Biochemical and Biophysical Characterization of Carbonic Anhydrase VI from Human Milk and Saliva

Abstract

Carbonic anhydrases (CA, EC 4.2.1.1) catalyze the hydration of carbon dioxide and take part in many essential physiological processes. In humans, 15 CAs are characterized, including the only secreted isoenzyme CA VI. CA VI has been linked to specific processes in the mouth, namely bitter taste perception, dental caries, and maintenance of enamel pellicle, and implicated in several immunity-related phenomena. However, little is known of the mechanisms of the above. In this study, we characterized human CA VI purified from saliva and milk with biophysical methods and measured their enzyme activities and acetazolamide inhibition. Size-exclusion chromatography showed peaks of salivary and milk CA VI corresponding to hexameric state or larger at pH 7.5. At pH 5.0 the hexamer peaks dominated. SDS- PAGE of milk CA VI protein treated with a bifunctional crosslinker further confirmed that a majority of CA VI is oligomers of similar sizes in solution. Mass spectrometry experiments confirmed that both of the two putative N-glycosylation sites, Asn67 and Asn256, are heterogeneously glycosylated. The attached glycans in milk CA VI were di- and triantennary complex-type glycans, carrying both a core fucose and 1 to 2 additional fucose units, whereas the glycans in salivary CA VI were smaller, seemingly degraded forms of core fucosylated complex- or hybrid-type glycans. Mass spectrometry also verified the predicted signal peptide cleavage site and the terminal residue, Gln 18, being in pyroglutamate form. Thorough characterization of CA VI paves way to better understanding of the biological function of the protein.

Keywords: CA VI; CA6; Glycosylation; Mass spectrometry; Oligomerization; Size exclusion chromatography.

Fig. 1

Determination of the oligomeric state of human CA VI by SEC and crosslinking a SDS-PAGE of purified CA VI, with molecular weight markers on the left and 2.5 µg protein of each CA VI sample. Salivary CA VI, ~ 40 kDa, and milk CA VI, ~ 38 kDa. b and c Size-exclusion chromatography of salivary CA VI (solid line) and milk CA VI (dashed line). Vertical axes shows A280 × 1000. The samples were run in 50 mM Tris–HCl, pH 7.5 (b) or 0.1 M Na-acetate buffer, pH 5.0 (c). d SDS-PAGE on a 12.5% acrylamide gel of crosslinked samples, with 50-fold molar excess of DSS (over protein) and increasing concentrations of milk CA VI on lanes 1 to 6: 0.25, 0.5, 0.75, 1, 1.5 and 2.0 mg/ml, respectively. MW markers are on the left, and lane 7 shows the protein without DSS treatment. 2.5 µg of CA VI was loaded on each of lanes 1 to 7

Fig. 2

Mass spectrometry analysis of human salivary and milk CA VI a and b Total ion chromatograms (TIC) for the LC–MS/MS runs of the tryptic digests of CA VI from saliva (a) and milk (b). c and d Glycan structures discovered by MS/MS in glycopeptides of CA VI from saliva (c) and milk (d). e and f Molecular models with representations of the largest discovered glycans of salivary CA VI (e) and milk CA VI (f). White/grey: human CA VI (PDB 3FE4, oriented with the active facing away from the viewer); glycan on the right is at Asn256 and glycan on the left is at Asn67. Monosaccharides are colored as in panels c and d, with a lighter blue distinguishing the second GlcNAc (from the reducing end) in the chain, and a darker green distinguishing the β-Man in the branchpoint of the glycan antennae

Fig. 3

Sequence coverage (75%) obtained for human salivary CA VI based on MS/MS-identified peptides numbering according to UniProt P23280. N-terminal pyroglutamate (pQ18) formation was observed in the peptides. Glycosylation sites (Asn67 and Asn265) marked with arrows and observed sequence variant sites with red color

Fig. 4

Sequence coverage (86%) obtained for human milk CA VI based on MS/MS-identified peptides. Numbering according to UniProt P23280. N-terminal pyroglutamate (pQ18) formation was observed in the peptides. Glycosylation sites (Asn67 and Asn265) are marked with arrows. Variants M68L and S90G as indicated in Fig. 3 were also observed

Fig. 5

MS/MS spectra of glycopeptides in human salivary and milk CA VI. Peptide sequence identified as GLNMTGYETQAGEFPMVNNGHTVQISLPSTMR [65–96] (fragmentation sites underlined). a and b Fragmentation patterns observed for peptides from salivary CA VI containing a single GlcNAc residue, b also contains an S90G variant. c and d Peptides observed in milk CA VI containing large complex-type glycan structures, d also contains an S90G variant

Fig. 6

Comparison of human CA VI models from PDB and AlphaFold. Blue, AlphaFold model (signal peptide 1–17 not shown). Pink, PDB 3FE4. Pink surface corresponds to 3FE4, and arrows indicate the regions with most difference between 3FE4 and the AlphaFold model in the vicinity of the active site. Brown ribbon, outside the surface, is for AlphaFold model residues 21–31 which are not seen in 3FE4, and yellow ribbon is residues 18–20 which are not even present in the construct that was made for crystallization of 3FE4 (18–19 missing, and 20 replaced by Met instead of Val) (Color figure online)

Fig. 7

Molecular models of full-length human CA VI. a and b Two views from opposite sides of a glycosylated dimer model based on the AlphaFold model of human CA VI. White and khaki regions in a are residues 18–31 which are not visible in PDB 3FE4. Pink and blue glycans are attached to the pink and blue monomers, respectively. Glycosylation site Asn256 is at the foreground in a, whereas the site at Asn67 is at the foreground in b. Hydrophobic residues in the C-terminal helices are indicated in red-brown. c Two representations of a model for dimerization of the amphipathic C-terminal helix, residues 290–308, with coloring as in panels a and b. d Hypothetical ring arrangement of a hexamerix CA VI complex (Color figure online)