Is the mitochondria part of the endomembrane system

2.1.2 Membrane-Bound Organelles in Eukaryotic Cells

The endomembrane system separates the cell into different compartments, or organelles, such as the nucleus, the endoplasmic reticulum (ER), the Golgi apparatus, and lysosomes (see Table 2.2). The endomembrane system is derived from the ER and flows to the Golgi apparatus, from which lysosomes bud. The ER is a continuous system of flattened membrane sacks and tubules that is specialized for protein processing and lipid biosynthesis. The endomembrane system is important for the cell’s compartmental organization to function independently and properly. Ribosomes that synthesize proteins destined for insertion into cellular membranes or for export from the cell associate with specialized regions of the ER, called the rough ER owing to the attached ribosomes.

Table 2.2. Most Common Characteristics of Membrane-Bound Organelles in Animal Eukaryotic Cells

For the most common organelles in eukaryotic cells, the structure and function of each are illustrated, using an animal cell as an example, in Figure 2.1 and Table 2.2.

3 Membrane Trafficking

The endomembrane system permits various functions of the eukaryotic cell to be compartmentalized (e.g., protein degradation occurs in the lysosome), allowing a higher degree of cell specialization. The system relies on dynamic interactions between different compartments, facilitated by vesicle trafficking between them. In general, intracellular membrane trafficking involves the formation and budding of membrane vesicles from a donor membrane, their transport and subsequent fusion with target membranes, leading to transport of cargo from the donor to the target organelle (Rothman, 2002). These events are orchestrated and regulated by several proteins and protein complexes, including adaptor and coat proteins, small GTP-binding proteins (also called GTPases), as well as SNARE and tethering proteins (Fig. 3A).

Vesicle formation starts by recruitment and assembly of cytosolic coat proteins onto the membrane, leading to membrane budding and recruitment of cargo into the forming vesicle, which is eventually pinched off and transported to the donor membrane (Bonifacino and Lippincott-Schwartz, 2003; Kirchhausen, 2000). Several protein coat complexes have been identified, including clathrin-coats, COP-I, and COP-II. Clathrin is recruited to membranes through its binding to APs, including AP-2 which mediates vesicle formation from the plasma membrane (PM) through clathrin-mediated endocytosis and AP-1-mediating vesicle transport from the trans-Golgi network (TGN) (Schmid, 1997). Uptake of cargo from the PM can also occur independently of clathrin, involving other coats which facilitate micropinocytosis, phagocytosis, and other small-scale endocytic processes (Mayor et al., 2014). The nonclathrin coats COP-I and COP-II drive vesicle formation and transport from intra-Golgi/Golgi-to-ER and ER-to-ERGIC/Golgi, respectively (Barlowe et al., 1994; Letourneur et al., 1994).

The large superfamily of GTPases includes the RAS, Rho, ADP-ribosylation factor (ARF), and RAN and RRAB families that all function as molecular switches between an active state upon GTP-binding or inactive state when bound to GDP. Moreover, they have intrinsic GTPase activity leading to hydrolysis of their bound GTP to GDP and phosphate (Stenmark, 2009; Takai et al., 2001). The largest group of GTPases is the RAB proteins (Lamb et al., 2016a). RAB proteins localize to specific intracellular compartments where they regulate several steps of membrane trafficking, including vesicle formation, vesicle transport along actin and tubulin filaments as well as membrane tethering and fusion. The RAB proteins are prenylated and thus membrane anchored, giving them specific roles in trafficking of membrane vesicles. The activity of RAB proteins is tightly regulated by guanine nucleotide exchange factors (GEFs) and GTPase-activating proteins (GAPs). GEFs catalyze the dissociation of GDP allowing GTP to bind. The GTP-bound RAB protein is in its active state and can further bind effector proteins, e.g., a membrane-bound tethering protein (Fig. 3B). GAPs catalyze hydrolysis of the bound GTP to GDP resulting in the inactivation of the RAB protein and its displacement from the membrane. This occurs through the binding of guanosine nucleotide dissociation inhibitors (GDI) proteins, which bind the prenyl groups of RAB GDP and delivers that RAB to its original membrane (Cherfils and Zeghouf, 2013).

