Many biotechnological and biomedical applications rely on the effective co-expression of multiple proteins. So far, multiple genes have been expressed via: (i) monocistronic vectors e.g. viral co-infection or co-transfection of plasmids expressing one protein each; fusion proteins; fusion proteins incorporating proteinase cleavage sites and (ii) polycistronic vectors where multiple genes are assembled either under the control of multiple promoters or a single promoter. IRES sequences have been used as a method to separate two coding sequences under the control of a single promoter. Despite their widespread use, they are relatively large (~600 base pairs) and expression of the downstream gene is significantly less efficient than the upstream gene [1]. A different approach using the 2A oligopeptide sequence allows multiple discrete proteins to be synthesized from a single strand of RNA, which also functions as a messenger RNA (mRNA). Not only is the 2A sequence smaller (54-174bp) than IRES elements, co-expression of proteins linked via 2A is independent of the cell type (cleavage activity is only dependent on eukaryotic ribosomes) and proteins are produced with equimolar stoichiometry [2-4]. The potential of this system is vividly demonstrated in the yeast Pichia pastoris by the functional co-expression of the carotenoid and violacein biosynthetic pathways encoded by nine genes from a polycistronic construct based on 2A peptides: the highest number of genes expressed in such a co-ordinated fashion so far [5]. IRES and 2A sequences have been used successfully in an impressive array of studies: at least 200 in the case of IRES elements and almost 900 in the case of 2A peptides making 2A the “go-to” option for protein coexpression [6-9]. Here we provide a short update on the history of 2Aand cover some applications of 2A co-expression technology.

The Foot-and-mouth disease virus (FMDV) 2A sequence (hereafter “F2A”) mediates “self-processing” by a novel translational effect variously referred to as ‘ribosome skipping’, ‘stop-go’ and ‘stop-carry on’ translation [10-12]. 2A peptide cleavage has been studied in various cell types using various recombinant polyproteins and artificial reporter polyprotein systems comprising chloramphenicol acetyltransferase (CAT), β-glucuronidase (GUS), and fluorescent proteins (FPs e.g. GFP, RFP, YFP) [13-16]. It was demonstrated that F2A plus the N-terminal proline of the 2B downstream protein co-translationally ‘self-cleaved’ at the glycyl-prolyl pair site corresponding to the 2A/2B junction (LLNFDLLKLAGDVESNPG↓P-) (Figure 1) [17-25]. Later work shows the length of the F2A used is also important for cleavage in vitro and in vivo - higher cleavage efficiency was reported when longer versions of 2A with N-terminal extensions derived from FMDV capsid protein1D upstream of 2A were used [18,20,26-33]. Importantly, the co- translational nature of this cleavage means that by including varioussignal sequences within 2A polyproteins, either up- or downstream of2A, proteins can be either targeted co-translationally to the exocyticpathway or post-translationally to different cellular compartments[34-36]. Of the many 2A peptides identified to date [10,27,37], fourhave been widely used in biotechnology and biomedicine: F2A, equinerhinitis A virus (“E2A”), porcine teschovirus-1 (“P2A”) and Thoseaasigna virus (“T2A”) (Table 1) [2-4,36,38-43].

Table 1: The resume of changes in ectonucleotidase activities in response to acute and chronic moderate exercise. é: increase; ê: decrease; NC: no change'

Figure 1: Schematic presentation of co-translational cleavage via “selfcleaving” 2A peptide. Gene sequences 1 (stop codon removed) and 2 are linked together into a single (trans) gene via a 2A sequence. The translation products are synthesized in an equimolar ratio, although, protein 1 upstream of 2A bears a C-terminal extension of 2A, and protein 2 bears an N-terminal proline residue.

Finally, it should be noted that (i) 2A remains as a C-terminal extension of the upstream gene, and (ii) proline forms the N terminus of the downstream gene. The presence of N-terminal proline does not seem to affect proteins which are metabolically stable [44], but when the authentic C-terminus is required for activity or subcellular targeting of certain proteins, they should either be encoded at the C-terminus of the polyprotein or followed by cleavage sequences of the mammalian Kex2p homologue, furin (-↓RRRR-, -↓RKRR-, -↓RRKR-), which removes the 2A “tag” [21,38]. The presence of 2A, however, can be used for detection of protein expression and localization using anti-2A antibodies [2,3].

Biotechnological and Biomedical Applications

F2A is the most widely used 2A sequence in plant biotechnology and has been used to target multiple proteins to various subcellular compartments [13,34,45,46], to improve disease resistance [47,48], drought-resistance [49] and nutritional value through metabolome engineering [10,50]. Vitamin A deficiency (VAD) is a major global health issue which affects hundreds of millions of people. This problem arises because rice, the staple food source in countries where VAD is prevalent, does not produce vitamin A or its precursor β-carotene, which have a number of vital functions in the body including growth. Since 2000, researchers have been engineering a transgenic variety of rice referred to as “golden rice” (Oryza sativa, GR) that includes the biosynthetic pathway for production of β-carotene [51,52]. Engineering the pathway into (carotenoid-free) rice endosperm requires two carotenoid biosynthetic genes, phytoene synthase (psy) and carotene desaturase (crtl) [53]. To avoid the problems of promoter interference (GR1 and 2 require two promoters), psy from Capsicum and crtl from Pantoea, were linked via 2A (psy-F2A-crtl; pPAC construct) or IRES (psy-IRES-crtl; pPIC construct) and placed under the control of the rice globulin promoter [50]. GR3 of transgenic PAC had a much more intense golden colour than did the PIC transformants, demonstrating that the 2A construct performed better than the IRES construct interms of carotenoid production.

F2A and ‘2A-like’ sequences have been used extensively in genetic engineering of T cells for adoptive cell therapies [39-41], human stem cells [42,54,55] and in the induction of pluripotent stem cells [23,25,43]. For example, the genes encoding four transmembrane proteins necessary for the assembly of the CD3 complex were coexpressed from a polycistronic vector containing three 2As [36] and coexpression of four transcription factors in a [KLF4-E2A-OCT3/4-T2ASOX2- P2A-c-Myc]-IRES-hrGFP construct resulted in the generation of induced pluripotent stem (iPS) cells from somatic cells [56]. High levels of functional monoclonal antibodies were produced by linking the antibody heavy and light chain sequences with F2A in the context of AAV-mediated gene transfer [21]. 2As were used to co-express tumor associated antigens (TCR) α- and β-chains to treat metastatic melanoma, colorectal cancer, renal cell carcinomas and many other types of malignant diseases [57-59]. Significant anti-tumour responses were observed in the clinic using monoclonal antibodies by coexpressing cytotoxic T-lymphocyte-associated antigen (CTLA-4) heavyand light chains [60].

Finally, we think ‘2A-like’ sequences are able to function both as a signal sequence and as a translational recoding element - this leads to partitioning of the translation products between two subcellular sites (dual protein targeting). We have identified some 2A-like sequences at the N-terminus of NLRs in the genome of the purple sea urchin Strongylocentrotus purpuratus that were putative signal sequences. Constructs encoding wild-type [Sp2A-cherryFPT2A- GFP] or a mutated, cleavage inactive form of Sp2A were used to transfect mammalian HeLa cells – with both constructs GFP wasevenly distributed throughout the cell, wild-type Sp2A lead to cherryFP localization throughout the cell and mutated Sp2A acting as a signallead to cherryFP localization in the exocytic pathway [37,61].

In conclusion, with the number of studies that have successfully used F2A and ‘2A-like’ sequences approaching 1000, these small peptides are proving to be the ‘go-to’ technology for co-expression ofmultiple proteins.