Hidden keyholders – exploring metabolite transport across the outer chloroplast membrane

ABSTRACT

The outer envelope of chloroplasts, which originates from the integration of a free-living bacterium into eukaryotic cells via endosymbiosis, is a unique membrane involved in various cellular processes. It not only forms the boundary of the organelle, but is also the site of contact between the plastids and the cytosol. Therefore, it is involved in ensuring an appropriate flux of metabolites to and from the chloroplast. To this end, the outer envelope is equipped with channel proteins, including several β-barrel proteins, i.e. OEP21, OEP24, OEP37 and OEP40, which play different roles in the transport of small molecules and metabolites across the membrane. Interestingly, these β-barrel proteins bear striking similarities to bacterial porins found in the outer membrane of Gram-negative bacteria, suggesting a common evolutionary origin. In addition to the β-barrels, the outer envelope contains other proteins with channel-like properties, notably OEP16 and JASSY, which function in the transport of amino acids and oxophytodieonic acid, respectively. This review highlights the importance of chloroplast outer membrane protein function in metabolite exchange and cellular homoeostasis. Furthermore, it provides insights into the phylogenetic origin of β-barrel proteins and their conserved membrane insertion mechanism.

KEYWORDS:

• Chloroplast outer envelope
• β-barrel proteins
• metabolite transport
• membrane protein insertion

Introduction

The production of oxygen and its release into the atmosphere by unicellular organisms through photosynthesis played a crucial role in the evolution of vascular plants Lyons et al. (2014). Around one billion years ago, a photosynthesising cyanobacterial-like cell was engulfed by a phagocytic eukaryote, which had already acquired another organelle, mitochondria, through a similar process Gould et al. (2008). In addition to these endosymbiotic organelles, higher plants now contain a number of membrane-bound compartments, including the vacuole, peroxisomes and the nucleus. This compartmentalisation is a prerequisite for the evolution of multicellular organisms, allowing cells to separate and run metabolic pathways independently without interference between substrates and products. However, effective communication between compartments is necessary to maintain and adapt cellular metabolite homoeostasis.

As sessile organisms, vascular plants are regularly exposed to various biotic and abiotic stresses, including variations in temperature, water availability, nutrient levels, and light conditions Kleine et al. (2021; Schwenkert et al. 2022). Light serves as the primary or secondary initiator not only for photosynthesis but also for other metabolic pathways. Cells need to respond to light on multiple levels, including the circadian rhythm and both long- and short-term changes in light intensity and availability. Therefore, the chloroplast, as the primary site of photosynthesis producing sugars as end products, has a high demand for the exchange of metabolites with the surrounding cytosol. In addition to their role in photosynthesis, chloroplasts are also the unique site for the biosynthesis of fatty acids as well as and nine out of the twenty amino acids Rolland et al. (2012; Holzl and Dormann 2019). They are involved in the production of NADPH, ATP and purines, which are subsequently exported from the organelle.

Due to their bacterial ancestry, plastids are surrounded by two membranes, the inner and outer chloroplast envelopes (IE and OE) Gray et al. (1999; Barth et al. 2022), thereby adding an extra layer of complexity to transport processes in and out of the organelle Breuers et al. (2011; Pottosin and Shabala 2016). Both membranes contain proteins that are crucial for the shuttling of metabolites. While the IE contains a number of well-described substrate-specific transporters, the OE was long considered non-osmotically active and easily permeable. However, in the last decades a number of OE proteins (OEPs) with channel properties have been identified and characterised. This review provides an overview of the known OE metabolite transport proteins to date.

The OE – a unique biomembrane

The transformation of once free-living bacteria into cellular organelles brought along a significant change in the function of the former bacterial outer membranes (OM) Breuers et al. (2011). In prokaryotes the OM mainly served to protect the organisms against the environment and to facilitate the import of nutrients. In endosymbiotic organelles, however, the need to import proteins and exchange ions and metabolites has diversified the function of this biomembrane. The OM is no longer just a barrier, but much rather is also involved in communication with the host organism and other cellular compartments.

