent experiments demonstrating that cargo can directly influence COPII vesicle formation (Guo and Linstedt 2006, Farhan et al. 2008). This is consistent with the fact that cargo has been shown to increase the affinity and stability of COPII components (e.g., Sec23–24 and Sar1) towards the membrane at the ERES (Forster et al. 2006, Tabata et al. 2009). Another mechanism wherein cargo could regulate COPII vesicle formation is by functioning as an initiator of a priming complex that recruits/stabilizes Sar1, its GAP and/or one or more COPII subunits. This model is supported by experimental data wherein COPII subunits interact physically with SNARE proteins (Mossessova et al. 2003). Concentration of an unassembled form of the SNARE Sec22 at ERES has been suggested to be mediated by the selective COPII recognition of a PN managed conformational epitope (Mancias and Goldberg 2007). Moreover, modeling along with experimental work has revealed that there is a strong correlation between the amount of available cargo and the number and size of ERES (Aridor et al. 1999) (Heinzer et al. 2008). Finally, as exemplified by the plasma cell, an increase in cargo load leads to increase in ERES size and number to cope with extra secretory flux (Farhan et al. 2008) that is closely linked to the up-regulation of the PN. Thus, we need to further consider the potential central role of cargo in the operation of the PN in COPII disease.

Diseases caused by aberrant COPII trafficking

A number of human diseases are now recognized to be caused by inherited mutations in COPII genes; to date they are all isoform specific and autosomal recessive syndromes. These are summarized in Table II. Salient features of how even the smallest variations in the function of COPII components can be detrimental to normal health and development are discussed below and summarized in Figure 1, highlighting the increasing awareness that the COPII machinery is a highly tuned machine driven by the energetic and kinetic parameters governing vesicle coat assembly in response to the protein fold and the operation of the PN. Each of the COPII component diseases described below illustrate defects in the export of cargo in specific cell, tissue and developmental environments, suggesting that COPII utilizes the activity of the PN system to specialize the exocytic pathway for optimal cell, tissue and organismal function.

Table II. Effects of mutations in essential COPII components.

Gene	COPII effects	Species	Disease	Phenotype	References
Sar1b	Truncated or non-functional Sar1b proteins decrease chylomicron transport in enterocytes	Human	Chylomicron retention disease (CMRD)/Anderson disease (ANDD)	Severe fat malabsorption, poor development in infancy	(Annesi et al. 2007, Charcosset et al. 2008, Jones et al. 2003, Treepongkaruna et al. 2009)
Sec23a	F382L substitution decreases Sec23a affinity for Sec13-31	Human	Craniolenticulo-sutural dysplasia (CLSD)	Skeletal and craniofacial development defects	(Boyadjiev et al. 2006, Fromme et al. 2007)
	L402X forms truncated protein	Zebrafish	crusher mutant		(Lang et al. 2006)
Sec23b	22 mis-sense or nonsense mutations result in non-functional Sec23b, which is critical for correct erythroid differentiation	Human	Congenital dyserythro-poietic anemia type II (CDAII)	Splenomegaly, gallstones and iron overload leading to liver failure	(Iolascon et al. 2009, Schwarz et al. 2009)
Sec24a	missense mutation prevents correct Sec24a ER distribution	Arabidopsis thaliana	N/A	Secretory and Golgi proteins accumulate in ER	(Faso et al. 2009)
Sec24b	Truncated Sec24b no longer recognizes a critical cargo, Vangl2	Mice	N/A	Severe birth defects such as failed neural tube fusion	(Merte et al. 2010)
Sec24d	Truncated Sec24d no longer recognizes collagen II for transport	Medaka	N/A	Skeletal and facial development defects	(Ohisa et al. 2010)
Sec13	Suppression of Sec13 expression leads to defects in proteoglycan deposition	Zebrafish	Similar to CLSD	Skeletal and cranial development defects	(Townley et al. 2008)

