Intelectin

The first intelectin was discovered in Xenopus laevis oocyte and is named XL35 or XCGL-1.^[9][10][11] X. laevis oocyte also contains a closely related XCGL-2.^[12] In addition, X. laevis embryos secrete Xenopus embryonic epidermal lectin into the environmental water, presumably to bind microbes.^[13][14] XSL-1 and XSL-2 are also expressed in X. laevis serum when stimulated with lipopolysaccharide.^[15] Two additional intestinal intelectins are discovered in X. laevis^[16]

Human has two intelectins: hIntL-1 (omentin) and hIntL-2.^[17] Mouse also has two intelectins: mIntL-1 and mIntL-2.^[18]

Several lines of evidence suggest that intelectins recognize microbes and may function as an innate immune defense protein. Tunicate intelectin is an opsonin for phagocytosis by hemocyte.^[19] Amphioxus intelectin has been shown to agglutinate bacteria.^[20][21] In zebrafish and rainbow trout, intelectin expression is stimulated upon microbial exposure.^[22][23][24] Mammals such as sheep and mice also upregulate intelectin expression upon parasitic infection.^[25][26] Increase in intelectin expression upon microbial exposure support the hypothesis that intelectins play a role in the immune system.

Although intelectins require calcium ion for function, the sequences bear no resemblance to C-type lectins.^[3] In addition, merely around 50 amino acids (the fibronogen-like domain) align with any known protein, specifically the ficolin family.^[2] The first structural details of an intelectin comes from the crystal structure of selenomethionine-labeled XEEL carbohydrate-recognition domain (Se-Met XEEL-CRD) solved by Se-SAD.^[5] XEEL-CRD was expressed and Se-Met-labeled in High Five insect cells using a recombinant baculovirus. The fibrinogen-like fold is conserved despite amino acid sequence divergence. However, extensive insertions are present in intelectin compared to ficolins, thus making intelectin a distinct lectin structural class.^[5] The Se-Met XEEL-CRD structure then enables the structure solution by molecular replacement of D-glycerol 1-phosphate (GroP)-bound XEEL-CRD,^[5] apo-human intelectin-1 (hIntL-1),^[4] and galactofuranose-bound hIntL-1.^[4]

Each polypeptide chain of XEEL and hIntL-1 contains three bound calcium ions: two in the structural calcium site and one in the ligand binding site.^[4][5] The amino acid residues in the structural calcium site are conserved among intelectins, thus it is likely that most, if not all, intelectins have two structural calcium ions.^[5]

In the ligand binding site of XEEL and hIntL-1, the exocyclic vicinal diol of the carbohydrate ligand directly coordinates to the calcium ion.^[4][5] There are large variations in the ligand binding site residues among intelectin homologs suggesting that the intelectin family may have broad ligand specificities and biological functions.^[5] As there is no intelectin numbering conventions in different organisms, one should not assume functional homology based on the intelectin number. For example, hIntL-1 has glutamic acid residues in the ligand binding site to coordinate a calcium ion, while zebrafish intelectin-1 are devoided of these acidic residues.^[5] Zebrafish intelectin-2 ligand binding site residues are similar to those present in hIntL-1.

hIntL-1 is a disulfide-linked trimer as shown by non-reducing SDS-PAGE^[3] and X-ray crystallography.^[4] Despite lacking the intermolecular disulfide bonds, XEEL-CRD is trimeric in solution.^[5] The N-terminal peptide of the full length XEEL is responsible for dimerizing the trimeric XEEL-CRD into a disulfide-linked hexameric full-length XEEL.^[5] Therefore, the N-termini of intelectins are often responsible for forming disulfide-linked oligomer. In intelectin homologs where the N-terminal cysteines are absent, the CRD itself may still capable of forming non-covalent oligomer in solution.