Glycoprotein import

Abstract

Complex plastids evolved by secondary endosymbiosis and are, in contrast to primary plastids, surrounded by 3 or 4 envelope membranes. Recently, we provided evidence that in diatoms proteins exist that get N-glycosylated during transport across the outermost membrane of the complex plastid. This gives rise to unique questions on the transport mechanisms of these bulky proteins, which get transported across up to 3 further membranes into the plastid stroma. Here we discuss our results in an evolutionary context and speculate about the existence of plastidal glycoproteins in other organisms with complex plastids.

The origin of so called complex plastids traces back to the engulfment of a photosynthetic active eukaryote by an eukaryotic host cell, a process known as secondary endosymbiosis.¹ In general, 2 major lineages are distinguished differing in the identity of the symbiont, which was either a red or green alga (Fig. 1A). Secondary endosymbiosis led to the formation of a diverse group of eukaryotic organisms, ranging from ecologically highly important algal groups to human pathogens such as Plasmodium falciparum and Toxoplasma gondii, which possess complex plastids surrounded by either 3 or 4 membranes. Comparable to primary plastids the majority of the plastid proteome is encoded in the host nucleus. Taken into account the complex architecture of these plastids, new strategies had to evolve to allow protein transport across multiple plastid membranes.^2,3

Figure 1.Evolution of complex plastids and models on glycoprotein transport into the plastid of P. tricornutum. (A) Secondary endosymbiosis, the engulfment and reduction of a red or green algal cell, led to the formation of complex plastids that are surrounded by 3 or 4 membranes. This process occurred twice in the green lineage, leading to the establishment of Chlorarachniophytes and Euglenophytes, and at least once, but potentially also several times, in the red lineage resulting in a diverse group of microalgae and non-photosynthetic organisms. (B) In the diatom P. tricornutum plastidal glycoproteins get N-glycosylated in the cER and are thereafter transported into the PPC or plastid stroma. Three models (I-III) on transport of stromal glycoproteins are shown. See text for details. cER, chloroplast ER; IMS, inter membrane space; mt, mitochondrium; nu, nucleus; pl, plastid; PPC, periplastidal compartment; SELMA, symbiont specific ERAD like machinery; TOC/TIC, translocon of the outer/inner chloroplast membrane.

Research, especially from the past 15 years, helped unscramble basic mechanisms on protein transport into different complex plastids, revealing most notably the recycling and modification of pre-existing transport machineries.^4,5 In all organisms harboring complex plastids nucleus-encoded plastid pre-proteins are characterized by an N-terminal bipartite targeting signal (BTS) consisting of a signal peptide followed by a transit peptide-like sequence. The signal peptide mediates cotranslational transport through the Sec61-complex into the endoplasmic reticulum (ER), which is in Stramenopiles, Haptophytes, and Cryptophytes equivalent with transport across the outermost plastid membrane as this membrane stands in continuum with the ER. The second part of the BTS, the transit peptide-like sequence, contains all the information for subsequent targeting steps, thereby defining the final location of the protein within the complex plastid. First, it facilitates transport across the second outermost membrane, which is supposed to be mediated by an ERAD (ER associated degradation) derived transport machinery in Stramenopiles, Haptophytes, Cryptophytes, and Apicomplexans.^6-10 According to this new context the translocation machinery of the second outermost plastid membrane was termed SELMA (symbiont-specific ERAD-like machinery).⁸ After translocation into the periplastidal compartment (PPC) proteins that stay in the PPC and stromal proteins are distinguished. Only stromal proteins are transported forward across the 2 innermost membranes, which seem to be homologous to the envelope of the primary plastid. Core components of the TOC/TIC import system were so far identified in Stramenopiles, Haptophytes, Cryptophytes, and Chlorarachniophytes.³ In Euglenophytes and peridinin-containing Dinoflagellates, which contain plastids surrounded by 3 membranes, details on protein import are less clear (for review see ref. 2).

