P-type lectins

Sequence alignment  
Interpro entries:
1) Mannose 6-phosphate receptor, binding
2) Cation-dependent mannose 6-phosphate receptor
3) Cation-independent mannose 6-phosphate receptor
4) Glucosidase II beta subunit-like
Structure of the bovine CD-MPR with bound M6P and Mn2+


Mannose 6-phosphate is shown in green and Mn2+ in purple.  Protein Data Bank structure ID: 1M6P.

Structure of domains 1-3 of the bovine CI-MPR with M6P

Mannose 6-phosphate is shown in green.  Protein Data Bank structure ID: 1SZ0.


Domain organization of proteins containing MRH domains


P-type CRDs and MRH domains

The P-type CRD was originally identified in two type I transmembrane proteins, the cation-dependent and cation-independent mannose 6-phosphate receptors (CD-MPR and CI-MPR, respectively).  The designation 'P-type' refers to the specificity of the CRD in these proteins for mannose 6-phosphate (M6P).  The CD-MPR and CI-MPR are conserved in vertebrates and are also present in evolutionarily more ancient chordate organisms such as sea squirts.  Similar proteins are present in invertebrates and fungi, and, at least in some cases, appear to perform equivalent, but non-identical, functions (see below).  Recently P-type CRD-like domains have been found in proteins with different architectures to the MPRs, and have been termed MPR homology (MRH) domains.  Some of the MRH domains in non-MPR proteins are known to have sugar-binding activity, and glycan recognition may be a general function of the MRH domain.

Mannose 6-phosphate receptors (MPRs)

The MPRs participate in the targeting of lysosomal enzymes in vertebrate animals.  Lysosomal hydrolases receive M6P tags as a sorting signal in the cis-Golgi.  This signal is generated by the addition of GlcNAc phosphate to terminal Man residues in N-linked glycans, followed by the removal of the GlcNAc moieties.  In the trans-Golgi, the MPRs bind M6P-tagged glycoproteins and transport them, in clathrin-coated vesicles, to the late endosome, where the lower pH triggers dissociation of receptor and cargo, allowing the receptors to recycle back to the Golgi.  Hydrolases which escape this system and are secreted can be recaptured by the CI-MPR, which also cycles between the cell surface and the late endosome.  Neither MPR can fully compensate for the loss of the other in sorting enzymes to the lysosome, probably because the number and arrangement of M6P residues on a glycoprotein ligand gives rise to differing affinities for the two receptors.  The CD-MPR has a luminal domain consisting of a single MRH domain, and forms dimers in which each MRH domain typically interacts with a separate glycan.  The luminal domain of the CI-MPR contains 15 MRH domains, of which numbers 3 and 9 bind M6P with high affinity (Dahms et al., 1989; domain 5 also exhibits low affinity binding to M6P).The CI-MPR is also known to bind and endocytose a variety of ligands connected with cell-cell signalling; insulin-like growth factor II, for example.  Bending of the luminal region of a CI-MPR molecule may enable it to interact simultaneously with two M6P residues on a single glycan.  Structures of the MRH domains in the CD- and CI-MPRs show that the sugar moiety of M6P is recognized by hydrogen bonds to the hydroxyl groups at positions 2, 3 and 4, the orientations of which identify the sugar as mannose.  The phosphate moiety makes hydrogen bonds with backbone atoms and interacts with a Mn2+ cation in the binding site of the CD-MPR (Roberts et al., 1998), whereas two serine side chains make hydrogen bonds to the phosphate group in the N-terminal M6P-binding MRH domain of the CI-MPR.

MPR-like proteins in non-vertebrates

The Drosophila lysosomal enzyme receptor protein (LERP) has five MRH domains and appears to perform an equivalent function to the CI-MPR in the sorting of lysosomal enzymes.  LERP does not bind M6P, but can recognize mammalian lysosomal enzymes through a different shared feature.  Similar proteins in other invertebrates and fungi may also to have roles in protein sorting that are independent of M6P.  Although M6P residues have not been found on yeast glycans, Saccharomyces cerevisiae expresses a protein, Mrl1, that is distantly related to mammalian MPRs, cycles though the late endosome, and appears to be involved in the sorting of hydrolases to the yeast vacuole.  Like LERP, Mrl1 lacks the residues identified in vertebrate MPRs as essential for M6P binding.

Other proteins containing MRH domains

MRH domains are present in four human proteins in addition to the MPRs, non of which have transmembrane regions.  An MRH domain is the only recognized domain in the gamma subunit of GlcNAc phosphotransferase.  This enzyme catalyzes the first step in the biosynthesis of M6P sorting signals in the cis-Golgi (see above) and the MRH domain may be involved in the recognition of substrate glycans.  GlcNAc phoshotransferase is present only in organisms which also possess MPRs, and raises the interesting possibility that the proteins involved in the synthesis and recognition of M6P evolved together from a common ancestor.  XTP3-B/erlectin has two MRH domains, and in Xenopus has been shown to be involved in the regulation of glycoprotein trafficking and to be essential during development.  It is present in vertebrate and invertebrate animals.  The beta subunit of ER glucosidase II contains an MRH domain preceded by two regions of coiled-coil and two complement-like repeats.  ER glucosidase II is involved in the biosynthesis of all proteins bearing N-linked glycans; it trims off the glucose residues from newly-added N-linked glycans to signify that the attached protein is correctly folded and ready to proceed to the cis-Golgi.  As in GlcNAc phosphotransferase, the MRH domain may serve to recognize the substrate glycan.  ER glucosidase II is conserved across the eukaryotic kingdom.  OS-9 is also conserved in eukaryotes.  The yeast protein, Yos9p, functions together with the M-type lectins in the ER-associated degradation (ERAD) pathway used to get rid of irreversibly misfolded glycoproteins.  OS-9 appears to identify glycoprotein targets through recognition of both N-linked glycans and other determinants.



This page last updated:
Wednesday, 01 January 2014
Animal lectins home
Contact information: This site is supported by:
Kurt Drickamer
Division of Molecular Biosciences
Faculty of Natural Sciences
Imperial College London
Email: k.drickamer@imperial.ac.uk