For a membrane vesicle to be able to fuse with other membrane vesicles or compartments, vesicle tethering proteins, and soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNAREs) are involved. These are proteins located on both the vesicle membranes (v-SNAREs) and the target membranes (t-SNAREs). SNARE proteins share a similar SNARE motif, and depending on whether or not this motif contains a conserved glutamate (Q) or an arginine (R) residue, SNAREs can also be classified into Q-SNARES or R-SNARES. t-SNAREs belong to the group of Q-SNAREs, and include syntaxins (STX) and synaptosomal-associated protein 25 (SNAP-25) proteins, whereas V-SNAREs belong to R-SNAREs and include the VAMP family of proteins (vesicle-associated membrane protein, also called synaptobrevin) (Ungermann and Langosch, 2005). Upon fusion, one R-SNARE and three Q-SNAREs (Qa, Qb, and Qc) interact with each other forming a trans-SNARE complex (Fasshauer et al., 1998). The subsequent formation of a α-helical bundle allows the two membranes to come in close proximity, leading to the displacement of water molecules and membrane fusion (Fig. 3C). The energy required for the membrane fusion is produced by the formation of the SNARE bundle (Hong and Lev, 2014; Südhof and Rothman, 2009). After fusion, the AAA-ATPase NSF (N-ethylmaleimide-sensitive factor) and its cofactor SNAP (soluble NSF attachment protein) allow the remaining SNARE complex, also called a cis-SNARE complex, to disassemble and be ready for another round of vesicle fusion (Jahn et al., 2003). An overview of the membrane trafficking components implicated in the biogenesis, transport, and maturation of autophagosomes is shown in Tables 1 and 2 and these components are discussed in more detail below.

Table 1. Trafficking and Membrane Remodeling Proteins in Autophagosome Formation

Table 2. Trafficking and Membrane Remodeling Proteins in Autophagosome Maturation

THE ENDOMEMBRANE SYSTEM

Fig. 1 shows the endomembrane system in the process of synthesizing a glycoprotein destined to become part of the plasma membrane; a similar scheme can be constructed for the synthesis of a secreted glycoprotein. There is considerable evidence that the peptide backbones of all glycoproteins are assembled on membrane-bound ribosomes (15,16); this has been most convincingly demonstrated by more recent work on enveloped viruses (such as vesicular stomatitis virus and Sindbis virus), which showed that messenger RNA coding for membrane glycoproteins was translated predominantly on membrane-bound ribosomes (17–19). In contrast, it has been suggested that nonglycosylated membrane proteins, destined for the cytoplasmic face of the endoplasmic reticulum or plasma membrane, are synthesized on free ribosomes and do not pass through the endomembrane assembly line; rather, these molecules possibly migrate through the cytosol (see 20 for a recent review).

The membrane glycoproteins of the endoplasmic reticulum present a special biosynthetic problem since glycosylation is completed in the Golgi complex and a mechanism is believed to exist for the transport of these molecules from the Golgi apparatus through the cytosol to the endoplasmic reticulum (21–24).