Sar1 isoform-specific disease

In mammals, there are two isoforms of Sar1, Sar1a and Sar1b. While Sar1a has yet to be associated with any human diseases, defects in the gene encoding for Sar1b are implicated in the rare recessive disorder Chylomicron Retention Disease (CMRD) – also known as Anderson disease (ANDD). CMRD is characterized by the selective retention of chylomicron-like particles within the ER of enterocytes. The lack of chylomicrons in the blood leads to severe fat mal-absorption, a deficiency of fat-soluble vitamins and ultimately, poor development during infancy (Jones et al. 2003, Annesi et al. 2007, Treepongkaruna et al. 2009). From the CMRD cases studied to date, 11 separate mutations in SARA2/SAR1B gene have been detected (Jones et al. 2003, Charcosset et al. 2008). They have been predicted to produce either truncated or non-functional Sar1b proteins (Jones et al. 2003, Charcosset et al. 2008), thus supporting a role for COPII in chylomicron export. While it is still uncertain how the COPII machinery can handle large cargo such as chylomicrons, as they vastly exceed the typical COPII particle size (Stagg et al. 2006, 2008), these results support the hypothesis that Sar1b is likely to have specific kinetic or thermodynamic properties of its protein fold that are required to facilitate the assembly of a more extensive coat lattice necessary to capture these large particles during exit from the ER (Stagg et al. 2006, 2008). Sar1's ability to modulate membrane curvature (Aridor et al. 2001, Long et al. 2010, Settles et al. 2010) may also play an important role in isoform-specific functions. Given that Sar1 is an essential gene in yeast it is interesting that Sar1a, like the many other COPII isoforms discussed below, appears to handle a sufficient level of general COPII activity to prevent embryonic lethality in the absence of functional Sar1b. Thus, Sar1b may have evolved to handle more specialized types of export.

Sec23 isoform-specific disease

In mammals, Sec23 has 2 isoforms, Sec23a and Sec23b. Defects in either isoform can have serious consequences. For example, a single amino acid substitution (F382L) in Sec23a causes CLSD or Cranio-Lenticulo-Sutural Dysplasia. CLSD is an autosomal recessive syndrome characterized by skeletal and facial defects (Boyadjiev et al. 2006). Patients have ER retention of secretory proteins that are required for normal morphogenesis. The molecular cause of trafficking disruption is believed to be that Sec23a–F382L has a reduced affinity for Sec13–31 (Fromme et al. 2007). The fact that its orthologous residue in the yeast Sec23a protein maps close to the binding site for the Sec31 active fragment promoting GTP hydrolysis (Bi et al. 2002) is further structural evidence for this hypothesis. Similarly, the mutation responsible for the crusher phenotype in of Zebrafish (D. rerio), which also induces craniofacial defects, has been mapped to a nonsense variant (L402X) of Sec23a (Lang et al. 2006). The crusher chondrocytes accumulate proteins in a distended ER, resulting in severe reduction of cartilage extracellular matrix (ECM) deposits, including type II collagen (Lang et al. 2006). These results highlight the predominant trafficking role of Sec23a for procollagen export in bone development, a role that is apparently not managed by the evolutionarily divergent Sec23b isoform that would still be present in the absence of Sec23a. The high degree of specialization of the PN for procollagen folding (Fromme and Schekman 2005) argues for a tight coupling between these processes.

In contrast to the requirement for Sec23a in procollagen secretion, Sec23b is the dominant isoform essential for red blood cell maturation (Schwarz et al. 2009). Mutations in the SEC23B gene are responsible for Congenital Dyserythropoietic Anemia type II (CDAII), the most prevalent form of CDA (Schwarz et al. 2009). The peripheral red blood cells of affected individuals are bi- or multi-nucleated, have abnormal protein and lipid glycosylation, and contain a double plasma membrane. The second membrane is thought to be an extension of the endoplasmic reticulum (Iolascon et al. 1996), possibly related to autophagy PN pathways (Cuervo 2008, Yang and Klionsky 2010). Individuals with CDAII show progressive splenomegaly, gallstones and iron overload with potential liver cirrhosis or cardiac failure (Iolascon et al. 1996). To date 22 different mutations have been identified and CDAII patients can be divided into two groups: (i) Patients with two mis-sense mutations, and (ii) patients with one non-sense and one mis-sense mutation (Iolascon et al. 2009). The mutations map throughout the Sec23b gene indicating important roles for different domains in protein function (Schwarz et al. 2009). As Sec23 is perhaps the lynchpin of the adaptor complex promoting cargo selection through its interactions with Sar1, Sec24 and Sec31, and as the target for regulation of multiple accessory factors, it is understandable how many different single residue substitutions and truncations may affect COPII transport mechanistically. Why erythrocyte development is dependent on the divergent sequence of the Sec23b isoform over that of Sec23a remains a mystery, but further studies will undoubtedly reveal important facets of specialization by each of these isoforms (Schwarz et al. 2009).