Glycosylation of proteins occurs in the ER and Golgi and the majority of proteins passing the secretory route get equipped with glycosylation residues, which play an important role in protein folding, structure, and function. For a long time cellular organelles, i.e., plastids and mitochondria, were not thought to harbor glycoproteins as synthesis of the nucleus-encoded proteins classically occurs in the cytosol. Surprisingly, though, in 2005 it was shown that there exist glycoproteins in primary plastids, which pass the ER and use an alternative vesicle mediated import pathway.^11-14

In contrast to protein import into primary plastids the import into complex plastids naturally includes the transport through the ER/chloroplast ER, and we recently demonstrated that glycoproteins are transported into the complex plastid of the diatom Phaeodactylum tricornutum.¹⁵ These findings confirmed the existence of an N-glycosylation machinery in the chloroplast ER and most notably raised the question how translocation of these bulky N-glycosylated proteins, which face multiple membrane barriers, is managed. Being an ERAD derived system, the SELMA translocation machinery of the second outermost membrane should be able to transport glycoproteins. However, for TOC and TIC channels glycoprotein transport is so far unknown and studies on the pore diameter suggest that these machineries might not be able to translocate N-glycosylated proteins. Hence, different models, explaining stromal glycoprotein transport across the 3 innermost plastid membranes of Stramenopiles, were proposed (Fig. 1B): 1) transport is mediated by the known translocation machineries SELMA, TOC/TIC, and sugar moieties might be sterically less problematic than expected. In the future, it will be interesting to investigate by in vitro import experiments whether isolated primary plastids are able to import N-glycosylated proteins. 2) Transport across the 2 innermost plastid membranes might be managed by hitherto unknown translocation machineries. In silico as well as proteomic analyses on the 2 innermost plastid membranes might help identifying new translocation components in future. 3) Stromal glycoproteins reach their final destination using vesicle transport from the second to the third outermost plastid membrane. However, to date there are no indications for PPC localized components, which mediate vesicle transport.¹⁶ In the case that so far unknown transport machineries are involved in stromal glycoprotein transport an additional specific targeting signal would be needed. The glycosylation itself in combination with the BTS could represent a possible signal, but inhibition of N-glycosylation by tunicamycin treatment did not alter protein targeting. Nevertheless, it cannot be ruled out that the absence of the glycosylation motif then led to a transport using the known translocation machineries. In future, broader bioinformatic analyses might help to possibly identify further targeting motifs.

The ERAD translocation machinery is one of a few translocator systems known to be able to transport bulky glycoproteins, which are usually transported via vesicles. Hence, from an evolutionary perspective the recycling of the red algal ERAD translocation machinery and the establishment of SELMA is very interesting as the ability to transport glycoproteins into the PPC might have served as a selection criterion. Such an evolutionary pressure would also explain the identification of SELMA in complex plastids of potentially independently evolved organisms. SELMA was identified so far in Stramenopiles and Apicomplexans, which most likely have a common phylogenetic origin and in Cryptophytes and Haptophytes that probably trace back to a common ancestor as well but might be the result of an independent secondary endosymbiosis.^7,17,18 Predictions on N-glycosylation of nucleus-encoded plastidal proteins suggest that also in these organisms plastidal glycoproteins might exist. But of course this needs to be verified experimentally. In some Apicomplexans, for example in P. falciparum, N-glycosylation seems to be unconventional resulting in truncated N-glycans. Previous studies proposed the existence of negative selection mechanisms as N-glycans might interfere with apicoplast targeting.¹⁹ In any case, similar approaches as applied in P. tricornutum might help to get more information on this issue. Chlorarachniophytes harbor complex plastids surrounded by 4 membranes, but pertain to the green lineage (Fig. 1A). In silico analyses reveal that also in Bigelowiella natans nucleus-encoded plastidal proteins with high confidence predictions for N-glycosylation sites exist. Since also in the green lineage plastidal proteins are transported through the ER it is very likely that these predicted recognition sites are modified by the N-glycosylation machinery. Interestingly, and in contrast to most organisms of the red lineage, there is no evidence for the existence of a SELMA translocation machinery in the second outermost plastid membrane of Chlorarachniophytes and so far it is not known how proteins, no matter if N-glycosylated or not, are transported into the PPC.²⁰ In the future, it will be very exciting to elucidate protein transport into the PPC of Chlorarachniophytes and to see whether the ability to transport bulky glycoproteins represents an evolutionary premise.

Glossary

Abbreviations:

BTS

bipartite targeting signal

cER

chloroplast ER

ERAD

ER associated degradation

PPC

periplastidal compartment

SELMA

symbiont specific ERAD like machinery

TOC/TIC

translocon of the outer/inner chloroplast membrane

Peschke M, Moog D, Klingl A, Maier UG, Hempel F. Evidence for glycoprotein transport into complex plastids. Proc Natl Acad Sci U S A. 2013;110:10860–5. doi: 10.1073/pnas.1301945110.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.