All chloroplast membranes, i.e. OE, IE and thylakoids, retain bacterial characteristics. Such characteristics are most evident in the thylakoids but are also found in the OE and IE. Like those of cyanobacteria, the IE and the thylakoids are largely composed of the galactolipids monogalactosyl diacylglycerol (MGDG) and digalactosyl diacylglycerol (DGDG) LaBrant et al. (2018), the synthesis of which occurs at the envelope membranes Douce (1974). While the OE also contains MGDG and DGDG, it additionally contains a considerable amount of phospholipids (approx. 48%), mostly phosphatidylcholine (PC) Block et al. (1983; Yu and Benning 2003). PC, which is asymmetrically distributed and found only in the outer leaflet of the OE, is considered to be a eukaryotic feature and may help to integrate newly acquired “eukaryotic” proteins into the membrane Dorne et al. (1985). In addition to the different lipid composition, the lipid-protein ratio is much higher in the OE (2.5–3.0) than in the IE (0.8–1.0) Block et al. (1983).

Proteins found in the OE include integral membrane proteins such as β-barrels, α-helical proteins, tail-anchored proteins and peripherally bound proteins Inoue (2015). Since β-barrels are found only in bacteria, chloroplasts and mitochondria, these proteins are considered to be a feature derived from the bacterial origin of the organelles Schleiff et al. (2003; Paschen et al. 2003). In contrast, α-helical or tail-anchored proteins are thought of as eukaryotic additions Barth et al. (2022). Nevertheless, all OE proteins are nuclear encoded, synthesised on cytosolic ribosomes and post-translationally targeted to the chloroplast, with each protein category requiring different targeting and membrane insertion mechanisms that must have adapted after the transition from a prokaryote to a eukaryotic organelle Fish et al. (2022).

β-barrel channels – a common feature among chloroplast, mitochondria and bacteria

The chloroplast OE of higher plants contains six β-barrel proteins known to date, Toc75, OEP80, OEP21, OEP24, OEP37 and OEP40, all named according to their molecular weight. While Toc75 is well known for its function in protein import and OEP80 is thought to be responsible for in the insertion of β-barrel proteins themselves into the membrane Baldwin et al. (2005; Day et al. 2014), OEP21, OEP24, OEP37 and OEP40 are involved in shuttling small molecules and metabolites across the membrane (Table 1 and Figure 1).

Figure 1. OEP16, OEP21, OEP23, OEP24, OEP37, OEP40, and JASSY are described as metabolite channels in the OE. Structural predictions were obtained by AlphaFold Jumper et al. (2021; Varadi et al. 2022). Experimental evidence for the topology are only provided for OEP21 and OEP24 (Gross et al. 2021; Gunsel et al. 2023). 6GP = glucose 6-phosphate, question marks represent yet unidentified metabolites.

Table 1. Overview of OE metabolite transporting proteins.

Protein name	(potential) Function	Locus	References	Structure
OEP24A OEP24B	Solute/ion channel Solute/ion channel	At1g45170 At5g42960	Pohlmeyer et al. (1998), Rohl et al. (1999)	β-barrel
OEP21A OEP21B	Solute/ion channel Solute/ion channel	At1g20816 At1g76405	Bolter et al. (1999), Hemmler et al. (2006); Gunsel et al. (2023)	β-barrel
OEP40	Glucose channel	At3g57990	Harsman et al. (2016)	β-barrel
OEP37	Solute/ion Channel	At2g43950	Goetze et al. (2006)	β-barrel
OEP16–1	Amino Acid Channel	At2g28900	Pohlmeyer et al. (1997), Steinkamp et al. (2000), Linke et al. (2004), Philippar et al. (2007), Pudelski et al. (2012), Zook et al. (2013)	α-helical
OEP16–2	Amino Acid Channel	At4g16160	Pohlmeyer et al. (1997), Steinkamp et al. (2000), Linke et al. (2004), Philippar et al. (2007), Pudelski et al. (2012), Zook et al. (2013)	α-helical
OEP16–4	Amino Acid Channel	At3g62880	Pohlmeyer et al. (1997), Steinkamp et al. (2000), Linke et al. (2004), Philippar et al. (2007), Pudelski et al. (2012), Zook et al. (2013)	α-helical
OEP23	Solute/ion Channel	At2g17695	Goetze et al. (2015)	Mixed
JASSY-1	OPDA Channel	At1g70480	Guan et al. (2019)	Mixed
JASSY-2	Unknown/OPDA Channel	At1g23560	Guan et al. (2019)	Mixed

