Solid state NMR reveals a parallel in register architecture for an infectious recombinant prion

Abstract

Two alternative architectures have been proposed for PrPSc, the most notorious prion: a parallel in register {beta} stack (PIRIBS) and a 4-rung {beta}-solenoid (4R{beta}S). We challenged these two models by measuring intermolecular 13C-13C dipole-dipole couplings of 13CO-labelled Phe residues in a fully infectious sample of recombinant bank vole PrPSc (recBVPrPSc) using a PITHIRDS-CT solid state NMR (ssNMR) experiment. To our surprise, data strongly support a PIRIBS architecture. However, the mean distance measured ([~]6.5 [A]) suggests that a minimum of two of the three Phe residues are not perfectly stacked at the canonical [~]5 [A] cross-{beta} distance. Additional ssNMR experiments show some local conformational variability of the Phe residues within limits of a relatively high rigidity. The most parsimonious interpretation of our data is that recBVPrPSc is arranged as a PIRIBS, although additional conformers with alternative architectures cannot be excluded, including a mixture of PIRIBS and 4R{beta}S. Author summaryPrPSc is the most notorious prion. It is an infectious protein that cuases fatal neurodegenerative diseases in humans and animals. PrPSc is the aberrant version of a brain protein, PrPC. PrPSc and PrPC have the same prinary structure, but different secondary, tertiaty and quaternary structures. PrPSc is capable of templating PrPC to convert to the PrPSc conformation, which is the basis of its capacity to propagate. Two plausible structural models of PrPSc have been proposed, the four-rung {beta}-solenoid (4R{beta}S) and the parallel in-register {beta} stack (PIRIBS) model. In both cases the driving force of the templating mechanism consists of "sticky" surface {beta}-strands; however, in the PIRIBS model all the {beta}-strands that conform a PrPSc monomer lie flat on a surface whereas in the 4R{beta}S model they wind in a corkscrew fashion. Here, we analyzed fully infectious recombinant PrPSc using a solid state NMR technique, PITHIRDS, that allows probing distances between specific labelled amino acid residues. To our surprise (as we have defended the 4R{beta}S model in the past), results clearly show the presence of a PIRIBS structure in our sample.