Evolutionary Emergence of Isoform-specific Regulation by N-glycosylation in Hyperpolarization-activated Cyclic Nucleotide-modulated (HCN) Channels

For certain ion channels in the Kv superfamily, N-glycosylation maintains stability and promotes cell surface expression; however, to our knowledge, none require it to form functional channels. To date, the role of N-glycosylation in Hyperpolarization-activated Cyclic Nucleotide modulated (HCN) channel function has been examined using only the mouse HCN2 isoform, in which, surprisingly, mutation of Asn to Gln at a predicted N-glycosylation site adjacent to the selectivity filter abolished functional expression in HEK cells. Nevertheless, other studies show that sea urchin HCN (spIH) channels are functional in HEK cells despite lacking the Asn-Xaa-Ser/Thr consensus sequon. These data raise three important questions about N-glycosylation: when in HCN evolution did it arise, do all mammalian HCN isoforms require it for function, given that they share a common ancestor with spIH, and does it affect HCN function? Here, we used phylogenetic analysis to show that invertebrates, but not chordate or urochordates, lack this N-glycosylation sequon, suggesting that it arose at a critical juncture in evolutionary time. We also show that individual mammalian HCN isoforms have distinct N-glycosylation requirements: mutation of Asn to Gln at the putative sequon renders mouse HCN2 non-functional, whereas mouse HCN1 is functionally expressed, albeit with reduced current density but minimally altered responses to voltage and cation selectivity. spIH yields robust currents but is not N-glycosylated, consistent with the absence of a predicted sequon. Taken together, these data suggest that N-glycosylation at this site emerged during chordate evolution as a regulatory mechanism that has developed uniquely for individual (uro)chordate HCN isoforms.