Toc75 and OEP80 are clearly of cyanobacterial origin as they are members of the bacterial Omp85 β-barrel protein family Gentle et al. (2005). OEP21, OEP24, OEP37 and OEP40 also show significant structural similarities to bacterial porins found in the OM of Gram-negative bacteria Bolter and Soll (2001). While the bacterial OM serves as a defence against hostile environmental contacts (e.g. antibiotics or bile salts), it also forms a barrier to nutrients and other hydrophilic or hydrophobic small molecules Prajapati et al. (2021). The bacterial OM is therefore densely packed with proteins, the majority of which are β-barrel proteins that facilitate nutrient uptake and waste removal. Many of these β-barrel proteins are known as non-selective porins, which allow diffusion depending on the concentration gradient of the substrate. These porins typically form water-filled pores with large diameters (>10 Å) that allow the passage of solutes up to 600 Da and require little interaction with the porin Kojima and Nikaido (2013). Prominent examples are the outer membrane proteins (Omp)F, OmpC and PhoE, which are organised as heterotrimers, each forming an hourglass-shaped channel with the narrow region limiting the size of the solutes passing through Cowan et al. (1992; Kefala et al. 2010). In addition to these non-specific porins, bacterial membranes also contain substrate-specific β-barrel channels that rely on transient channel-substrate interactions Nikaido (2003). Typically, these channels are equipped with 2–4 flexible extracellular loops that fold into the channel, reducing its inner diameter. In contrast to non-specific channels, transport rates are saturated at high substrate concentrations. Depending on the substrate, the inner cavities of these channels are lined with different amino acids to allow the binding of substrates with different biochemical properties. For example, the pore axis of LamB is lined with aromatic amino acids to facilitate the passage of maltose molecules Schirmer et al. (1995; Dutzler et al. 1996).

Likewise, several β-barrel proteins are also found in the mitochondrial outer membrane. Among them are prominent proteins such as the translocase of the outer membrane (Tom)40, sorting and assembly machinery (Sam)50 and the voltage-dependent anion-selective channel (VDAC) Hill et al. (1998; Kozjak et al. 2003; Wiedemann et al. 2003; Mertins et al. 2014; Becker and Wagner 2018). Tom40 is part of the TOM complex that serves as an entry point for protein import, while Sam50 facilitates the integration of β-barrel proteins (Hill et al. 1998; Wiedemann et al. 2003). Like Toc75 and OEP80, Sam50 belongs to the Omp85 family of β-barrel proteins (Paschen et al. 2003). For metabolite transport, VADC is known as a highly abundant universal channel. VDAC plays a critical role in the transport of a wide range of compounds, ranging from inorganic ions such as K⁺, Na⁺ and Cl⁻ to metabolites of different sizes and charges (Colombini 1989; Mertins et al. 2014). The channel also facilitates the transport of large macromolecules up to 6 kDa, such as tRNAs (Salinas et al. 2006). Despite functional structural similarities to bacterial porins and the fact that VDAC and Tom40 share a common evolutionary origin, phylogenetic analysis revealed that these proteins have followed independent evolutionary pathways, generating paralogues in animals and plants, with common themes observed in their structures rather than high sequence identity (Bay et al. 2012). While Tom40, Sam50 and VDAC are conserved across all species, an additional plant-specific mitochondrial β-barrel protein of the bacterial porin III superfamily, the outer membrane protein of 47 kDa (AtOM47), has recently been described as a potential function in metabolite transport (Li et al. 2016).

Metabolite β-barrel channels in the OE of chloroplasts

In contrast to mitochondria, metabolite transport across the OE in chloroplasts appears to be more specialised, as indicated by the presence of four different typical β-barrel channels, OEP21, OEP24, OEP37 and OEP40.