The “signal” causes attachment of ribosome to endoplasmic reticulum membrane. (2) The ribosome becomes attached to endoplasmic reticulum membrane by binding to a special protein in the membrane; this protein may serve to form a channel for passage of nascent peptide through the membrane. (3) Translation of messenger RNA occurs and nascent peptide undergoes “vectorial discharge” into the lumen of the rough endoplasmic reticulum. (4) Glycoproteins destined for secretion pass completely into the lumen and may remain only loosely bound to membrane; glycoproteins destined for the plasma membrane probably contain a hydrophobic region which keeps them bound tightly to membrane. Presumably when this hydrophobic region is translated, the ribosome lifts off the membrane and the remainder of the nascent peptide is released into the cytoplasmic side of the endoplasmic reticulum membrane. (5) The ribosome has lifted off the membrane and translation continues. Some proteins may undergo processing at this stage by a proteinase which removes the “signal” sequence from the amino-terminal end (26). Also, as discussed in the text, some carbohydrate incorporation may occur while the peptide is still nascent on the ribosome. (6) Translation is completed and peptide detached from the ribosome. Incorporation of oligosaccharide into peptide from dolichyl oligosaccharide pyrophosphate probably occurs predominantly at this stage. The multiglycosyltransferase systems catalysing carbohydrate incorporation are directed towards the intravesicular space. (7) The glycoprotein migrates towards the Golgi apparatus where elongation of the core to N-acetyllactosamine-type oligosaccharide occurs. It is now believed that a second type of processing occurs either in the rough endoplasmic reticulum, smooth endoplasmic reticulum, or Golgi apparatus, such that a large protein-bound oligosaccharide is cleaved to a smaller unit; this small oligosaccharide serves as the starting point for elongation (see text). (8) Vesicles migrate from the Golgi apparatus to the plasma membrane where fusion occurs. (9) Secretory proteins are released from the cell; lateral migration causes insertion of membrane glycoprotein into the plasma membrane.

The mechanism for the segregation of messenger RNA between free and membrane-bound ribosomes has long been a topic of heated speculation. A “signal hypothesis” has been proposed (25–27), which suggests that translation products destined for transfer across the endoplasmic reticulum membrane carry a “signal” to initiate binding of ribosome to the endoplasmic reticulum (Fig. 1). The “signal hypothesis” envisaged by Blobel and Dobberstein (26) applied to proteins destined for secretion from the cell and suggests that the signal is a sequence of hydrophobic amino acids near the amino-terminal end of the nascent polypeptide chain (Fig. 1); this signal sequence is not present in the final secreted product and is cleaved off within the endomembrane system by a specific protease. Presumably, proteins lacking such a signal sequence cannot initiate ribosome binding to membrane, are translated on free ribosomes, and do not enter the endomembrane assembly line.

There is only limited evidence to support the “signal hypothesis”. Nor is it certain that the mechanism just outlined applies to all secreted proteins or to all glycoproteins. However, a recent report by Wirth et al. (18) strongly suggests that the signal hypothesis applies to at least two viral membrane glycoproteins. Sindbis virus has three structural proteins, i.e., two envelope glycoproteins (E1 and E2) and a nonglycosylated core protein. A single polycistronic messenger RNA codes for these three proteins; there is only a single initiation site. When chicken embryo fibroblasts are infected with Sindbis virus, messenger RNA is found mainly on membrane-bound ribosomes; however, newly synthesized core protein is localized on the cytoplasmic side of endoplasmic reticulum membranes, whereas newly synthesized envelope glycoproteins are sequestered within membrane vesicles. Thus, nascent peptide precursor of envelope glycoprotein must have a signal that directs the binding of ribosomes to endoplasmic reticulum membranes. It is important to point out that this signal may not be an extra piece, which is subsequently cleaved off, but may be part of the final protein product; further it need not be a hydrophobic amino acid sequence, but may be due to some special, three-dimensional folded structure on the nascent peptide.

Once the ribosome is bound to endoplasmic reticulum, nascent peptide enters the intra-vesicular space, a process often termed “vectorial discharge” (29). A small amount of carbohydrate may be incorporated into the nascent peptide attached to ribosomes (30–34), but most glycosylation reactions probably occur after release of the peptide from the ribosomes (15,16,35). For oligosaccharides of the Asn-GlcNAc linkage type, only the core sugars (D-mannose and N-acetyl-D-glucosamine) are incorporated in the rough endoplasmic reticulum (15,16). This is probably achieved by transfer of oligosaccharide to peptide from dolichyl oligosaccharide pyrophosphate (see Spiro, in this volume). Recent evidence has suggested that the precursor of both oligomannoside and N-acetyllactosamine-type oligosaccharides (Scheme 1) is a large, protein-bound oligosaccharide, which is subsequently “processed” to a smaller size (see articles by S. Kornfeld and by P. W. Robbins in this volume). This large oligosaccharide is transferred from dolichyl oligosaccharide pyrophosphate and consists of an oligomannoside (and possibly some glucosyl residues) attached to the Man3GlcNAc2-Asn core structure.

