Certain protein molecules in the mushroom, called enzymes, enable the chemical reactions necessary to combine or convert the substrate molecules into the desired products. The image to the right is a ribbon presentation of the P. cubensis enzyme PsiD, showing aspartate mapping and key residues in our computational PsiD AlphaFold model. The model includes the α-chain as an integral part of the antiparallel β-sheet of the catalytic domain core. The most likely candidate for completion of the non-classical catalytic triad remains E242. The modeling of the flexible N-terminus is uncertain.
Note that the depicted molecular model is a first computer model, based on similarity to other protein serine decarboxylases. Ultimately, the accurate structure has to be determined experimentally. The methods we use to determine the enzyme structures are X-ray crystallography and cryo Electron Microscopy.
As you probably can deduce from the previous page, the enzyme depicted in the second figure is PsiM, where SAM donates in two successive reactions methyl groups to norbaeocystin and then baeocystin, yielding the end product psilocybin.
While we know much about the required chemistry, we did not know how the detailed mechanism of the methyl transfer works in PsiM. Interesting parts such as the norbaeocystin binding site have no similarity to known enzyme structures and had to be determined experimentally, which we accomplished in early 2024.
a Ribbon representation of the orthorhombic crystal structure at 0.91 Å resolution (PDB 8PB4). The Rossmann fold is depicted in beige, the N-terminal jaw-like domain in blue and two helices forming a unique C-terminal extension in red. The region 189–221, which bears no homology to other methyltransferases and acts as a substrate recognition loop (SRL), is coloured green. SAH and norbaeocystin are represented as yellow and pink space-filling models, respectively. The N-terminal extension preceding the jaw domain (shown in grey) mainly consists of a TEV protease site originating from the expression vector. b Topology diagram for the PsiM fold (colours as in a). c Wall-eyed stereo figure showing OMIT density for the two ligands in the monoclinic crystal form, PDB 8PB3 (2mFo–DFc with SAH and norbaeocystin omitted simultaneously, map contoured at 1.0 σ). Also shown are functionally relevant protein residues (stick models) and key hydrogen bonds (dashed lines). An arrow indicates the direction of the nucleophilic attack.