OEP21 has long been shown to function as a solute-selective porin and to function as a major export site for primary photosynthetic products, such as triose phosphates. It has also been suggested to be regulated in an ATP-dependent manner (Bolter et al. 1999; Hemmler et al. 2006). While the transport of triose phosphates across the IE via the triosephosphate/phosphate translocator (TPT) is well understood at a mechanistic level, the molecular functioning of OEP21 has only recently been elucidated (Gunsel et al. 2023). An NMR structure revealed that the channel is cone-shaped, with its wider side orientated towards the intermembrane space and positively charged amino acids facing towards the inside of the cone. Fittingly, negatively charged molecules were shown to be preferred as substrates with a cut-off at >1 kDa. While this applies to triose phosphates, it may also allow the passage of other small metabolites of similar size and charge. Interestingly, OEP21 tends to form oligomers, which can lead to pore closure driven by conformational changes. The presence of ATP can stabilise the open state of the channel, supporting speculation that lower ATP levels caused by oxidative stress result in pore closure, thereby preventing triose phosphate depletion of the chloroplast (Gunsel et al. 2023).

Besides OEP21, OEP24 and OEP37 are two phylogenetically related β-barrel channels present in the chloroplast OE (Reddy et al. 2016). Interestingly, OEP24 and OEP37 are more abundant in the mesophyll chloroplasts of the C₄ plant maize, whereas OEP21 is more highly expressed in C₃ plants (Brautigam et al. 2008). This may reflect an adaptation of the fluxes of small metabolites specifically required for C₄ photosynthesis.

OEP24 appears to be less specific than OEP21 and has been proposed to be a functional homologue of VDAC, acting as a general solute channel (Pohlmeyer et al. 1998). Although there are no similarities at the sequence level, a yeast strain lacking VDAC could be fully complemented by the expression of OEP24, which is inserted into the outer mitochondrial membrane in this system (Rohl et al. 1999). In reconstituted proteoliposomes, OEP24 was able to function as a high conductance, slightly cation-selective ion channel allowing the passage of triosephosphate, dicarboxylic acids, positively or negatively charged amino acids, sugars, ATP and Pi (Pohlmeyer et al. 1998).

Likewise, OEP37 was shown to be a high conductance, cation-selective channel with an hourglass shape and, interestingly, was also able to partially complement VDAC in yeast (Goetze et al. 2006; Ulrich et al. 2012). However, OEP37 mutants do not show an obvious phenotype under normal growth conditions (Goetze et al. 2006). The similar properties of OEP37 and OEP24 might suggest an at least partially overlapping function. However, to date there is no information on OEP24 loss-of-function mutants, let alone OEP37 and OEP24 double mutants.

More recently, OEP40 has been described as a high conductance channel permeable to glucose and its phosphorylated derivatives, glucose 1-phosphate and glucose 6-phosphate (Harsman et al. 2016). These sugars are important metabolites that need to be exchanged between the chloroplast and the cytosol. In particular, glucose must be exported from the chloroplast during starch breakdown at night. The export of glucose and maltose across the IE during the night is facilitated by the plastid glucose transporter and MEX1, respectively (Weber et al. 2000). Interestingly, OEP40 was found to be impermeable to maltose in vitro. Further structural studies would be required to determine whether the specificity is mediated by a mechanism similar to that of the previously described bacterial LamB. Interestingly, trehalose-6-P is not transported by OEP40, but appears to be able to interact with the channel and may therefore be involved in its regulation. Trehalose-6-P plays a role in the induction of flowering, providing insight into the observation that loss of OEP40 leads to early flowering when plants are grown at 10°C (Harsman et al. 2016).