Since “oligosaccharide processing” has been observed during the synthesis of vesicular stomatitis virus glycoprotein G (which has only oligosaccharides of the N-acetyllactosamine-type), it has been concluded that the precursor for the N-acetyllactosamine structure (Man3GlcNAc2-Asn protein) is derived by removal of mannosyl (and possibly glucosyl) residues from a large, protein-bound oligosaccharide (Kornfeld, Robbins, this volume). Evidence in support of this concept has been obtained from a study of a human genetic disorder, α1-antitrypsin deficiency (36). Hepatocytes from individuals with homozygous α1-antitrypsin deficiency (phenotype Pi ZZ) accumulate inclusion bodies within the rough endoplasmic reticulum (37–40). These inclusion bodies contain α1-antitrypsin which has not been secreted by the liver. Hercz et al. (36) isolated this hepatic α1-antitrypsin and showed that it contained only mannose and N-acetylglucosamine; galactose and sialic acid were absent from this material although these sugars are present in normal serum α1-antitrypsin. Further, whereas normal serum protein appears to contain four oligosaccharides with three mannose residues each, the Pi ZZ liver protein appears to carry only three oligosaccharides with an average of seven mannose residues on each oligosaccharide. Thus, these findings suggest that the precursors of the N-acetyllactosamine-type oligosaccharides of secreted α1-antitrypsin are larger, mannose-containing oligosaccharides.

One can speculate on two possible mechanisms for the defect in α1-antitrypsin deficiency. It is possible that an abnormality in the amino acid sequence of Pi ZZ α1-antitrypsin prevents normal processing within the rough endoplasmic reticulum, and that this in turn prevents movement from the rough endoplasmic reticulum to the Golgi apparatus. Alternatively, processing may occur within the smooth endoplasmic reticulum or Golgi apparatus and the defect in the disease may be a block in transport of α1-antitrypsin out of the rough endoplasmic reticulum.

The rough endoplasmic reticulum may therefore be involved in two types of processing, i.e., cleavage of “signal” sequences from the polypeptide backbone and cleavage of mannose (and possibly glucose) residues from oligosaccharide side-chains. The mechanism by which the glycoprotein is transported out of the rough endoplasmic reticulum to the Golgi apparatus (Fig. 1) is not known; as suggested above, “processing” may be an essential component of this transport. In the Golgi apparatus, the final stages of glycoprotein assembly are completed; this process has been termed “elongation” and is discussed in detail in the following sections.

The completed glycoprotein is then either secreted out of the cell or becomes part of the plasma membrane (Fig. 1). The transport vehicles for both these products are probably secretory vesicles. The membrane proteins are probably distinguished from the secreted proteins by having an intimate hydrophobic interaction with the membrane (Fig. 1). The model shown in Fig. 1 is based on the structure of glycophorin (41); this membrane protein has its amino-terminal end on the outside of the red cell. It is now clear that some intrinsic membrane glycoproteins may have their amino-terminal ends on the inside of the cell and this implies that the amino-terminal end may, after vectorial discharge into the lumen of the rough endoplasmic reticulum, snake back through the membrane to the cytoplasmic side of the endoplasmic reticulum. The carbohydrate of plasma membrane-bound glycoproteins appears always to be localized on the outside of the cell, suggesting that the glycosylation reactions occur within the lumen of the endomembrane system as depicted in Fig. 1.

