Dual roles of the sterol recognition region in Hedgehog protein modification

Abstract

Nature provides a number of mechanisms to encode dynamic information in biomolecules. In metazoans, there exist rare chemical modifications that occur in entirely unique regimes. One such example occurs in the Hedgehog (Hh) morphogens, proteins singular across all domains of life for the nature of their covalent ligation to cholesterol. The isoform- and context-specific efficiency of this ligation profoundly impacts the activity of Hh morphogens and represents an unexplored facet of Hh ligand-dependent cancers. To elucidate the chemical mechanism of this modification, we have defined roles of the uncharacterized sterol recognition region (SRR) in Hh proteins. We use a combination of sequence conservation, directed mutagenesis, and biochemical assays to specify residues of the SRR participate in cellular and biochemical aspects of Hh cholesterolysis. Our investigations offer a functional portrait of this region, providing opportunities to identify parallel reactivity in nature and a template to design tools in chemical biology.

Fig. 1. The SRR is a helix-loop-helix motif.

a The Hint domain within full-length Hedgehog (Hh) proteins catalyzes formation of a thioester intermediate in the peptide backbone. The SRR facilitates cholesterol attack on the thioester, cleaving the Hh protein and ligating cholesterol to the last residue of the N-terminus. In a separate process, palmitate is attached to the first residue of the N-terminus after cleavage of the signal peptide. Conventional Hint domain numbering is shown; R = palmitoyl. b Sequence logo for PROMALS3D alignment (ref. ) of the SRR of 700 manually annotated Hedgehog protein sequences above the corresponding sequence of hSHH (SRR, residues 363–462). Heatmap below shows the percent helical character predicted by the JNet4 algorithm (ref. ). c Model of the SRR of hSHH (residues 363–462), created using ab initio Rosetta prediction (ref. ). d Helical wheel diagrams of the 1st and 2nd SRR helices from HELIQUEST (ref. ). The 1st helix is twisted at P379; axial views of each segment are shown. Residues are colored according to the Kyte Doolittle hydrophobicity scale (white = hydrophobic, blue = hydrophilic). e Circular dichroism (CD) spectra of peptides encompassing the 1st helix (residues 368–391) and 2nd helix (residues 431–449) in liposomes.

Fig. 2. Conserved residues in the SRR helices enable cellular hSHH cholesterolysis.

a Left: Schematic of deletion mutants lacking individual SRR secondary structure elements. Center: Western blot analysis of cell-associated hSHH-N protein (hSHH-N_L) produced by SRR deletion mutants in overexpressing HEK293T cells. Right: Plot of relative hSHH-N_L production versus wild-type protein for each mutant. b and c Alanine scanning of the 1st and 2nd helices, respectively, showing % hSHH-N_L versus wild-type produced by each mutant. Above: sequence logos for the corresponding regions (from Fig. 1b). d Monte Carlo simulation of 1^st helix residues W372-L390 in a phospholipid bilayer, showing leucine residues L382, L386, and L390 embedded in the membrane (ref. ). e Relative hSHH-N_L produced by alanine and glutamate mutants of the three predicted membrane-embedded (L382, L386, L390A) residues and one surface (L387A) residue in the 1st helix of the SRR. f Relative hSHH-N_L produced by alanine, aspartate, and phenylalanine mutants of Y435 in the 2nd helix of the SRR. For a, b, c, e, and f: The ratio of pixel intensity of hSHH-N_L to hSHH-FL for each mutant was compared to the same ratio for wild-type protein and expressed as %WT. A biological replicate for wild-type protein was analyzed in each blot. Symbols represent the mean of n = 3–10 biological replicates for each mutant ± s.d. Mutants that produced ≤50% hSHH-N_L protein relative to wild-type protein are indicated in red.

Fig. 3. Specific SRR residues control biochemical cholesterolysis and cellular localization.

a Wild-type hSHH-FL isolated from HEK293T cells shows no cleavage in 1 mM DTT, non-cholesteroylative cleavage with 50 mM DTT, and cholesteroylative cleavage with 1 mM DTT + 0.5 mM cholesterol. Cholesteroylated hSHH-N (hSHH-N_C) shows a characteristic migration shift relative to non-cholesteroylated hSHH-N, as demonstrated by protein generated from a construct expressing hSHH-N only (hSHH-N*, residues 1-197). b Cleavage and/or cholesterolysis hSHH mutants by 50 mM DTT or 1 mM DTT + 0.5 mM cholesterol. For additional blots, see Supplementary Fig. 5. c SRR model showing residues required for cholesterol modification both in cells and in vitro (violet), in cells but not in vitro (green), and Y435 (blue), which is required both in cells and in vitro but can be functionally replaced by phenylalanine. Helical wheels show the facial positions of functional residues; the 1st helix is divided into coaxial segments before and after the P379 kink. d Confocal microscopy images of HEK293T cells expressing EGFP-SRR fusion proteins. e Images of EGFP-SRR coexpressed with mCherry fused to the Golgi-targeting sequence from β4-galactosyltransferase-1 (β4Gal-T1) or the lysosome-targeting sequence from LAMP1. Scale bar = 10 µM.

Fig. 4. Models for SRR-promoted cholesterolysis of Hh proteins.

a The SRR localizes the Hint domain to a cholesterol-rich membrane interface, enabling direct access of cholesterol to the thioester. b The SRR extracts membrane cholesterol and delivers it to the thioester within the Hint domain. c The SRR interacts with the Hint domain to create a hydrophobic conduit, enabling cholesterol to access the thioester.