Sec24 isoform-specific phenotypes

Unlike Sec23, which serves as a link to Sar1 and Sec13-31, Sec24's primary role in COPII vesicles is as the cargo “adaptor” element (Gurkan et al. 2006, Wendeler et al. 2007, Russell and Stagg 2010). To accommodate the large range of cargos (and their specific exit signals) that must be recognized for export, humans have evolved four isoforms of Sec24, each thought to contain at least three binding sites. The appearance of isoform-specific signal recognition can be explained structurally; the surface groove for binding the conserved IxM packing signal is open in Sec24c and Sec24d, but is occluded in Sec24a and Sec24b (Mancias and Goldberg 2008). Conversely, LxxLE/ME and DxE signals are only recognized by Sec24a and Sec24b subunits due to a particular Leu to Asp acid substitution near the B binding site in both Sec24c and Sec24d (Mancias and Goldberg 2008). Thus, human Sec24 isoforms expand the repertoire of protein cargo for signal-mediated ER export, but are in part functionally redundant, implying that exit codes must be presented in the proper context of the protein fold and that this can be disrupted by mutation (Wang et al. 2004). Therefore, it is not surprisingly that Sec24 mutations could cause marked reduction in the export of specific types of protein cargo that are dependent on particular isoform and/or tissue-specific expression patterns in response to developmental/environmental cues.

To date there are no known human diseases caused by a Sec24 mutation. However, clues as to the cargo selective nature of Sec24a were gained from a mutant of Arabidopsis thaliana (Faso et al. 2009). Whilst total loss of Sec24a function was lethal, a mis-sense mutation in the SEC24a gene (R693K substitution) led to a phenotypic partial accumulation of a select group of Golgi and secretory markers in globular, distended ER tubules still connected to the bulk ER. Although R693K mapped to a cargo binding region conserved in the Sec24 proteins of other organisms (Miller et al. 2003), it was shown that the mutation clearly affected the distribution of Sec24a to ER export sites (Faso et al. 2009). Further clues to Sec24a signal specificity was highlighted by vitro knockdown of Sec24a in HeLa cells. Knockdown selectively impaired transport mediated by di-leucine export signals (Wendeler et al. 2007). Double knockdown of different combinations of Sec24 isoforms uncovered a dominant role of Sec24a (Wendeler et al. 2007).

Sec24b has been shown to be essential for the proper ER export of proteins critical for neural tube development in mice (Merte et al. 2010). Mice expressing a truncated form of Sec24b, the Sec24b^Y613 mutant, exhibited Craniorachischisis, a rare but severe birth defect that resulted in a completely open neural tube with congenital fissure of the skull and vertebral column. Consistent with this proposal, Vangl2, a protein essential for proper neural tube closure, is selectively sorted by Sec24b for export from the ER (Merte et al. 2010).

The third isoform, Sec24c, has been implicated in the specificity of the docking reaction of the pre-chylomicron transport vesicle (PCTV) onto the Golgi (Siddiqi et al. 2009). The PCTV is a specialized ER to Golgi transport vesicle, and its docking onto the Golgi has been speculated to be a rate-limiting step in the transit of dietary fat across the intestinal epithelium. Immuno-depletion of Sec24c caused a nearly complete cessation of PCTV docking activity onto the Golgi in vitro (Siddiqi et al. 2009). On the addition of recombinant Sec24c, docking activity was restored. They conclude that the COPII proteins are present at the docking site of PCTV with the Golgi and that Sec24c is required for this event (Siddiqi et al. 2009). A potential role for the Sec23–24 coat components to also participate in late tethering steps of vesicle traffic from the ER to the Golgi is consistent with the interaction of the TRAPPI tethering complex (which modulates COPII vesicle delivery to the Golgi) with Sec23 (Cai et al. 2007, Trahey and Hay 2010). Sec24c (along with Sec24d) have recently been demonstrated to be favoured for GPI-anchored protein export (Bonnon et al. 2010), again emphasizing the selective role of Sec24 isoforms in protein cargo export.

