Three Decades of Advances in Arabinogalactan-Protein Biosynthesis

Abstract

Arabinogalactan-proteins (AGPs) are a large, complex, and highly diverse class of heavily glycosylated proteins that belong to the family of cell wall hydroxyproline-rich glycoproteins. Approximately 90% of the molecules consist of arabinogalactan polysaccharides, which are composed of arabinose and galactose as major sugars and minor sugars such as glucuronic acid, fucose, and rhamnose. About half of the AGP family members contain a glycosylphosphatidylinositol (GPI) lipid anchor, which allows for an association with the outer leaflet of the plasma membrane. The mysterious AGP family has captivated the attention of plant biologists for several decades. This diverse family of glycoproteins is widely distributed in the plant kingdom, including many algae, where they play fundamental roles in growth and development processes. The journey of AGP biosynthesis begins with the assembly of amino acids into peptide chains of proteins. An N-terminal signal peptide directs AGPs toward the endoplasmic reticulum, where proline hydroxylation occurs and a GPI anchor may be added. GPI-anchored AGPs, as well as unanchored AGPs, are then transferred to the Golgi apparatus, where extensive glycosylation occurs by the action of a variety glycosyltransferase enzymes. Following glycosylation, AGPs are transported by secretory vesicles to the cell wall or to the extracellular face of the plasma membrane (in the case of GPI-anchored AGPs). GPI-anchored proteins can be released from the plasma membrane into the cell wall by phospholipases. In this review, we present an overview of the accumulated knowledge on AGP biosynthesis over the past three decades. Particular emphasis is placed on the glycosylation of AGPs as the sugar moiety is essential to their function. Recent genetics and genomics approaches have significantly contributed to a broader knowledge of AGP biosynthesis. However, many questions remain to be elucidated in the decades ahead.

Keywords: arabinogalactan-proteins, arabinogalactan-proteinbiosynthesis; cell wall; glycosylation; glycosyltransferases; glypiation; hydroxyproline; proline hydroxylation.

Figure 1

Detailed steps of the biosynthesis of AGPs in the endoplasmic reticulum (ER). (A) The N-terminal sequence is translated on the ribosomes, allowing the entry of the AGP into the ER. The N-terminal signal is removed and the AGP mRNA continues to be translated to produce the AGP protein backbone. (B) In the ER lumen, proline (Pro) residues are converted to hydroxyproline (Hyp) residues by prolyl-4-hydroxylases (P4Hs; Pro hydroxylation) and the C-terminal GPI anchor signal sequence is removed. The arrows indicate the site of action of P4Hs. (C) The preassembled GPI anchor is attached to the ω site of the mature protein via a transamidation reaction catalyzed by the transamidase complex (GPI-T). AtAGP13 (At4g26320; NM_118765) and AtAGP42 (At1g51915; NM_104072), two AG peptides were used as models in the schematic (left and right AGP, respectively). AtAGP13 (Q9STQ3-1) and AtAGP42 (Q8L9S8-1) have the smallest amino-acid sequence predicted to have or not a GPI anchor, respectively (Showalter et al., 2010). The extended Pro hydroxylation code was applied to determine, which Pro residues are hydroxylated (Canut et al., 2016; Duruflé et al., 2017). In these cases, only Pro residues after alanine (Ala) residues were converted to Hyp. Signal peptides and C-terminal anchor addition sequence positions were determined using UniProt (The UniProt Consortium, 2019). The GPI model structure was based on the GPI of PcAGP1 (Oxley and Bacic, 1999), which consists of phosphoethanolamine attached to the protein, three mannoses, one galactose, and glucosamine-inositol linked to phosphoceramide. Created by BioRender.com.

Figure 2

Detailed steps of the biosynthesis of AGPs after transport from the endoplasmic reticulum (ER). (A) AGPs are transported to the Golgi, where type II arabinogalactan polysaccharides (AGs) are O-glycosidically linked to hydroxyproline (Hyp) residues by glycosyltransferases (GTs). (B) O-glycosylated AGPs are transported via Golgi vesicles to the cell wall, where they remain temporarily attached to the outer leaflet of the plasma membrane (PM) in the case of GPI-anchored AGPs. GPI-anchored AGPs may be released from the PM by PI-PLC phospholipase. The AGs may be cleaved by glycoside hydrolases (GHs). The arrows indicate the site of action of GTs (A), GHs and PI-PLC (B). AtAGP13 (At4g26320; NM_118765) and AtAGP42 (At1g51915; NM_104072), two AG peptides were used as models in the schematic (left and right AGP, respectively). AtAGP13 (Q9STQ3-1) and AtAGP42 (Q8L9S8-1) have the smallest amino-acid sequence predicted to have or not a GPI anchor, respectively (Showalter et al., 2010). Based on the Hyp contiguity hypothesis, non-contiguous Hyp residues are arabinogalactosylated. The GPI model structure was based on the GPI of PcAGP1 (Oxley and Bacic, 1999), which consists of phosphoethanolamine attached to the protein, three mannoses, one galactose, and glucosamine-inositol linked to phosphoceramide. Created by BioRender.com.

Figure 3

Model structure of type II arabinogalactan polysaccharides (AGs) and sites of action of known glycosyltransferases (GTs) and glycoside hydrolases (GHs) acting on AGPs. Type II AGs are O-glycosidically linked to hydroxyproline (Hyp) and consist of a β-(1→3)-linked backbone of galactose (Gal) with β-(1→6)-galactan side chains. Further modifications involve the addition of arabinose (Ara), fucose (Fuc), rhamnose (Rha), glucuronic acid (GlcA), 4-O-methylglucuronosyl (4-O-MeGlcA), and xylose (Xyl). Solid line arrows (left) represent sites of action of GTs and dotted line arrows (right) indicate sites of action of GHs. This structure is based on AGs analyzed from Arabidopsis leaves (Tryfona et al., 2012).