β-barrel membrane insertion mechanisms are evolutionary conserved

All mitochondrial and chloroplast β-barrel proteins are translated in the cytosol and must be targeted and inserted into the appropriate membrane. Strikingly, the insertion mechanism of β-barrel proteins into bacterial, mitochondrial and chloroplast outer membranes is facilitated by homologous systems, each involving a β-barrel protein belonging to the Omp85 protein family that drives the insertion process (Diederichs et al. 2021). Figure 2 provides a schematic comparison of the insertion mechanisms in bacteria, chloroplasts and mitochondria. In bacteria, the β-barrel assembly machine (BAM) assists in the folding and insertion of all β-barrel proteins, which arrive at the OM from the periplasmic side. The BAM complex consists of the central β-barrel protein BamA, which is associated with several auxiliary proteins that facilitate substrate interaction. In the last two decades, considerable insight has been gained into the structural details, but the mechanism is still not fully understood and two theoretical models are currently under discussion (Figure 2(a)). The first model proposes that chaperones assist partial or even complete folding of the β-barrel protein in the membrane and BamA assists insertion by locally destabilising the membrane. The second model, known as the BamA budding mechanism, involves the so-called β-signal of the new β-barrel protein, which are semi-degenerate C-terminal peptides motifs found in all prokaryotic and eukaryotic outer membrane β-barrel proteins. The first β-strand of BamA seems to act as a template for strand formation of the new β-barrel protein, resulting in a hybrid β-barrel intermediate. Each added strand facilitates the formation of the next strand until the new β-barrel is fully folded by continuous strand exchange. To prevent the formation of a super-pore, the new β-barrel is separated from BamA by the opening of a lateral gate. Experimental evidence, including the lateral opening of the BamA barrel and cross-linking studies, supports this mechanism (Wu et al. 2020; Diederichs et al. 2021).

Figure 2. Schematic overview of β-barrel insertion in membranes. (a) insertion into bacterial membranes is facilitated by the SAM complex. Two insertion mechanisms are proposed: 1) an intermediate pore of the Sam50 and the new β-barrel is formed by consecutive subsitution of β-sheets. 2) the new β-barrel is folded in the periplasm and BamA assists the insertion by destabilising the membrane. (b) β-barrel insertion in mitochondria requires the prior translocation via the TOM complex. β-barrels are bound by small TIM proteins in the IMS and inserted by the opening of a lateral gate through Sam50. (c) β-barrel insertion in chloroplasts requires translocation by Toc75. OEP80 is involved in membrane insertion, but mechanistic details are lacking. For clarity only the core proteins of the BAM, SAM, TOM and TOC complexes are shown.

Unlike bacteria, the outer membrane of mitochondria and chloroplasts faces the cytosol rather than the extracellular space. The protein targeting process is consequently more complex, as proteins are arriving at the membrane from the cytosolic side and must first cross their destination membrane (Figure 2(b)). In mitochondria, β-barrel insertion involves the TOM complex and the mitochondrial distribution and morphology (MDM) in addition to the sorting and assembly machinery (SAM), which serves as the mitochondrial counterpart of the bacterial BAM complex (Meisinger et al. 2007; Hohr et al. 2015, 2018). Typically, organellar proteins have evolved N-terminal targeting signals that are cleaved after translocation and are translocated by Tom40. However, mitochondrial β-barrels do not contain an N-terminal targeting signal, but rather a β-signal in the C-terminal hairpin, similar to that described for bacterial β-barrel proteins. As a first step, mitochondrial β-barrel proteins are translocated to the inner membrane space (IMS) by Tom40, where they are protected from aggregation by the chaperoning small TIM proteins (Hohr et al. 2018). The β-barrel protein Sam50 shares sequence homology with BamA, in contrast to the auxiliary proteins of the complexes (Diederichs et al. 2021). In a mechanism similar to the BAM budding hypothesis, Sam50 facilitates interaction with the β-signal and subsequent protein insertion. Sam50 also contains a β-signal in strand 16 that can be displaced by the substrate β-signal, resulting in lateral opening of the pore. Recent studies have shown that Sam50 can also interact with Mdm10, another β-barrel protein (Takeda et al. 2023). It has been suggested that the SAM complex can exist either as a dimer, in complex with a substrate or in complex with Mdm10. The SAM complex seems to undergo a reaction cycle in which β-barrel switching is required for the release of a fully folded β-barrel into the membrane by the SAM complex (Takeda et al. 2021). Mdm10, however, is absent from plants and it remains to be investigated whether plant mitochondria rely on a similar mechanism.