Compartmentalization Ensures Fidelity and Directionality of Transport

The principal compartments of the endomembrane system are the nuclear envelope, the endoplasmic reticulum (ER), the Golgi apparatus, endosomes, and lysosomes (Figure 1). Proteins destined to different compartments of the endomembrane system and those destined for secretion (approximately one-third of the proteins encoded by the human genome) are translocated into the ER during their synthesis on ER-attached ribosomes. The ER contains the machinery for the proper folding, assembly, and, in some cases, modification of the proteins to enhance their stability and render them suitable for further modification as they pass through the next compartment (i.e., the Golgi apparatus) along the secretory pathway. Incorrect assembly can lead to protein degradation, while correctly folded and modified proteins are selected and organized into domains that exit the ER in the membrane-bound carriers. The process involved in the formation of the carriers is mediated by a core set of evolutionarily conserved molecules called the COPII coat machinery that selects cargo molecules to be transported and bends the membrane to form small round or large pleomorphic membrane carriers. The carriers that leave the ER are then destined for the next compartment of the pathway, the Golgi apparatus. The membrane carriers fuse with the membrane of the Golgi and the proteins pass through a series of subcompartments (cisternae) where they may be sequentially modified to produce the mature functional form of the protein. Having traversed the Golgi, the proteins are then sorted for delivery to their final destination, such as lysosomes or the cell surface. This, so far incompletely understood, sorting process involves the interaction of sorting motifs on the cargo molecules with specific adaptor proteins and is also influenced by the lipid composition of the membranes. Here, for example, another vesicle coat, clathrin, mediates traffic from the trans-Golgi network (TGN) through the endocytic system. Efficient sorting may occur in endosomal compartments that also operate in the sorting of molecules that are endocytosed from the plasma membrane. Thus, the compartmental organization of the secretory pathway ensures fidelity and directionality and a stringent selection process so that the proteins are delivered to their target organelles/locations.

Abstract

Eukaryotic organisms are defined by their endomembrane system and various organelles. The membranes that define these organelles require complex protein sorting and molecular machines that selectively mediate the import of proteins from the cytosol to their functional location inside the organelle. The plastid possibly represents the most complex system of protein sorting, requiring many different translocons located in the three membranes found in this organelle. Despite having a small genome of its own, the vast majority of plastid-localized proteins is nuclear encoded and must be posttranslationally imported from the cytosol. These proteins are encoded as a larger molecular weight precursor that contains a special “zip code,” a targeting sequence specific to the intended final destination of a given protein. The “zip code” is located at the precursor N-terminus, appropriately called a transit peptide (TP). We aim to provide an overview of plastid trafficking with a focus on the mechanism and regulation of the general import pathway, which serves as a central import hub for thousands of proteins that function in the plastid. We extend comparative analysis of plant proteomes to develop a better understanding of the evolution of TPs and differential TP recognition. We also review alternate import pathways, including vesicle-mediated trafficking, dual targeting, and import of signal-anchored and tail-anchored proteins.

Accumulating proteins in the apoplast

Secretion is the default pathway of the plant endomembrane system and without addition of specific signals for sorting or retention, proteins that traffic through the endomembrane system will typically be secreted to the extracellular space. Most large recombinant proteins accumulate within the apoplast — the region between the plasma membrane and the cell wall — as diffusion through the cell wall matrix is size delimiting. However, recombinant protein strategies using plant cell cultures are often employed to recover the target protein in the culture medium, which decreases the complexity of the initial purification stream and minimizes exposure to vacuolar/intracellular proteinases. Strategies to direct and enhance recovery of secreted proteins in plant cell culture systems are discussed in the next section.

II Introduction

Eukaryotic cells are characterized by an elaborate endomembrane system encapsulated by the plasma membrane. Comprised of the endoplasmic reticulum, Golgi apparatus, endosomes and lysosomes, the endomembrane system, and the plasma membrane constitute a dynamic membrane network. Physical and functional compartmentalization of this network is achieved by the production and consumption of pools of compositionally distinct transport vesicles. Vesicles are produced from a donor compartment by membrane budding and fission and consumed at the acceptor compartment by membrane fusion. Owing to the development of assays that accurately report lipid and content mixing in membrane vesicles in the late 1970s and early 1980s (Struck, Hoekstra, & Pagano, 1981; Wilschut & Papahadjopoulos, 1979), reconstitution of membrane fusion using viral and SNARE protein have significantly advanced our understanding of how transport vesicles are consumed at the acceptor compartment (Jahn & Sudhof, 1999). In contrast, mechanisms underlying membrane budding and fission remain largely unclear due to the lack of sensitive assays that can distinguish membrane fission from physical shearing of and protein desorption from membranes.