Finally, the Sec24d isoform has been shown to be critical during developmental stages that rely heavily on extracellular matrix components (ECM) (Ohisa et al. 2010). A medaka (O. latipes) mutant, vertebra imperfecta (vbi), displays skeletal defects such as malformed craniofacial cartilage and vertebra. Positional cloning analysis of vbi genes revealed a nonsense mutation in the SEC24D gene leading to a premature stop codon in the C-terminal domain. Although the Sec24d mutant was stably expressed as truncated form in ECM-rich tissues of the vbi mutant, ER distention was visible and type II collagen accumulated cytoplasmically (Ohisa et al. 2010). Given that Sec23a is also required for type II collagen export, these results suggest that Sec23a and Sec24d may form a heterodimeric “team” optimized for this specific protein cargo, assuming that isoform function is evolutionarily conserved.

Sec13–Sec31 isoform-specific disease

The export of collagen has been shown to also heavily rely upon the correct recruitment of Sec13–31 (Townley et al. 2008). Sec13 depletion results in concomitant loss of Sec31 and a juxtanuclear clustering of pre-budding complexes containing Sec23–24 and procollagen cargo (Townley et al. 2008). Suppression of Sec13 expression in Zebrafish causes defects in proteoglycan deposition and skeletal abnormalities that are grossly similar to the craniofacial abnormalities of crusher mutant Zebrafish and CLSD patients in response to Sec23b mutations. Thus, efficient coupling of the inner (Sec23–24) and outer (Sec13–31) layers of the COPII coat is required to drive the export of procollagen from the ER, and that highly efficient COPII assembly is essential for normal craniofacial development during embryogenesis (Townley et al. 2008).

In addition to its central role in proteoglycan deposition, Sec13 has recently been suggested to interact directly with the cytosolic N-terminal fragment of presenilin-1 responsible for correct embryonic development in mice (Bonnon et al. 2010). Presenilin-1 is a major contributing factor to Alzheimer's disease (AD) in post-ER compartments during aging. Thus, given unknown PN sensitive events responsible for AD, Sec13 may have a surprising role in insuring a close coupling between the PN and trafficking components that contribute to the slow disease onset of AD in response to proteome imbalance (Hutt and Balch 2010).

The PN-COPII-system as an integrated system managing human health and disease

While the above analysis of human mutations probably only represents the tip of the iceberg in terms of insights into the specificity of COPII isoforms in human biology, it is consistent with the idea that the COPII system functions as part of a larger schema of interacting proteins to facilitate cargo movement through the exocytic pathway. We suggest that there exists a tight coordination between the variable ER dynamics of protein folding controlled by the PN and the composition of the COPII-based trafficking machinery, to drive the generation and function of specialized organelles that direct the function of different cell types. While the current set of human diseases seemingly highlight the importance of COPII isoforms on demanding protein cargo such as type II procollagen, this is likely to simply reflect the experimental approaches currently available to detect and characterize these gross phenotypic responses that occur in the absence of embryonic lethality. COPII selectivity for many non-essential genes may be difficult to interrogate given the subtlety of the phenotype: other essential genes may be embryonic lethal.

What is becoming clear is that the PN and COPII systems are both finely tuned, closely evolutionary linked machines; single point mutations can have disastrous effects on embryonic fate and/or tissue specific development through defects in protein cargo recognition transport. What we would like to propose is that cargo export from the ER by COPII is an extension of the proteostasis network. The COPII system detects kinetic and energetic parameters that influence the export ‘permissive’ state of the cargo it encounters in the context of PN function. Therefore, export from the ER through the coupled activities of these two systems provides an unanticipated degree of flexibility and specificity that is likely responsible for evolutionary diversity of the eukaryotic cell and the impact of PN regulated evolutionarily processes that define protein function in biology.

At a more practical level, learning from the diseases associated with COPII may be critical for understanding how COPII transport can be manipulated to alleviate diseases states associated with misfolding and trafficking. Given the dynamic signaling pathways that impinge on the PN to protect us from stress and aging, it is apparent that the PN-COPII system offers unanticipated levels of novel interactions that may potentially be used to restore function to a wide range of human inherited diseases (Balch et al. 2008, Prahlad and Morimoto 2009, Gidalevitz et al. 2010, Hutt and Balch 2010). Funding for this research supported by the National Institutes of Health grants GM42336, GM33301, NS067643 and DK51870 to WEB, and by the generous support of Cystic Fibrosis Foundation Postdoctoral Research Fellowships to KER and VG.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.