For chloroplast β-barrel insertion, mechanistic details have only recently begun to emerge. As mentioned above, the OE contains two proteins belonging to the Omp85 family – Toc75 and OEP80. It is hypothesised that these two proteins arose from a gene duplication event and that while Toc75 specialised in protein import, OEP80 retained its original function in β-barrel insertion (Day et al. 2014). Like mitochondrial proteins, most chloroplast proteins contain an N-terminal cleavable targeting sequence. In contrast to all other OM β-barrel proteins, this is also the case for Toc75 and OEP80, with Toc75 even containing an additional polyglycine stretch that is required for sorting (Inoue and Keegstra 2003). However, it remains unclear how much of Toc75 is translocated across membranes before being inserted into the OE. The transit peptide of OEP80 has been shown to be important for the import of the IMS-localised POTRA domains, but not for the insertion of the β-barrel itself (Day et al. 2014, 2019). For the remaining β-barrel OEPs, OEP21, 24 and 37, not much is known about their insertion mechanism. They do not contain a targeting signal, but the signal for translocation seems to be present in the 5–6 N-terminal β-strands. It was found that the most C-terminal strand is required for membrane insertion and presumably, the OEPs are first translocated across the OE membrane via Toc75 (Figure 2(c)) (Gross et al. 2021). Although these data suggest a mechanism similar to that found in bacteria and mitochondria, detailed structural analysis describing the insertion mechanism with accessory proteins and potentially involved IMS chaperones is still lacking.

OEP16 an α-helical eukaryotic addition to OE metabolite channels

In addition to β-barrel proteins, α-helical proteins with (potential) channel function are also found in the OE. The most prominent protein is OEP16, which consists of four hydrophobic α-helices (Figure 1). Based on its structural features, OEP16 can be considered a eukaryotic complement to the OEPs and is found in charophytes, bryophytes and all higher plants (Pudelski et al. 2012). It has been shown to assemble as homodimers and to form a high conductance channel that is slightly cation selective, with helix I and helix II responsible for pore formation (Steinkamp et al. 2000; Linke et al. 2004). NMR analysis has revealed breaks in each of the helices, which may contribute to substrate specificity. In addition, helix I is amphiphilic, with two charged amino acids facing the water-filled pore (Zook et al. 2013). OEP16 is a member of the preprotein and amino acid transporter (PRAT) superfamily, which also includes the inner mitochondrial membrane protein translocation channels Tim17, Tim22 and Tim23 (Rassow et al. 1999; Murcha et al. 2007, 2016). Four closely related genes are encoded in Arabidopsis: OEP16-1, OEP16-2, and OEP16-4 are targeted to the chloroplast, while OEP16-3 is located in the inner mitochondrial membrane. The latter has recently been shown to interact with Tim22, a component of the mitochondrial protein import machinery (Zhang et al. 2023).

Plastids are the primary site of amino acid production, and amino acid transport is essential not only for polypeptide production, but also for plant development and stress response (Batista-Silva et al. 2019). Interestingly, in vitro transport assays have shown that OEP16 is permeable to amino acids but not to phosphoglyceric acid and uncharged sugars (Pohlmeyer et al. 1997). While OEP16-1 is the most abundantly expressed homolog, OEP16-2 is mainly expressed during late seed development and early germination (Pudelski et al. 2012). Its expression is under the control of the phytohormone abscisic acid (ABA), and oep16-2 loss-of-function mutants show an ABA-hypersensitive germination phenotype. An amino acid imbalance was observed during this developmental stage, further suggesting a role for OEP16 in amino acid transport (Pudelski et al. 2012). The role of OEP16 in stress tolerance is emphasised by the fact that OEP16-2 was also found to be up-regulated under heat and drought, and expression of the wheat homolog TaOEP16-2-5B resulted in increased heat tolerance (Zang et al. 2017). In contrast, the expression of OEP16-1 in Arabidopsis and barley is under the control of the C-repeat binding factor (CBF) regulatory pathway and is up-regulated in cold (Fowler and Thomashow 2002; Thomashow 2010).

Another α-helical protein with a potential function in metabolite transport across the OE is ABCG7/WBC7, a protein of the ATP-binding cassette (ABC) protein superfamily that has been found in several OE proteomics studies (Simm et al. 2013; Bouchnak et al. 2019). These proteins, known as ABC transporters, are responsible for shuttling a wide range of substrates across different membranes (Sanchez-Fernandez et al. 2001). However, the localisation of this protein needs to be confirmed and its function remains to be elucidated.