Of the numerous transport pathways involved in generation of vesicular intermediates in cells, the fission machinery involved in coated vesicular transport has been relatively well characterized. Coat complexes perform the dual role of concentrating proteins and budding membranes in the process of generating nascent transport vesicles (Brodsky, Chen, Knuehl, Towler, & Wakeham, 2001; Traub, 2009). Clathrin-coated vesicles mediate transport between the plasma membrane, endosome and trans Golgi compartments. Release of coated vesicles requires the membrane necks of coated buds to undergo scission, a process requiring GTP hydrolysis, and is catalyzed by members of the dynamin superfamily of large GTPases (Praefcke & McMahon, 2004). Early indication of the involvement of dynamin in scission of clathrin-coated endocytic buds came from analysis of a temperature sensitive Drosophila mutant shibire that displayed rapid paralysis at nonpermissive temperatures (Koenig & Ikeda, 1983). Thin-section electron micrographs of synaptic termini of paralyzed flies revealed a dramatic depletion of synaptic vesicles and accumulation of coated pits with electron dense material at the necks of coated pits resembling a collar. The shibire locus was subsequently identified as the Drosophila homologue of dynamin (van der Bliek & Meyerowitz, 1991).

In an attempt to understand the mechanistic basis of membrane fission, several reconstitution efforts have since been directed toward understanding dynamin behavior on artificial membrane templates. Much like the viral and SNARE proteins in membrane fusion research, these studies constitute a framework to understand membrane fission mechanisms.

Transport System and Vesicle Pools

Directed and regulated transport between different compartments of the endomembrane system and organelles of a cell, but also endo- and exocytosis are operated through coated vesicles, which underlie a steady circle of budding and fusion reactions at endomembranes of organelles and the cellular plasma membrane. Vesicles are formed by strictly regulated self-assembly of proteins, budding, and finally scission from the membrane in a GTP-dependent manner. Budding is an energy consuming process that occurs on the surface of membranes when cytoplasmic coat protein (COP) complexes assemble on the membrane surface. Within eukaryotes COPI buds vesicles from the Golgi, COPII functions at the ER, and the clathrin/adaptin system is operating at plasma membranes and endosomes. Taken together, at present three main coats are known: (1) COPI, (2) COPII, and (3) clathrin/adaptin.

While all of them exhibit similar function and comparable ancestry, severe differences exist regarding structure and assembly to the target and of the load concentrated within the specialized vesicle [75–77]. COPI vesicles may be understood as “Golgi vesicles.” They shuttle within the Golgi and from the Golgi back to the ER. COPII vesicles on the other hand export proteins from the ER to their target. Third, clathrin-coated vesicles (CCVs) provide the late secretory pathway and the endocytic pathway at the plasma membrane of cells.

Generally, vesicle forming is an energy consuming process and participating proteins can be functionally divided into two subgroups: first, adaptor-proteins and second, cage forming, that is, “coating” proteins. Vesicle formation is tightly regulated by GTP-binding proteins and activated by a GTPase that is stimulated by guanine exchange factors. The protein complex is anchored to the membrane by an amphipathic alpha-helix, recruits coat-proteins to form a complex, which further interacts with recognition and targeting sequences of possible cargo structures. Thus, specific cargo can be concentrated within a vesicle.

Generally, formation of these vesicle-forming protein-complexes differs between the three mentioned coats and can be observed (1) at the endomembrane itself or (2) within the cytosol, which leads to subsequent recruitment to the endomembrane and following vesicle formation.

During COPI-guided vesicle formation adaptor and cage-forming complexes are associated as a single heptameric complex, which is recruited afterwards to the membrane. Briefly, an ARF family G protein and coatomer, a 550 kDa cytoplasmic complex of seven COPs: α-, β-, β’-, γ-, δ-, ɛ-, and ζ-COP is formed and is then recruited to the membrane as a solid complex. By exchanging GDP for GTP on ARF (by guanine nucleotide exchange factor; GEF), budding is initiated. The early vesicle recruits cargo through coatamer-ARF-GTP interaction and forms spherical cages, which are coated by COPI thereafter [78–80].

During vesicle formation by COPII and clathrin, the adaptor proteins are first bound to the membrane. Subsequently, cage complexes are polymerized to form the coated vesicle. The adaptor complexes consist of AP1–5, AP180, and the Golgi-localizing, γ-adaptin ear containing, ARF-binding (GGA) proteins for clathrin, and Sec23–24, for COPII [77]. In contrast, endocytosis does not require GTPases but is initiated by phoshoinisitol-dependent recruitment of AP2 adaptor complexes to the membrane [81].

Lately, vesicular transport originating from the mt has been identified. Small vesicles derived from mt (MDVs) carry outer mitochondrial membrane protein MAPL to peroxisomes and furthermore, subpopulations of MDV targeted to the endosome and Golgi have been identified as well [82–84]. Thus, besides direct contact between mt and other organelles to interact and to transfer ions, metabolites and proteins, vesicular transport facilitates directional transfer [85,86]. MDV are sized between 70–150 nm. The “budding” and scission is independent from mitochondrial fission protein DrpI. MDV directed to lysosomes and Golgi have been shown to be enriched with oxidized proteins, the purpose of vesicles delivering to peroxisomes is not understood so far [83,87]. Based on the stress induced to mitochondria, selective incorporation of cargo into MDV is distinguishable. ROS originating from metabolic stress by xanthine oxidase/xanthine induce genesis of MDVs carrying the outer membrane pore voltage-dependent anion channel (VDAC). Otherwise oxidative stress resulting from dysfunction of the respiratory chain has been shown to lead to MDVs carrying an oxidized Complex III subunit (core2) without VDAC enrichment [88]. However, attribution of cargo, aggregation, or oligomerization in MDVs seems to depend on the structure damaged or oxidized first and may affect each mitochondrial structure. The generation of MDVs destined for lysosomes requires the protein kinase PINK1 and the cytosolic ubiquitin E3 ligase Parkin [89]. More recent data suggest that MDVs may be a first-line defense mechanisms to mt enabling export of damaged proteins to avoid mitochondrial dysfunction or failure without activation of autophagy [84].

Evolution of the Golgi Apparatus

The secretory pathway is a central component of the eukaryotic endomembrane system and the Golgi apparatus plays an essential role in it. Nevertheless, and rather surprisingly, there are organisms that do not seem to have a recognizable Golgi apparatus and that led to the suggestion these organisms might represent the earliest branches of the eukaryotic tree before the evolution of a functional Golgi apparatus (Cavalier-Smith, 1987). Later phylogenetic studies indicated that this was not the case (reviewed in Mowbrey and Dacks, 2009) and suggested that the Golgi apparatus is a universal feature of the eukaryotic cells and probably the LCEA did have a functional Golgi apparatus.

The Golgi apparatus varies in shape from a ribbon-like Golgi in mammalian cells to scattered stacks in the plants or insect cells to unstacked fragmented Golgi cisterna in S.cerevisiae. This led to the question about the nature of Golgi apparatus present in the LCEA. Phylogenetic studies have shown that organisms that lacked Golgi stacks are embedded in clades that had stacked Golgi apparatus suggesting that the absence of a stack in certain organisms could be a secondary loss and there were at least eight such independent instances of Golgi unstacking during the eukaryotic evolutionary history (Mowbrey and Dacks, 2009). In addition, a recent phylogenomic study has shown that almost all the categories of Golgi proteins including SNAREs, Rabs, and matrix proteins are conserved across all the eukaryotes with secondary losses seen in some branches (Klute et al., 2011). Given the importance of matrix proteins in the stacking of the Golgi (Seemann et al., 2000; Shorter and Warren, 2002), this supports the idea that early eukaryotes probably had a stacked Golgi.