The unusual OEPs – JASSY and OEP23

More recently, two other OE channel proteins, JASSY and OEP23, have been described that are conserved in land plants, bryophytes and green algae (Goetze et al. 2015; Guan et al. 2019). Based on their structural predictions, the proteins are neither classical β-barrel proteins nor do they resemble the structure of an α-helical transport protein (Figure 1). Instead, they are composed of a mixture of α-helices and β-sheets, and although electrophysiological experiments clearly show channel properties (Goetze et al. 2015; Guan et al. 2019), their structural basis for pore formation within the membrane remains elusive. JASSY belongs to the START/RHO_alpha_C/PITP/Bet_v1/CoxG/CalC (SRPBCC) superfamily, which contains a conserved ligand-binding domain. The ubiquitous domain consists of an incomplete β-barrel forming a large hydrophobic binding cavity capable of binding hydrophobic compounds such as fatty acids and derivatives (Radauer et al. 2008). Members of the SRPBCC superfamily typically have low sequence identity but share a similar three-dimensional structure. Interestingly, structural prediction analysis of OEP23 using Phyre2 models OEP23 on structures of proteins from the Bet v1-like superfamily with 89.9% confidence, thereby suggesting the presence of a Bet v1-like/SRPBCC domain as found in JASSY (Kelley et al. 2015). Despite its unusual predicted structure, JASSY has been shown to have channel-like properties and to function in the export of oxylipin-12-oxophytodieonic acid (OPDA), the precursor of the plant hormone jasmonate (JA) (Guan et al. 2019). OPDA is derived from α-linolenic acid from chloroplast membranes and must therefore be exported. JA plays an important role in various cellular responses, including the response to biotic and abiotic stresses and the development of reproductive organs. Its active form, JA-Ile, activates various transcription factors in the nucleus (Wasternack and Song 2017). The absence of JASSY in Arabidopsis leads to a deficiency in JA accumulation, which results in impaired expression of JA target genes upon exposure to various stresses. This defect not only renders plants more susceptible to pathogen attack, but also leads to impaired cold acclimation (Guan et al. 2019).

OEP23 was originally identified in a proteomic approach using isolated pea OE membranes (Gutierrez-Carbonell et al. 2014; Goetze et al. 2015; Guan et al. 2019). The protein forms a high conductance cation-selective ion channel that appears to represent a single channel unit rather than forming multiple pore complexes (Goetze et al. 2015). However, no specific substrates of OEP23 have been identified. Interestingly, in a separate proteomic approach on cold and non-cold treated Arabidopsis envelope membranes, it was found to be strongly down-regulated after cold treatment (Trentmann et al. 2020). This suggests that the substrate(s) shuttled by OEP23 and a change in its cellular distribution have a major impact on cold acclimation. However, the analysis of loss-of-function mutants is required to further investigate the role of OEP23.

Conclusion

The OE is a striking example of how a biological membrane has evolved to enable metabolic homoeostasis and communication within a eukaryotic cell. Metabolite transport processes across the membrane are not only facilitated by β-barrel proteins as found in bacteria, but eukaryotic features have been added to allow more specific passages. Although it has become clear over the past decades that chloroplast OEPs provide a degree of selectivity and thus contribute to the ability of plants to adapt to changing growth conditions, further work is needed to elucidate the molecular basis of the OEP substrate specificity and their in vivo roles. Structural studies are particularly important to elucidate the precise mechanisms by which these proteins function and how they interact with other membrane components. Overall, the study of OEPs promises to improve our understanding of plant metabolism and stress adaptation. The knowledge gained from these studies can contribute to the development of strategies to enhance plant growth, productivity, and stress tolerance, with potential applications in agriculture and biotechnology.

Acknowledgment

We would like to thank Jürgen Soll for helpful discussions.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data sharing is not applicable to this article as no new data were created or analysed in this study.

Additional information

Funding

This work was supported by the DFG under Grant CRC TR175, B06 to SS and B07 and C05 to DL.