Apart from this conserved properties, evolutionary innovations have played a part in modifying and adapting the Golgi to the need of the organisms. A clear example is the evolution of ribbon-like organization in the vertebrates (Wei and Seemann, 2010). This structure is also present in sea urchin (Terasaki, 2000), suggesting that probably it is a deuterostome innovation. Cell-biological studies have pointed to a role for the Golgi ribbon in the efficiency of glycosylation (Puthenveedu et al., 2006; Xiang et al., 2013) and transition to deuterostomes is accompanied by huge increase in the diversity of glycosylation products made by the Golgi apparatus with a pronounced increase in sialylated products (Varki et al., 2009a). Thus, the formation of the ribbon-like Golgi apparatus may have accompanied or contributed to the evolution of deuterostomes.

Another innovation during the course of evolution was the development of the TGN. While the TGN is rudimentary in most organisms, it is a well-developed structure in the mammalian cells (probably also in other vertebrates). Moreover, in some organisms the TGN and endosomes are identical (e.g., in plants) (Richter et al., 2009) while the mammalian TGN is distinct from the endosomes, with extensive membrane traffic between them. The expansion of the TGN and development of the ribbon-like structure was also accompanied by expansion in the repertoire of membrane traffic regulators with the human genome encoding nearly 30 known Rab proteins (Diekmann et al., 2011). Whether the evolution of TGN also parallels an increase in the number of sorting pathways exiting the TGN and how this may have contributed to the evolution of mammals remain open questions.

Introduction

Protein function is just as much about where a protein is located as it is about the shape of that protein. Proteins that are synthesized on the surface of the RER travel through a series of membrane compartments known as the endomembrane system, which functions to modify and sort proteins to their correct destinations inside and outside the cell. The endomembrane system is made up of the RER, the Golgi complex, vacuoles, endosomes, lysosomes, and small transport vesicles.

Protein synthesis starts in the cytoplasm when an initiation complex is formed between a mature mRNA and a small ribosomal subunit. With the addition of the large ribosomal subunit and additional tRNAs, the process of translation begins. Proteins that are destined to enter the endomembrane system carry a signal peptide that acts like an address label to direct these mRNAs to the surface of the RER. As translation continues, the proteins move through a translocon or protein channel embedded in the RER membrane and enter the interior space or lumen of the RER. Soluble proteins are released into the lumen while membrane proteins become anchored in the RER membrane by their stop transfer sequences.

One of the main modifications that occur to proteins that travel through the endomembrane system is glycosylation. As proteins enter the lumen of the RER glycosylating enzymes catalyze the formation of a covalent bond between the side chain (R group) of the amino acid asparagine and a large oligosaccharide made up of the sugars glucose, mannose, and N-acetylglucosamine in a process known as core glycosylation. As these core-glycosylated proteins move from the RER to the Golgi complex, distinct populations of enzymes located in each of the compartments or cisternae catalyze changes to the original oligosaccharide. Mannosidases, located in the cis Golgi, remove some of the mannose from the oligosaccharide while GlcNAc transferases, located in the medial Golgi, add N-acetylglucosamine to it. Enzymes in the final compartment of the Golgi, the trans Golgi, complete the glycosylation of a protein with the addition of galactose, fucose, and sialic acid.

Sketch the layout of a generic eukaryotic cell illustrating the relative positions of the nucleus, RER, Golgi, endosomes, lysosomes, a vacuole, and the plasma membrane. Label your drawing.

Research/review the steps involved in translation of an mRNA.

Research/review the signal hypothesis.

Explain why newly synthesized proteins move through a translocon and not directly through the membrane of the RER.

Viruses can hijack a cell’s endomembrane system to transport viral proteins to the plasma membrane. Along the way, the viral proteins can be glycosylated just like a cellular protein. Provide an explanation why viral proteins can be glycosylated.